Method, medium, and system visually compressing image data

BACKGROUND

1. Field of the Invention

0001One or more embodiments of the present invention relate to a system of compressing and/or reconstructing image data, and more particularly, to an image encoder/decoder system suitable for Liquid Crystal Display Dynamic Capacitance Compensation (LCD DCC) devices.

2. Description of the Related Art

0002<figref idref="f0001">FIG. 1</figref> illustrates a portion of a conventional Liquid Crystal Display Dynamic Capacitance Compensation (LCD DCC) device.

0003Referring to <figref idref="f0001">FIG. 1</figref>, the conventional LCD DCC device includes a memory 11 and a Lookup Table (LUT) module 12. The LCD DCC device is used to apply a voltage higher than a required voltage to the pixels of a Thin Film Transistor (TFT)-LCD panel, in order to enhance the response time of the TFT-LCD panel. For example, if a current voltage of a pixel in the TFT-LCD panel is A and the following voltage of the pixel is B, the LCD DCC device applies the voltages to the pixel in the order of A->C->B (where C is a voltage higher than one of A and B) without applying the voltages to the pixel in the order of A->B. Hereinafter, a process in which a LCD DCC device calculates a voltage value for enhancing the response time of a TFT-LCD panel will be described, and then problems of such conventional techniques will be described.

0004Referring to <figref idref="f0001">FIG. 1</figref>, the memory 11 receives and stores information of a current picture. When the information of the current picture is received, information of the previous picture is already stored in the memory 11.

0005The LUT module 12 calculates a voltage value required to achieve a target response time of a TFT-LCD panel, with reference to a lookup table. In more detail, the LUT module 12 searches for voltage value information corresponding to a difference between the luminance value of a certain pixel of a currently received picture and the luminance value of the corresponding pixel of the previous picture stored in the memory 11, from the lookup table, and calculates a voltage value required to achieve the target response time of the TFT-LCD panel, using the voltage value information and information on the target response time of the TFT-LCD panel.

0006As described above, in order to calculate the voltage value required to achieve the target response time of the TFT-LCD panel, information of the previous picture must be stored in the memory 11. The higher the picture quality, the larger the amount of data of the picture. Accordingly, a variety of techniques for compression-storage of pictures in a memory have been proposed. Conventional image compression techniques include the JPEG (Joint Photographic Experts Group) standard, the H.264 (intra coding) standard, the JPEG-LS standard, etc. However, these standards exhibit the following problems when they are applied to LCD DCC devices.

0007First, there is a picture quality problem that occurs. When image data is compressed to 1/2-1/3 the size of the original image data, according to any one of the above-mentioned image compression methods, no significant problems occur in view of objective picture quality, e.g., measured by a Peak Signal-to-Noise Ratio (PSNR). However, in the case of the JPEG or H.264 standard based on transform coding, subjective picture quality performance deteriorates. In particular, based on a phenomenon in which the edge portion of an image appears crushed, a subjective interpretation of video quality may be that picture quality has deteriorated along the edge portions. Since LCD DCC devices drive a TFT-LCD channel that is to be shown directly to people, such subjective picture quality tends to be more important than objective picture quality in compression codecs for LCD DCC. That is, the compression codecs for LCD DCC devices typically should provide high picture quality performance so that people should not recognize the difference between images that have undergone image compression.

0008Another problem related to transform coding compression occurs when images shift in units of a pixel, e.g., with an object within the image moving, thereby shifting pixel values among neighboring pixels. Though there may be no apparent difference between an original image and its reconstructed image when the original image is shifted in a unit of each pixel, when an original image is transform coding compressed and reconstructed, a slight difference between pixel values within the same particular block has substantial influence on the reconstructed image. Accordingly, when an image is input to an LCD DCC device and shifted in a unit of a pixel, even a slight difference between pixel values will be easily recognized by users.

0009Second, since all of the above-mentioned image compression techniques are based on entropy coding, a critical path is created when images are decoded, and accordingly the complexity of a decoder significantly increases. Here, "critical path" means the longest process path among process paths that must be concurrently performed to complete all required processes. That is, "critical path" means a path requiring the longest process time among all processes currently being performed. In particular, in the case of the JPEG and H.264 standards that perform transform coding in a unit of a block, the complexity of a decoder further increases. Furthermore, since the H.264 standard uses intra prediction, the amount of memory use increases and the complexity of a decoder further increases.

0010Third, all of the above-mentioned image compression techniques allow bit rate control to some degree, however, they cannot generate an exact bit rate. For example, when the above-mentioned image compression techniques control a bit rate to achieve 1/3 compression, the bit rate will be limited within 1/3 compression due to the capacity limitation of a memory used in the corresponding LCD DCC device. For example, if bit rate control is successfully performed according to one of the above-mentioned image compression techniques, the data is compressed to less than 1/3 of the original amount. However, if bit rate control fails, the data cannot be compressed to less than 1/3 of the original amount. Therefore, the above-mentioned techniques are not suitable for image compression for LCD DCC devices which use a fixed capacity of a memory.

SUMMARY

0011One or more embodiments of the present invention provides a system, medium, and method enhancing subjective picture quality, thereby significantly lowering the complexity of an image encoder/decoder system, and providing a picture-based Control Bit Rate (CBR) suitable for a Liquid Crystal Display Dynamic Capacitance Compensation (LCD DCC) device, when images are reconstructed.

0012Additional aspects and/or advantages will be set forth in part in the description which follows and, in part, will be apparent from the description, or may be learned by practice of the invention.

0013To achieve the above and/or other aspects and advantages, embodiments of the present invention include a data compression method, including selecting a compression mode, from among a plurality of compression modes to selectively differently compress components of a data, so that a select compressed data, from a plurality of respective corresponding compressions of the data according to the plurality of compression modes, is represented by predetermined fixed lengths of bits, and outputting the select compressed data.

0014To achieve the above and/or other aspects and advantages, embodiments of the present invention include a system with data compression, including a selection unit to select a compression mode, from among a plurality of compression modes to selectively differently compress components of a data, so that a select compressed data, from a plurality of respective corresponding compressions of the data according to the plurality of compression modes, is represented by predetermined fixed lengths of bits, and a bit packeting unit to output the select compressed data.

0015To achieve the above and/or other aspects and advantages, embodiments of the present invention include a data reconstruction method for reconstructing a current data from compressed current data represented by predetermined fixed lengths of bits, including recognizing a compression mode, from among a plurality of predetermined compression modes to selectively differently compress components of the current data, from the compressed current data as a corresponding compression of the current data, and reconstructing the current data from the compressed current data based on the recognized compression mode by selectively performing one of adding the compressed current data to reference data and adding a predetermined binary value to the compressed current data.

0016To achieve the above and/or other aspects and advantages, embodiments of the present invention include system with data reconstruction, to reconstruct a current data from compressed current data represented by predetermined fixed lengths of bits, the system including a recognition unit to recognize a compression mode, from among a plurality of predetermined compression modes to selectively differently compress components of the current data, from the compressed current data as a corresponding compression of the current data, and a decoding unit to reconstruct and output the current data from the compressed current data based on the recognized compression mode by selectively performing one of adding the compressed current data to reference data and adding a predetermined binary value to the compressed current data.

BRIEF DESCRIPTION OF THE DRAWINGS

0017These and/or other aspects and advantages will become apparent and more readily appreciated from the following description of the embodiments, taken in conjunction with the accompanying drawings of which: <ul id="ul0001" list-style="none" compact="compact"><li><figref idref="f0001">FIG. 1</figref> illustrates a portion of a conventional Liquid Crystal Display Dynamic Capacitance Compensation (LCD DCC) device;</li><li><figref idref="f0001">FIG. 2</figref> illustrates a portion of a LCD DCC device/system, according to an embodiment of the present invention;</li><li><figref idref="f0002">FIG. 3</figref> is a view for explaining an example to which a Differential Pulse Code Modulation (DPCM) method performed by a data compression system, such as that illustrated in <figref idref="f0001">FIG. 2</figref>, is applied, according to an embodiment of the present invention;</li><li><figref idref="f0002">FIG. 4</figref> is a view for explaining an example to which a Pulse Code Modulation (PCM) method performed by a data compression system, such as that illustrated in <figref idref="f0001">FIG. 2</figref>, is applied, according to an embodiment of the present invention;</li><li><figref idref="f0003">FIG. 5</figref> illustrates a data format compressed by the a compression system, such as that illustrated in <figref idref="f0001">FIG. 2</figref>, according to an embodiment of the present invention;</li><li><figref idref="f0004">FIG. 6</figref> illustrates a data compression system, such as that illustrated in <figref idref="f0001">FIG. 2</figref>, according to an embodiment of the present invention;</li><li><figref idref="f0004">FIG. 7</figref> illustrates prediction directions used by an intra prediction unit, such as that illustrated in <figref idref="f0004">FIG. 6</figref>, according to an embodiment of the present invention;</li><li><figref idref="f0005">FIG. 8</figref> illustrates pixels to which the prediction directions illustrated in <figref idref="f0004">FIG. 7</figref> are applied;</li><li><figref idref="f0005">FIG. 9</figref> is a table that lists modes used by a DPCM/PCM encoding unit 63 illustrated in <figref idref="f0004">FIG. 6</figref>, according to an embodiment of the present invention;</li><li><figref idref="f0006">FIG. 10</figref> is a view for explaining a zero mode in which a data compression system, such as that illustrated in <figref idref="f0004">FIG. 6</figref>, compresses data without shifting the bits of any color components, according to an embodiment of the present invention;</li><li><figref idref="f0007">FIG. 11</figref> is a view for explaining a first mode in which a data compression system, such as that illustrated in <figref idref="f0004">FIG. 6</figref>, compresses data after shifting R and B color components by 1 bit each without shifting a G color component, according to an embodiment of the present invention;</li><li><figref idref="f0008">FIG. 12</figref> is a view for explaining a second mode in which a data compression system, such as that illustrated in <figref idref="f0004">FIG. 6</figref>, compresses data after shifting all color components by 1 bit each, according to an embodiment of the present invention;</li><li><figref idref="f0009">FIG. 13</figref> is a view for explaining a third mode in which a data compression system, such as that illustrated in <figref idref="f0004">FIG. 6</figref>, compresses data after shifting a G color component by 1 bit and shifting R and B color components by 2 bits each, according to an embodiment of the present invention;</li><li><figref idref="f0010">FIG. 14</figref> is a view for explaining a fourth mode in which a data compression system, such as that illustrated in <figref idref="f0004">FIG. 6</figref>, compresses data after shifting all color components by 2 bits each, according to an embodiment of the present invention;</li><li><figref idref="f0011">FIG. 15</figref> is a view for explaining a fifth mode in which a data compression system, such as that illustrated in <figref idref="f0004">FIG. 6</figref>, compresses data after shifting a G color component by 2 bits and shifting R and B color components by 3 bits each, according to an embodiment of the present invention;</li><li><figref idref="f0012">FIG. 16</figref> is a view for explaining a sixth mode in which a data compression system, such as that illustrated in <figref idref="f0004">FIG. 6</figref>, compresses data after shifting all color components by 3 bits each, according to an embodiment of the present invention;</li><li><figref idref="f0013">FIG. 17</figref> is a view for explaining a seventh mode in which a data compression system, such as that illustrated in <figref idref="f0004">FIG. 6</figref>, compresses data after shifting a G color component by 3 bits and shifting R and B color components by 4 bits each, according to an embodiment of the present invention;</li><li><figref idref="f0014">FIG. 18</figref> is a view for explaining an eighth mode in which a data compression system, such as that illustrated in <figref idref="f0004">FIG. 6</figref>, compresses data after shifting all color components by 4 bits each, according to an embodiment of the present invention;</li><li><figref idref="f0015">FIG. 19</figref> is a view for explaining a ninth mode in which a data compression system, such as that illustrated in <figref idref="f0004">FIG. 6</figref>, compresses data after shifting a G color component by 4 bits and shifting R and B color components by 5 bits each, according to an embodiment of the present invention;</li><li><figref idref="f0016">FIG. 20</figref> is a view for explaining a tenth mode in which a data compression system, such as that illustrated in <figref idref="f0004">FIG. 6</figref>, compresses data after shifting all color components by 5 bits each, according to an embodiment of the present invention;</li><li><figref idref="f0017">FIG. 21</figref> illustrates a mode selection unit, such as that illustrated in <figref idref="f0004">FIG. 6</figref>, according to an embodiment of the present invention;</li><li><figref idref="f0017">FIG. 22</figref> illustrates a data reconstruction system, such as that illustrated in <figref idref="f0001">FIG. 2</figref>, according to an embodiment of the present invention;</li><li><figref idref="f0018">FIG. 23</figref> is a view for explaining a process of reconstructing 3 bits of compressed data, e.g., by a DPCM/PCM decoding unit such as illustrated in <figref idref="f0017">FIG. 22</figref>, according to an embodiment of the present invention;</li><li><figref idref="f0019">FIG. 24</figref> illustrates a data compression system, such as illustrated in <figref idref="f0001">FIG. 2</figref>, according to another embodiment of the present invention;</li><li><figref idref="f0019">FIG. 25</figref> illustrates a data reconstruction system, such as illustrated in <figref idref="f0001">FIG. 2</figref>, according to another embodiment of the present invention;</li><li><figref idref="f0020">FIG. 26</figref> illustrates a data compression method, according to an embodiment of the present invention; and</li><li><figref idref="f0021">FIG. 27</figref> illustrates a data reconstruction method, according to an embodiment of the present invention.</li></ul>

DETAILED DESCRIPTION OF THE EMBODIMENTS

0018Reference will now be made in detail to embodiments, examples of which are illustrated in the accompanying drawings, wherein like reference numerals refer to the like elements throughout. Embodiments are described below to explain the present invention by referring to the figures.

0019<figref idref="f0001">FIG. 2</figref> illustrates a portion of a LCD DCC device/system, according to an embodiment of the present invention.

0020Referring to <figref idref="f0001">FIG. 2</figref>, the LCD DCC device may include a data compression system 21, a memory 22, a data reconstruction system 23, and a LUT module 24, for example.

0021The data compression system 21 may compress data, e.g., using one of two data compression methods, that is, a Differential Pulse Code Modulation (DPCM) method and a Pulse Code Modulation (PCM) method, for example.

0022The memory 22 may receive and store the compressed data from the data compression system 21. By performing this operation, when a current picture is received, the previous picture may already be stored in the memory 22.

0023The data reconstruction system 23 may reconstruct the compressed data using one of the two data compression methods, that is, the DPCM method and the PCM method, for example.

0024The LUT module 24 may calculate a voltage value required to achieve a target response time of a TFT-LCD panel, with reference to a lookup table. In more detail, the LUT module 24 may search the lookup table for voltage value information corresponding to a difference between the luminance value of a certain pixel in the currently received picture and the luminance value of the corresponding pixel in the previous picture, e.g., as reconstructed by the data reconstruction system 23, and calculate the voltage value required to achieve the target response time of the TFT-LCD panel, using the voltage value information and the target response time of the TFT-LCD panel.

0025<figref idref="f0002">FIG. 3</figref> is a view for explaining an example in which a DPCM method is performed by the data compression system 21, for example.

0026Referring to <figref idref="f0002">FIG. 3</figref>, the data compression system 21 may calculate a difference between current image data and reference data, using the DPCM method, and compress the current image data based on the calculated difference. In particular, in the example illustrated in <figref idref="f0002">FIG 3</figref>, 32 bits corresponding to 4 pixels are compressed to 17 bits.

0027<figref idref="f0002">FIG. 4</figref> is a view for explaining an example in which a PCM method is performed by the data compression system 21, for example.

0028Referring to <figref idref="f0002">FIG. 4</figref>, the data compression system 21 may merely truncate a portion of current image data, using the PCM method, to compress the current image data. In particular, in the example illustrated in <figref idref="f0002">FIG. 4</figref>, the 5 lowest significant bits of original data 63 (00111111) are truncated. As an example, 5 bits of "10000" may later be added back when the original data is reconstructed, so that data representing a value of 56 (00110000) is reconstructed in place of the original data 63 (00111111). Herein, the term "PCM method" is used to explain a technical concept that is different from the "DPCM" method, and may also, for example, further be different from the technical concept of a typical PCM method of converting analog signals into digital signals. Thus, alternative terms can be used for this PCM method described herein, e.g., a truncation compression method, noting that alternatives are equally available.

0029From the above descriptions, it should be understood that the DPCM and PCM methods, which may be used by the data compression system 21 and the data reconstruction system 23, for example, can have very low complexity compared to conventional image compression methods, such as JPEG, H.264, and JPEG-LS standards. In particular, in an embodiment, by using the DPCM and PCM methods, for example, data can be easily compressed at a constant amount. According to one embodiment, a bit rate required by LCD DCC devices/systems can be exactly achieved.

0030<figref idref="f0003">FIG. 5</figref> illustrates a data format compressed by a data compression system, such as system 21 illustrated in <figref idref="f0001">FIG. 2</figref>, according to an embodiment of the present invention. Below, and further herein, the reference to the data compression system 21 is used only as an example, and alternative data compression systems are equally available.

0031Referring to <figref idref="f0003">FIG. 5</figref>, the data compression system 21 may 1/3 compresses 96 bits of image data, for example, corresponding to a 2x2 size pixel block (hereinafter, simply referred to as a "2x2 block") made up of 4 pixels P0, P1, P2, and P3, thereby generating 32 bits of compressed data. Here, each pixel of the 2x2 block includes 8 bits of current image data corresponding to a Red (R) component, 8 bits of current image data corresponding to a Green (G) component, and 8 bits of current image data corresponding to a Blue (B) component, for example. Accordingly, in this example, each pixel data is originally represented by 24 bits, and the overall 2x2 block is represented by 96 bits.

0032The data compression system 21 may compress the 2x2 block to 4 bits of mode data and 28 bits of compressed data, for example. In more detail, in this example, 2 bits may be assigned to the 8 bits corresponding to the R component of each pixel, 3 bits may be assigned to the 8 bits corresponding to the G component of each pixel, and 2 bits may be assigned to the 8 bits corresponding to the B component of each pixel, again noting that alternative embodiments are equally available. In an embodiment, in order to achieve an exact 1/3 compression rate of 96 bits, 4 bits can be assigned to the mode data of each 2x2 block, 3 bits assigned to the G component of each pixel of the block, and 2 bits assigned to each of the R and B components of each pixel of the block.

0033As described above, in this embodiment, since the DPCM or PCM compression methods are being used with 2x2 blocks, it is possible to minimize picture quality deterioration when images shift, compared to conventional image compression methods where images are compressed in units of 16×16 macroblocks or 8x8 blocks. That is, since this embodiment compresses and/or reconstructs images in units of very small pieces, e.g., the 2x2 blocks, compared to the conventional methods, a slight difference between the values of pixels which appears in a certain block does little to substantially influence the subjective observable result obtained by compressing and reconstructing the corresponding image. Accordingly, this embodiment can significantly enhance subjective picture quality performance.

0034Embodiments of the present invention can be widely applied to image compression fields requiring low complexity and subjective lossless picture quality, other than the above-mentioned LCD DCC device/system illustrated in <figref idref="f0001">FIG. 2</figref>. For example, embodiments may be applied to image compression for Display Driver IC (DDI), reference picture compression of an image encoder/decoder system, or the like.

0035<figref idref="f0004">FIG. 6</figref> illustrates a data compression system 21, such as that illustrated in <figref idref="f0001">FIG. 2</figref>, according to an embodiment of the present invention.

0036Referring to <figref idref="f0004">FIG. 6</figref>, the data compression system 21 may include a splitter 61, an intra prediction unit 62, a DPCM/PCM encoder 63, a length check unit 64, a mode selection unit 65, a reconstruction unit 66, and a bit packeting unit 67, for example.

0037The splitter 61 may receive a current picture, divide the current picture in units of 2x2 blocks, e.g., as illustrated in <figref idref="f0003">FIG. 5</figref>, and output at least one divided 2x2 blocks to the intra prediction unit 62.

0038The intra prediction unit 62 may calculate pixel value differences between each of 4 pixels forming each 2x2 block and its neighboring pixels that are respectively positioned in four prediction directions of the pixel. In addition, in an embodiment, the intra prediction unit 62 may determine the prediction direction corresponding to a minimum value of the pixel value differences.

0039<figref idref="f0004">FIG. 7</figref> illustrates example prediction directions that may be used by the intra prediction unit 62.

0040Referring to <figref idref="f0004">FIG. 7</figref>, the intra prediction unit 62, for example, may calculate a pixel value difference between a certain pixel and its neighboring pixel, e.g., positioned in a direction among the vertical, horizontal, right-down, and left-down directions of the pixel. In particular, as illustrated in <figref idref="f0004">FIG. 7</figref>, a value indicating the vertical direction may be set to "0", a value indicating the horizontal direction may be set to "1", a value indicating the right-down direction may be set to "2", and a value indicating the left-down direction may be set to "3". Thus, in such an embodiment, the intra prediction unit 62 can represent values indicating the four prediction directions through 2 bits of binary data. Thus, the intra prediction unit 62 may output 2 bits of binary data representing the prediction direction corresponding to the minimum value of the pixel value differences, e.g., to the DPCM/PCM encoder 63.

0041<figref idref="f0005">FIG. 8</figref> illustrates example pixels to which such prediction directions illustrated in <figref idref="f0004">FIG. 7</figref> may be applied.

0042Referring to <figref idref="f0005">FIG. 8</figref>, the intra prediction unit 62, for example, may calculate corresponding differences D0, D1, D2, and D3 in vertical directions between pixels P0, P1, P2, and P3 and pixels N3, N4, P0, and P1, located in a vertical direction to the pixels P0, P1, P2, and P3, respectively, as follows.

0043<maths id="math0001" num=""><math display="block"><mi mathvariant="normal">D</mi><mo>⁢</mo><mn mathvariant="normal">0</mn><mo mathvariant="normal">=</mo><mi mathvariant="normal">P</mi><mo>⁢</mo><mn mathvariant="normal">0</mn><mo mathvariant="normal">-</mo><mi mathvariant="normal">N</mi><mo>⁢</mo><mn mathvariant="normal">3</mn></math><img file="EP1968323A2_D0001.tif" /></maths><maths id="math0002" num=""><math display="block"><mi mathvariant="normal">D</mi><mo>⁢</mo><mn mathvariant="normal">1</mn><mo mathvariant="normal">=</mo><mi mathvariant="normal">P</mi><mo>⁢</mo><mn mathvariant="normal">1</mn><mo mathvariant="normal">-</mo><mi mathvariant="normal">N</mi><mo>⁢</mo><mn mathvariant="normal">4</mn></math><img file="EP1968323A2_D0002.tif" /></maths><maths id="math0003" num=""><math display="block"><mi mathvariant="normal">D</mi><mo>⁢</mo><mn mathvariant="normal">2</mn><mo mathvariant="normal">=</mo><mi mathvariant="normal">P</mi><mo>⁢</mo><mn mathvariant="normal">2</mn><mo mathvariant="normal">-</mo><mi mathvariant="normal">P</mi><mo>⁢</mo><mn mathvariant="normal">0</mn></math><img file="EP1968323A2_D0003.tif" /></maths><maths id="math0004" num=""><math display="block"><mi mathvariant="normal">D</mi><mo>⁢</mo><mn mathvariant="normal">3</mn><mo mathvariant="normal">=</mo><mi mathvariant="normal">P</mi><mo>⁢</mo><mn mathvariant="normal">3</mn><mo mathvariant="normal">-</mo><mi mathvariant="normal">P</mi><mo>⁢</mo><mn mathvariant="normal">1</mn></math><img file="EP1968323A2_D0004.tif" /></maths>

0044Further, the intra prediction unit 62, for example, may calculate corresponding differences D0, D1, D2, and D3 in horizontal directions between the pixels P0, P1, P2, and P3 and pixels N1, P0, N0, and P2, located in a horizontal direction to the pixels P0, P1, P2, and P3, respectively, as follows.

0045<maths id="math0005" num=""><math display="block"><mi mathvariant="normal">D</mi><mo>⁢</mo><mn mathvariant="normal">0</mn><mo mathvariant="normal">=</mo><mi mathvariant="normal">P</mi><mo>⁢</mo><mn mathvariant="normal">0</mn><mo mathvariant="normal">-</mo><mi mathvariant="normal">N</mi><mo>⁢</mo><mn mathvariant="normal">1</mn></math><img file="EP1968323A2_D0005.tif" /></maths><maths id="math0006" num=""><math display="block"><mi mathvariant="normal">D</mi><mo>⁢</mo><mn mathvariant="normal">1</mn><mo mathvariant="normal">=</mo><mi mathvariant="normal">P</mi><mo>⁢</mo><mn mathvariant="normal">1</mn><mo mathvariant="normal">-</mo><mi mathvariant="normal">P</mi><mo>⁢</mo><mn mathvariant="normal">0</mn></math><img file="EP1968323A2_D0006.tif" /></maths><maths id="math0007" num=""><math display="block"><mi mathvariant="normal">D</mi><mo>⁢</mo><mn mathvariant="normal">2</mn><mo mathvariant="normal">=</mo><mi mathvariant="normal">P</mi><mo>⁢</mo><mn mathvariant="normal">2</mn><mo mathvariant="normal">-</mo><mi mathvariant="normal">N</mi><mo>⁢</mo><mn mathvariant="normal">0</mn></math><img file="EP1968323A2_D0007.tif" /></maths><maths id="math0008" num=""><math display="block"><mi mathvariant="normal">D</mi><mo>⁢</mo><mn mathvariant="normal">3</mn><mo mathvariant="normal">=</mo><mi mathvariant="normal">P</mi><mo>⁢</mo><mn mathvariant="normal">3</mn><mo mathvariant="normal">-</mo><mi mathvariant="normal">P</mi><mo>⁢</mo><mn mathvariant="normal">2</mn></math><img file="EP1968323A2_D0008.tif" /></maths>

0046Further, the intra prediction unit 62, for example, may calculate corresponding differences D0, D1, D2, and D3 in right-down directions between the pixels P0, P1, P2, and P3 and pixels N2, N3, N1, and P0, located in an lower right direction to the P0, P1, P2, and P3, respectively, as follows.

0047<maths id="math0009" num=""><math display="block"><mi mathvariant="normal">D</mi><mo>⁢</mo><mn mathvariant="normal">0</mn><mo mathvariant="normal">=</mo><mi mathvariant="normal">P</mi><mo>⁢</mo><mn mathvariant="normal">0</mn><mo mathvariant="normal">-</mo><mi mathvariant="normal">N</mi><mo>⁢</mo><mn mathvariant="normal">2</mn></math><img file="EP1968323A2_D0009.tif" /></maths><maths id="math0010" num=""><math display="block"><mi mathvariant="normal">D</mi><mo>⁢</mo><mn mathvariant="normal">1</mn><mo mathvariant="normal">=</mo><mi mathvariant="normal">P</mi><mo>⁢</mo><mn mathvariant="normal">1</mn><mo mathvariant="normal">-</mo><mi mathvariant="normal">N</mi><mo>⁢</mo><mn mathvariant="normal">3</mn></math><img file="EP1968323A2_D0010.tif" /></maths><maths id="math0011" num=""><math display="block"><mi mathvariant="normal">D</mi><mo>⁢</mo><mn mathvariant="normal">2</mn><mo mathvariant="normal">=</mo><mi mathvariant="normal">P</mi><mo>⁢</mo><mn mathvariant="normal">2</mn><mo mathvariant="normal">-</mo><mi mathvariant="normal">N</mi><mo>⁢</mo><mn mathvariant="normal">1</mn></math><img file="EP1968323A2_D0011.tif" /></maths><maths id="math0012" num=""><math display="block"><mi mathvariant="normal">D</mi><mo>⁢</mo><mn mathvariant="normal">3</mn><mo mathvariant="normal">=</mo><mi mathvariant="normal">P</mi><mo>⁢</mo><mn mathvariant="normal">3</mn><mo mathvariant="normal">-</mo><mi mathvariant="normal">P</mi><mo>⁢</mo><mn mathvariant="normal">0</mn></math><img file="EP1968323A2_D0012.tif" /></maths>

0048Still further, the intra prediction unit 62, for example, may calculate corresponding differences D0, D1, D2, and D3 in left-down directions between the pixels P0, P1, P2, and P3 and pixels N4, N5, P1, and N6, located in a lower left direction to the pixels P0, P1, P2, and P3, respectively, as follow.

0049<maths id="math0013" num=""><math display="block"><mi mathvariant="normal">D</mi><mo>⁢</mo><mn mathvariant="normal">0</mn><mo mathvariant="normal">=</mo><mi mathvariant="normal">P</mi><mo>⁢</mo><mn mathvariant="normal">0</mn><mo mathvariant="normal">-</mo><mi mathvariant="normal">N</mi><mo>⁢</mo><mn mathvariant="normal">4</mn></math><img file="EP1968323A2_D0013.tif" /></maths><maths id="math0014" num=""><math display="block"><mi mathvariant="normal">D</mi><mo>⁢</mo><mn mathvariant="normal">1</mn><mo mathvariant="normal">=</mo><mi mathvariant="normal">P</mi><mo>⁢</mo><mn mathvariant="normal">1</mn><mo mathvariant="normal">-</mo><mi mathvariant="normal">N</mi><mo>⁢</mo><mn mathvariant="normal">5</mn></math><img file="EP1968323A2_D0014.tif" /></maths><maths id="math0015" num=""><math display="block"><mi mathvariant="normal">D</mi><mo>⁢</mo><mn mathvariant="normal">2</mn><mo mathvariant="normal">=</mo><mi mathvariant="normal">P</mi><mo>⁢</mo><mn mathvariant="normal">2</mn><mo mathvariant="normal">-</mo><mi mathvariant="normal">P</mi><mo>⁢</mo><mn mathvariant="normal">1</mn></math><img file="EP1968323A2_D0015.tif" /></maths><maths id="math0016" num=""><math display="block"><mi mathvariant="normal">D</mi><mo>⁢</mo><mn mathvariant="normal">3</mn><mo mathvariant="normal">=</mo><mi mathvariant="normal">P</mi><mo>⁢</mo><mn mathvariant="normal">3</mn><mo mathvariant="normal">-</mo><mi mathvariant="normal">N</mi><mo>⁢</mo><mn mathvariant="normal">6</mn></math><img file="EP1968323A2_D0016.tif" /></maths>

0050The DPCM/PCM encoder 63 may, thus, compress current image data received from the splitter 61, using the DPCM or PCM method, according to a plurality of modes for compressing 8 bits of current image data corresponding to the respective R, G, and B components of each of 4 pixels which form a 2x2 block received from the splitter 61. That is, the DPCM/PCM encoder 63 may calculate differences between 8 bits of current image data corresponding to the respective R, G, and B components of a certain pixel, and reference image data corresponding to the respective R, G, and B components of a neighboring pixel positioned in the prediction direction decided by the intra prediction unit 62, for example, according to some modes of the plurality of modes. In addition, the DPCM/PCM encoder 63 may truncate a portion of the 8 bits of current image data corresponding to the respective R, G, and B components of the pixel, and thereby compresses the current image data, also according to particular modes. Here, the reference image data may be data reconstructed by the reconstruction unit 66, for example, though the reference image data may be original data, noting that alternatives are also available.

0051Thus, in an embodiment, the DPCM/PCM encoder 63 may compress the current image data by, e.g., in parallel, calculating differences between the current image data and the reference image data and truncating the current image data in the plurality of modes. This operation may accomplish a compressing of current image data at high speed by concurrently performing operations of the respective modes. However, in this regards, it should be understood that a method of at least sequentially performing the operations of the respective modes may also be easily implemented, again noting that alternatives are also available.

0052<figref idref="f0005">FIG. 9</figref> is a table listing example modes that may be used by a DPCM/PCM encoding unit 63, such as that illustrated in <figref idref="f0004">FIG. 6</figref>, according to an embodiment of the present invention.

0053Referring to <figref idref="f0005">FIG. 9</figref>, the DPCM/PCM encoder 63 may use a DPCM method, e.g., in particular modes 0 through 10, e.g., among 12 example modes, and then use a PCM method in a mode 11. As described above, in one embodiment, since 4 bits are assigned to mode data, 4 additional modes may be used.

0054In particular, in an embodiment, when a DPCM method is used, the DPCM/PCM encoder may 63 may selectively shift 8 bits of current image data and 8 bits of reference image data in a right direction by a predetermined number of bits depending on which of the example modes 0 through 10 is implemented, and calculate differences between the selectively shifted 8 bits of current image data and the corresponding selectively shifted 8 bits of reference image data. In this embodiment, shifting the bits of current image data and the bits of reference image data by 1 bit in a right direction means shifting the image data 1 bit in the direction of its Least Significant Bit (LSB). Accordingly, such 1-bit shifting is equivalent to dividing the 8 bits of current image data and the 8 bits of reference image data by 2, respectively. Hereinafter, a method in which the DPCM/PCM encoder 63 compresses current data in each of the example modes 0 through 10 will be described in greater detail.

0055<figref idref="f0006">FIG. 10</figref> is a view for explaining a mode 0 in which a data compression system 21, such as that illustrated in <figref idref="f0004">FIG. 6</figref>, compresses data without shifting bits of any color components.

0056Referring to <figref idref="f0004">FIGS. 6</figref> and <figref idref="f0006">10</figref>, the DPCM/PCM encoder 63 may calculate a difference between 8 bits of current image data corresponding to the G component of a pixel, e.g., among 4 pixels which form a 2x2 block received from the splitter 61, and 8 bits of reference image data corresponding to the G component of a neighboring pixel, e.g., positioned in a prediction direction decided by the intra prediction unit 62. Particularly, in an embodiment, in the case of the G component, mode 0 may be used if the difference is in a range of -4 to 3, as the difference can be represented by 3 bits as illustrated in <figref idref="f0006">FIG. 10</figref>, and no error would be generated between the original data and its reconstructed data since no bits of the original data are ignored.

0057Similarly, the DPCM/PCM encoder 63 may process the R and B components of the pixel in the same way as the processing of the G component. In particular, in an embodiment, in the case of the R and B components, this mode 0 may be used if the corresponding difference is in a range of -2 to 1, as the difference could be represented by 2 bits, e.g., as illustrated in <figref idref="f0001">FIG. 2</figref>, and no error would be generated between the original data and its reconstructed data since no bits of the original data are ignored. The DPCM/PCM encoder 63 may process the remaining pixels in the same way.

0058<figref idref="f0007">FIG. 11</figref> is a view for explaining a mode 1 in which a data compression system 21, such as that illustrated in <figref idref="f0004">FIG. 6</figref>, compresses data after shifting R and B color components respectively by 1 bit and without shifting a G color component.

0059Referring to <figref idref="f0007">FIG. 11</figref>, the DPCM/PCM encoder 63 may calculate a difference between 8 bits of current data corresponding to the G component of a pixel, e.g., among 4 pixels which form a 2x2 block received from the splitter 61, and 8 bits of reference image data corresponding to the G component of a neighboring pixel, e.g., positioned in a prediction direction decided by the intra prediction unit 62. Thus, in this embodiment, the DPCM/PCM encoder 63 may shift bits of current image data corresponding to the R component of the pixel and bits of reference image data corresponding to the R component of the neighboring pixel, positioned in the prediction direction decided by the intra prediction 62, respectively, by 1 bit, and then calculate a difference between the shifted 8 bits of current image data and the shifted 8 bits of reference image data. In addition, the DPCM/PCM encoder 63 may process the B component of the pixel in the same way.

0060Here, since the R and B components are shifted by 1 bit, this mode 1 may be used if the difference is in a range of -4 to 3, as the difference may be represented by 2 bits, as illustrated in <figref idref="f0006">FIG. 10</figref>. In addition, since 1 bit of original data is ignored, a maximum error which can be generated between the original data and its restored data becomes "1", if the ignored 1 bit is restored to "1". The DPCM/PCM encoder 63 may further process the remaining 3 pixels, except for the pixel which has just been processed in the same way.

0061<figref idref="f0008">FIG. 12</figref> is a view for explaining a mode 2 in which the data compression system 21, such as that illustrated in <figref idref="f0004">FIG. 6</figref>, compresses data after shifting all color components respectively by 1 bit.

0062Referring to <figref idref="f0008">FIG. 12</figref>, the DPCM/PCM encoder 63 may shift bits of current image data corresponding to the G component of a pixel, e.g., among 4 pixels which form a 2x2 block received from the splitter 61, and bits of reference image data corresponding to the G component of a neighboring pixel, e.g., positioned in a prediction direction decided by the intra prediction 62, respectively, by 1 bit, and calculate a difference between the 8 bits of current image data shifted by 1 bit and the 8 bits of reference image data shifted by 1 bit.

0063In particular, here, since the G component is shifted by 1 bit, this mode 2 may be used if the difference is in a range of -8 to 7, as the difference can be represented by 3 bits, as illustrated in <figref idref="f0006">FIG. 10</figref>. Since 1 bit of original data is ignored, a maximum error value which can be generated between the original data and its reconstructed data becomes "1", since the ignored 1 bit is restored to "1". The DPCM/PCM encoder 63 may process the R and B components of the pixel in the same way. Further, the DPCM/PCM encoder 63 may process the remaining 3 pixels in the same way.

0064<figref idref="f0009">FIG. 13</figref> is a view for explaining a mode 3 in which the data compression system 21, such as that illustrated in <figref idref="f0004">FIG. 6</figref>, compresses data after shifting a G color component by 1 bit and shifting R and B color components by 2 bits each.

0065<figref idref="f0010">FIG. 14</figref> is a view for explaining a mode 4 in which the data compression system 21, such as that illustrated in <figref idref="f0004">FIG., 6</figref> compresses data after shifting all color components by 2 bits each.

0066<figref idref="f0011">FIG. 15</figref> is a view for explaining a mode 5 in which the data compression system 21, such as that illustrated in <figref idref="f0004">FIG. 6</figref>, compresses data after shifting a G color component by 2 bits and shifting R and B color components by 3 bits each.

0067<figref idref="f0012">FIG. 16</figref> is a view for explaining a mode 6 in which the data compression system 21, such as that illustrated in <figref idref="f0004">FIG. 6</figref>, compresses data after shifting all color components by 3 bits each.

0068<figref idref="f0013">FIG. 17</figref> is a view for explaining a mode 7 in which the data compression system 21, such as that illustrated in <figref idref="f0004">FIG. 6</figref>, compresses data after shifting a G color component by 3 bits and shifting R and B color components by 4 bits each.

0069<figref idref="f0014">FIG. 18</figref> is a view for explaining a mode 8 in which the data compression system 21, such as that illustrated in <figref idref="f0004">FIG. 6</figref>, compresses data after shifting all color components by 4 bits each.

0070<figref idref="f0015">FIG. 19</figref> is a view for explaining a mode 9 in which the data compression system 21, such as that illustrated in <figref idref="f0004">FIG. 6</figref>, compresses data after shifting a G color component by 4 bits and shifting R and B color components by 5 bits each.

0071<figref idref="f0016">FIG. 20</figref> is a view for explaining a mode 10 in which the data compression system 21, such as that illustrated in <figref idref="f0004">FIG. 6</figref>, compresses data after shifting all color components by 5 bits each.

0072Embodiments illustrated in <figref idref="f0009 f0010 f0011 f0012 f0013 f0014 f0015 f0016">FIGS. 13 through 20</figref> have a similar technical concept as those of <figref idref="f0006 f0007 f0008">FIGS. 10 through 12</figref>, as described above, and although the numbers of bit shifting operations and modes are different from those of the embodiments illustrated in <figref idref="f0006 f0007 f0008">FIGS. 10 through 12</figref>, further detailed descriptions thereof will be omitted; the above description of <figref idref="f0006">FIGS. 10</figref> and <figref idref="f0008">12</figref> can be used for understanding <figref idref="f0009 f0010 f0011 f0012 f0013 f0014 f0015 f0016">FIGS. 13-20</figref>. In addition, here, a maximum error value that can be generated between original data and its reconstructed data may be calculated under the assumption that the MSBs of ignored bits of the original data are restored to "1" and the remaining bits of the ignored bits of the original data are restored to "0", noting that alternatives are equally available.

0073In an embodiment, the DPCM/PCM encoder 63 may truncate 5 bits of the example 8 bits of current image data corresponding to the G component, e.g., of a pixel among 4 pixels forming a 2x2 block received from the splitter 61, and thus compresses the current image data. Further, in an embodiment, the DPCM/PCM encoder 63 may truncate a maximum of 6 bits of the example 8 bits of current image data respectively corresponding to the R and B components of the pixel, and thus compresses the current image data. Here, the DPCM/PCM encoder 63 may process the remaining 3 pixels in the same way. Here, these embodiments fall under the described mode 11.

0074In one embodiment, the length check unit 64 may determine whether the difference calculated by the DPCM/PCM encoder 63, e.g., in the respective 11 modes according to the DPCM method, can be represented by a fixed length of bits. That is, here, the length check unit 64 may determine whether the difference between the 8 bits of current image data corresponding to the G component of the pixel and the 8 bits of reference image data can be represented by 3 bits, and whether the difference between the 8 bits of current image data corresponding to each of the R and B components of the pixel and the 8 bits of current image data corresponding to each of the R and B components of the pixel can be represented by 2 bits, for example.

0075For example, the length check unit 64 may process the difference calculated in mode 0 illustrated in <figref idref="f0006">FIG. 10</figref>, for example. That is, mode 0 may be used if the difference between the 8 bits of current image data corresponding to the G component and the 8 bits of reference image data corresponding to the G component, without bit shifting, is within a range of -4 to 3, and then the length check unit 64 may determine that the difference can be represented by 3 bits. Similarly, in such an example, if the difference between the 8 bits of current image data corresponding to the R component and the 8 bits of reference image data corresponding to the R component is within a range of -2 to 1, the length check unit 64 may determine that the difference can be represented by 2 bits, and, if the difference value between the 8 bits of current image data corresponding to the B component and the 8 bits of reference image data corresponding to the B component is within a range of -2 to 1, the length check unit 64 may determine that the difference can be represented by 2 bits.

0076Thus, the mode selection unit 65 may select an appropriate mode in which the maximum error value that can be generated between current image data and its reconstructed data is at a minimum, from among modes in which the difference can be represented by a fixed length of bits.

0077Here, in an embodiment, since maximum error values corresponding respectively to the R, G, and B components exist separately, the mode selection unit 65 may select the mode in which a sum of maximum error values corresponding respectively to the R, G, and B components is minimum.

0078<figref idref="f0017">FIG. 21</figref> illustrates a mode selection unit 65, such as that illustrated in <figref idref="f0004">FIG. 6</figref>, according to an embodiment of the present invention.

0079Referring to <figref idref="f0017">FIG. 21</figref>, the mode selection unit 65 may include a priority check unit 211 and a mode upgrade unit 212, for example.

0080Thus, in an embodiment, the priority check unit 211 may check priorities of the modes in which the difference can be represented by a fixed length of bits, as determined by the length check unit 64, and select the mode having the highest priority from among the modes. Here, the higher priority may be assigned to a mode when the maximum error value is least, that is, the priorities of the modes may correspond to the numbers of the modes. For example, the priority check unit 211 may identify modes 5 through 11 as where the difference can be represented by a fixed length of bits, e.g., as determined by the length check unit 64, and select mode 5 having the highest priority from among modes 5 through 11.

0081Referring to <figref idref="f0017">FIG. 21</figref>, the mode upgrade unit 212 may include a first error value calculator 2121, a DPCM corrector 2122, a second error value calculator 2123, and a comparator 2124, for example.

0082The first error value calculator 2121 may calculate a maximum error value based on the mode selected by the priority check unit 211. As only an example, if mode 5 is selected by the priority check unit 211, the first difference value calculator 2121 may calculate a sum 50 of maximum error values 16, 8, and 16 of R, G, and B components according to mode 5.

0083The DPCM corrector 2122 may further upgrade the current mode to a mode having a higher priority, e.g., by 1 level, compared to the mode selected by the priority check unit 211, and correct a portion of difference values according to the upgraded mode. In alternative embodiments, it should be understood that the DPCM corrector 2122 may, thus, upgrade the current mode to a mode having a priority higher by two or more levels, noting that alternatives are also available. As one example, if mode 5 is selected by the priority check unit 64, a difference between 8 bits of current image data corresponding to the G component of each of 4 pixels forming a 2x2 block, and 8 bits of reference image data corresponding to the G component may respectively be (10, 10, 10, 10), and a difference between 8 bits of current image data corresponding to each of the R and B components of each of the four pixels and 8 bits of reference image data corresponding to each of the R and B components may respectively be (7, 7, 8, 7), the DPCM corrector 2122 may upgrade the mode to mode 4 having a higher priority, e.g., by 1 level, compared to the originally selected fifth mode 5, and may change the difference represented by (7, 7, 8, 7) to the difference (7, 7, 7, 7), without changing the difference (10, 10, 10, 10) between the 8 bits of current image data and the 8 bits of reference image data corresponding to the G component.

0084The second error value calculator 2123 may calculate a maximum error value, e.g., according to the differences corrected by the DPCM corrector 2122. For example, the second difference value calculator 2123 may calculate a sum 25 of maximum error values 8, 8, and 8, respectively of the R, G, and B components, according to mode 4, and an error value 1, e.g., according to the difference correction of the DPCM corrector 2122.

0085The comparator 2124 may, thus, compare the maximum error value calculated by the first error value calculator 2121, for example, with the maximum error value calculated by the second error value calculator 2123, also as an example, and select a mode having a smaller error value according to the result of the comparison. For example, if the maximum error value calculated by the first error value calculator 2121 is 50 and the maximum error value calculated by the second error value calculator 2123 is 25, the comparator 2124 may selects mode 4 for having the lower maximum error value of 25.

0086Thus, in an embodiment, the reconstruction unit 66 illustrated in <figref idref="f0004">FIG. 6</figref>, for example, may add compressed data corresponding to the R, G, and B components of a pixel among 4 pixels forming a 2x2 block, according to the mode selected by the mode selection unit 65, to reference image data corresponding to the R, G, and B components of a neighboring pixel which is positioned in a prediction direction decided by the intra prediction unit 62, or add a predetermined binary value to the compressed data corresponding to the R, G, and B components of the pixel, thereby reconstructing current image data corresponding to the R, G, and B components. Here, the operation of the reconstruction unit 66 may be similar to the data reconstruction system 23 illustrated in <figref idref="f0017">FIG. 22</figref>. Accordingly, a further detailed corresponding description for the reconstruction unit 66 is omitted in consideration to the following description related to the DPCP/PCM decoder 223.

0087In addition, in an embodiment, if reference image data is equal to original data, the reconstruction unit 66 can be omitted.

0088Returning again to <figref idref="f0004">FIG. 6</figref>, the bit packeting unit 67 may generate a total of 32 bits, for example, of a data packet having 28 bits of data compressed, according to a mode selected by the mode selection unit 65, and 4 bits of data indicating the mode selected by the mode selection unit 65, among data compressed respectively according to the 12 example modes by the DPCM/PCM encoder 63, and output the 32 bits of the data packet to the memory 22. Alternatively, the bit packeting unit 67 may generate a total of 34 bits of a data packet, for example, which further includes 2 bits of data indicating a prediction direction decided by the intra prediction unit 62, other than the 28 bits of compressed data and the 4 bits of mode data.

0089Thus, <figref idref="f0017">FIG. 22</figref> illustrates a data reconstruction system 23, such as that illustrated in <figref idref="f0001">FIG. 2</figref>, according to an embodiment of the present invention.

0090Referring to <figref idref="f0017">FIG. 22</figref>, the data reconstruction system 23 may include a bit parser 221, a mode recognition unit 222, a DPCM/PCM decoder 223, and a merger 224, for example.

0091The bit parser 221 may read the 32 bits of the data packet from the memory 22, for example, parse the 32 bits of the data packet, and extract 4 bits of mode data and 28 bits of compressed data from the 32 bits of the data packet. In addition, the bit parser 221 may further output the 4 bits of mode data to the mode recognition unit 222, and output the 28 bits of compressed data to the DPCM/PCM decoder 223. Alternatively, in another embodiment, the bit parser 221 may read the 34 bits of the data packet, for example, instead of the 32 bits of the data packet, from the memory 22, and extract 4 bits of mode data, 28 bits of compressed data, and 2 bits of predicted direction data from the 34 bits of the data packet. Still further, the bit parser 221 may output the 4 bits of mode data to the mode recognition unit 222, and output the 28 bits of compressed data and the 2 bits of predicted direction data to the DPCM/PCM decoder 223.

0092Here, the mode recognition unit 222 may recognize the mode that had been selected, e.g., by the data compression unit 21, among the 12 example modes for compressing current image data corresponding to a 2x2 block, from the 4 bits of mode data received from the bit parser 221, for example.

0093The DPCM/PCM decoder 223 may reconstruct 96 bits of a 2x2 block from 28 bits of compressed data received from the bit parser 221, according to the mode recognized by the mode recognition unit 222. In an embodiment, for example, if the mode recognized by the mode recognition unit 222 is one among modes 0 through 10, that is, if the mode recognized by the mode recognition unit 222 is a mode using the DPCM method, for example, the DPCM/PCM decoder 223 may shift compressed data corresponding to each of the R, B, and G components of a pixel received from the bit parser 221, by the number of bits corresponding to the mode recognized by the mode recognition unit 222, for 4 pixels forming a 2x2 block. Then, the DPCM/PCM decoder 223 may add a binary value corresponding to the mode recognized by the mode recognition unit 222, to the compressed data subjected to bit-shifting, thereby reconstructing differences between current image data corresponding to the R, B, and G components and reference image data corresponding to the R, B, and G components.

0094In particular, the DPCM/PCM decoder 223 may shift the compressed data in a left direction, e.g., opposite to the bit-shifting direction of when the current image data was compressed. Here, shifting 4 bits of compressed data in units of 1 bit means shifting the compressed data in the direction of its MSB. Accordingly, such a shifting of 1 bit is equivalent to multiplying the value of compressed data by 2. In addition, the DPCM/PCM decoder 223 may add the reconstructed difference value to 8 bits of reference data corresponding to the R, G, and B components of a neighboring pixel which is positioned in a prediction direction, e.g., indicated by the potential 2 bits of predicted direction data received from the bit parser 221, thereby reconstructing 8 bits of current image data corresponding to the R, G, and B components.

0095In an embodiment, if the mode recognized by the mode recognition unit 222 is mode 11, that is, if the mode recognized by the mode recognition unit 222 is a mode using the PCM method, for example, the DPCM/PCM decoder 223 may add 5 bits of a binary value to the 3 bits of compressed data corresponding to the G component among the 28 bits of compressed data received from the bit parser 221, and add 6 bits of a binary value to the 2 bits of compressed data corresponding to each of the R and B components among the 28 bits of compressed data, for 4 pixels forming a 2x2 block, thereby reconstructing 8 bits of current image data corresponding to each of the R, G, and B components.

0096<figref idref="f0018">FIG. 23</figref> is a view for explaining a process of reconstructing 3 bits of compressed data, e.g., by the DPCM/PCM decoder 223 illustrated in <figref idref="f0017">FIG. 22</figref>, according to an embodiment of the present invention.

0097As indicated by the illustrated portion 231 of <figref idref="f0018">FIG. 23</figref>, and in this example embodiment, the DPCM/PCM decoder 223 may basically assign "10000" to the 3 bits of compressed data corresponding to a G component, and assign "100000" to the 3 bits of compressed data corresponding to each of the R and B components. However, due to the characteristics of image data, if the 3 bits of compressed data corresponding to the G component is "111", its original image data will be a maximum value "1111111" or approach "1111111". If the 3 bits of compressed data corresponding to the G component is "000", its original image data will be a minimum value "00000000" or approach "00000000". Accordingly, if the 3 bits of compressed data is a maximum value, among values that can be represented by 3 bits, the DPCM/PCM decoder 223 may increases the binary value that is to be assigned to the compressed data by a predetermined size, and assign the increased binary value to the compressed data. That is, if the compressed data is a maximum value (that is, "111"), among values that can be represented by 3 bits, the DPCM/PCM decoder 223 may increase the binary value that is to be assigned to the compressed data by a predetermined size, and assign the increased binary value to the compressed data. In addition, if the compressed data is a minimum value (that is, "000"), among values that can be represented by 3 bits, the DPCM/PCM decoder 223 may decrease the binary value that is to be assigned to the compressed data by a predetermined size, and assign the decreased binary value to the compressed data.

0098As indicated by the illustrated portion 232 of <figref idref="f0018">FIG. 23</figref>, if the compressed data is the maximum value, the DPCM/PCM decoder 223 may increase the binary value that is to be assigned to the compressed data by a value corresponding to the amount of bit-shifting or truncation according to the mode recognized by the mode recognition unit 222, for example, and assign the increased binary value to the compressed data. Further, if the compressed data is the minimum value, the DPCM/PCM decoder 223 may decrease a binary value that is to be assigned to the compressed data by a value corresponding to the amount of bit-shifting or truncation according to the mode recognized by the mode recognition unit 222, for example, and assign the decreased binary value to the compressed data.

0099For example, in an embodiment, if the amount of bit-shifting or truncation, e.g., according to the mode recognized by the mode recognition unit 222, is 5 bits and the compressed data is "111", the DPCM/PCM decoder 223 may increase the binary value "10000" that is to be assigned to the compressed data by 4, and assign the increased binary value "10100" to the compressed data "111", thereby reconstructing current image data to "111 10100". In addition, if the amount of bit-shifting or truncation, e.g., according to a mode recognized by the mode recognition unit 222, is 5 bits and the compressed data is "000", the DPCM/PCM decoder 223 may decrease a binary value "10000" that is to be assigned to the compressed data by 4, and assign the decreased binary value "01100" to the compressed data "111", thereby reconstructing current image data to "000 01100".

0100The merger 224 may merge the resultant total of 96 bits of a 2x2 block and thus reconstruct the corresponding picture, wherein the 2x2 block is made up of 4 pixels, and each pixel includes R, G, B components having 8 bits of current image data, as respectively reconstructed by the DPCM/PCM decoder 223.

0101<figref idref="f0019">FIG. 24</figref> illustrates a data compression system 21, such as illustrated in <figref idref="f0001">FIG. 2</figref>, according to another embodiment of the present invention.

0102Referring to <figref idref="f0019">FIG. 24</figref>, the data compression system 21 may include a splitter 241, an intra prediction unit 242, a DPCM/PCM encoder 243, a length check unit 244, a mode selection unit 245, a reconstruction unit 246, an entropy encoder 247, and a bit packeting unit 248, for example. The data compression system 21 illustrated in <figref idref="f0019">FIG. 24</figref> includes the entropy encoder 247, compared to the similar data compression system illustrated in <figref idref="f0004">FIG. 6</figref>. Accordingly, hereinafter, only the entropy encoder 247 will be further described. In addition, the bit packeting unit 248 may generate a data packet including the result of entropy-encoding by the entropy encoder 247.

0103The entropy encoder 247 may perform entropy-encoding on 32 bits of data having 28 bits of data compressed, e.g., according to the mode selected by the mode selection unit 65, and 4 bits of data indicating the mode selected by the mode selection unit 65, among data compressed according to the example 12 modes by the DPCM/PCM encoder 63. Otherwise, the entropy encoder 247 may perform entropy-encoding on 34 bits of data, which further includes 2 bits of data indicating the prediction direction decided by the intra prediction unit 242, for example, other than the 28 bits of compressed data and the 4 bits of mode data. Here, entropy-encoding means an encoding method of assigning codes to symbols so that the probabilities of the symbols are matched to the lengths of the codes. Representative entropy encoding techniques include Huffman coding, Arithmetic coding, etc.

0104<figref idref="f0019">FIG. 25</figref> illustrates a data reconstruction system 23, such as that illustrated in <figref idref="f0001">FIG. 2</figref>, according to another embodiment of the present invention.

0105Referring to <figref idref="f0019">FIG. 25</figref>, the data reconstruction system 23 may include a bit parser 251, an entropy decoder 252, a mode recognition unit 253, a DPCM/PCM decoder 254, and a merger 255, for example.

0106The data reconstruction system 23 illustrated in <figref idref="f0019">FIG. 25</figref> may also include the entropy decoder 252, compared to the similar data reconstruction system 23 illustrated in <figref idref="f0017">FIG. 22</figref>. Accordingly, hereinafter, only the entropy encoder 252 will be described.

0107The entropy decoder 252 may receive the result of the parsing from the bit parser 251, and perform entropy-decoding on the result of the parsing. In addition, the entropy decoder 252 may output the 4 bits of mode data among the result of the entropy-decoding to the mode recognition unit 222, and output 28 bits of compressed data among the result of the entropy-decoding to the DPCM/PCM decoder 223. Alternatively, for example, the entropy decoder 252 may output 4 bits of mode data among the result of the entropy-decoding, to the mode recognition unit 222, and output 28 bits of compressed data and 2 bits of predicted direction data among the result of the entropy-decoding to the DPCM/PCM decoder 223.

0108As described above, if the entropy-encoding operation is added to the data compression process, and the entropy-decoding operation is added to the data reconstruction process, since the entropy-encoding operation is a lossless compression method, it may be possible to achieve a higher data compression rate while enhancing subjective picture quality when images are reconstructed. However, when the complexity of an image encoder/decoder system increases, it is difficult to exactly meet a picture-based Control Bit Rate (CBR) suitable for a Liquid Crystal Display Dynamic Capacitance Compensation (LCD DCC) system.

0109<figref idref="f0020">FIG. 26</figref> illustrates a data compression method, according to an embodiment of the present invention.

0110Referring to <figref idref="f0020">FIG. 26</figref>, the data compression method may include operations which are sequentially processed, e.g., by the data compression system 21 illustrated in <figref idref="f0004">FIG. 6</figref>, noting that alternative embodiments are equally available. Accordingly, in such an embodiment, above descriptions about the data compression system 21 illustrated in <figref idref="f0004">FIG. 6</figref> may similarly be applied to the data compression method. Further, an operation of performing entropy-encoding, e.g., in correspondence to the data compression system 21 illustrated in <figref idref="f0019">FIG. 24</figref>, may also be included.

0111In operation 261, a picture may be received, and the picture may be divided into units of 2x2 blocks as illustrated in <figref idref="f0003">FIG. 5</figref>.

0112In operation 262, differences may be calculated between each of 4 pixels forming a 2x2 block , e.g., as divided in operation 261, and neighboring pixels which are respectively positioned in 4 prediction directions, and a prediction direction may be determined corresponding to a minimum value of the differences.

0113In operation 263, differences may be calculated between 8 bits of current image data corresponding to the respective R, G, and B components of a pixel according to example modes 0 through 10 and 8 bits of reference image data corresponding to the respective R, G, and B components of a neighboring pixel which is positioned in the prediction direction determined in operation 262, and a portion of the 8 bits of current image data corresponding to the respective R, G, and B components may be truncated according to a mode 11, for the 4 pixels forming the 2x2 block divided in operation 261, thereby compressing the current image data.

0114In operation 264, it may be determined whether the difference between the 8 bits of current image data corresponding to the G component and the 8 bits of reference image data corresponding to the G component can be represented by 3 bits, for example, and whether the difference between the 8 bits of current image data corresponding to the respective R and B components and the 8 bits of reference image data corresponding to the respective R and B components can be represented by 2 bits, also for example.

0115In operation 265, a mode can be selected corresponding to a maximum error value that can be generated between the current image data and its reconstructed data being minimum, from among modes in which the difference can be represented by a fixed length of bits.

0116In operation 266, the compressed data corresponding to the respective R, G, and B components of the pixel may be added to reference image data corresponding to the R, G, and B components of a neighboring pixel which is positioned in the prediction direction decided in operation 262, for the 4 pixels forming the 2x2 block, according to the mode selected in operation 265, or a predetermined value may be added to the compressed data corresponding to the respective R, G, and B components, thereby reconstructing the current image data corresponding to the respective R, G, and B components.

0117In operation 267, a total of 32 bits of a data packet may be generated having of 28 bits of data compressed according to the mode selected in operation 265, and 4 bits of data indicating the mode selected in operation 265, among data compressed respectively according to 12 modes in operation 263, and the 32 bits of the data packet may be output to the memory 22. Alternatively, in operation 267, a total of 34 bits of a data packet may be generated, such that the data packet further includes 2 bits of data indicating a prediction direction decided by the intra prediction unit 62, further to the 28 bits of compressed data and the 4 bits of mode data.

0118In operation 268, it may be determined whether all pixels forming a picture are completely compressed, and operation 262 may be returned to if all pixels forming a picture are not completely compressed or the process may be terminated if all pixels forming a picture are completely compressed.

0119<figref idref="f0021">FIG. 27</figref> illustrate a data reconstruction method, according to an embodiment of the present invention.

0120Referring to <figref idref="f0021">FIG. 27</figref>, the data reconstruction method includes operations that may be sequentially processed, e.g., by the data reconstruction system 23 illustrated in <figref idref="f0017">FIG. 22</figref>, noting that alternative embodiments are equally available. Accordingly, in such an embodiment, the above descriptions about the data reconstruction system 23 illustrated in <figref idref="f0017">FIG. 22</figref> may be similarly applied to the data reconstruction method.

0121In operation 271, 32 bits of a data packet may be read from the memory 22, for example, then parsed, and 4 bits of mode data and 28 bits of compressed data may be extracted from the 32 bits of the data packet. Alternatively, in operation 271, 34 bits of a data packet may be read, rather than the above-mentioned 32 bits of the data packet, from the memory 22, then parsed, and 4 bits of mode data, 28 bits of compressed data, and 2 bits of predicted direction data may be extracted from the 34 bits of the data packet.

0122In operation 272, a mode selected by the data compression system 21, for example, from among 12 modes for compressing current image data corresponding to a 2x2 block, may be recognized from the 4 bits of mode data extracted in operation 271, and operation 273 may be proceeded to if the recognized mode is one of modes 0 through 10, and operation 275 may be proceeded to if the recognized mode is mode 11, for example.

0123In operation 273, compressed data corresponding to the respective R, B, and G components of a pixel extracted in operation 271 may be shifted by the number of bits corresponding to the mode recognized in operation 272, for the 4 pixels forming the 2x2 block. Then, a binary value corresponding to the mode recognized in operation 272 may be added to the compressed data subjected to bit-shifting, thereby reconstructing differences between current image data corresponding to the respective R, G, and B components and reference image data corresponding to the respective R, G, and B components.

0124In operation 274, the differences, e.g., reconstructed in operation 273 may be added to 8 bits of reference data corresponding to the respective R, G, and B components of a neighboring pixel which is positioned in the prediction direction indicated by the 2 bits of prediction direction data extracted in operation 271, thereby reconstructing 8 bits of current image data corresponding to each of the R, G, and B components.

0125In operation 275, 5 bits of a binary value may be added to the 3 bits of compressed data corresponding to the G component among the 28 bits of compressed data received in operation 271, and 6 bits of a binary value may be added to the 2 bits of compressed data corresponding to the respective R and B components among the 28 bits of compressed data, for four pixels forming a 2x2 block, thereby reconstructing 8 bits of current image data corresponding to the respective R, G, and B components.

0126In operation 276, it may be determined whether all pixels forming a picture are completely reconstructed, and operation 271 may be returned to if all pixels forming a picture are not completely reconstructed, or the process may be terminated if all pixels forming a picture are completely reconstructed.

0127In operation 277, a total of 96 bits of 2x2 a blocks may be merged, thereby reconstructing a picture, wherein the 2x2 block is made up of 4 pixels, and each pixel includes R, G, B components which each have 8 bits of current image data and are reconstructed in operation 274 or 275.

0128In addition to the above described embodiments, embodiments of the present invention can also be implemented through computer readable code/instructions in/on a medium, e.g., a computer readable medium, to control at least one processing element to implement any above described embodiment. The medium can correspond to any medium/media permitting the storing and/or transmission of the computer readable code.

0129The computer readable code can be recorded/transferred on a medium in a variety of ways, with examples of the medium including recording media, such as magnetic storage media (e.g., ROM, floppy disks, hard disks, etc.) and optical recording media (e.g., CD-ROMs, or DVDs), and transmission media such as carrier waves, as well as through the Internet, for example. Thus, the medium may further be a signal, such as a resultant signal or bitstream, according to embodiments of the present invention. The media may also be a distributed network, so that the computer readable code is stored/transferred and executed in a distributed fashion. Still further, as only an example, the processing element could include a processor or a computer processor, and processing elements may be distributed and/or included in a single device.

0130As described above, according to one or more embodiments of the present invention, by compressing and/or reconstructing image data selectively using one of a DPCM and PCM methods, it is possible to significantly lower the complexity of an image encoder/decoder system, and, for example, exactly meet a picture-based Control Bit Rate (CBR) suitable for a Liquid Crystal Display Dynamic Capacitance Compensation (LCD DCC) device/system. In particular, according one or more embodiments of the present invention, since data is compressed or reconstructed in units of 2x2 pixel blocks, with each block made up of 4 pixels, using the DPCM method and the PCM method, it is possible to potentially 1/3-compress original data without deteriorating subjective picture quality of the original data.

0131Although a few embodiments have been shown and described, it would be appreciated by those skilled in the art that changes may be made in these embodiments without departing from the principles and spirit of the invention, the scope of which is defined in the claims and their equivalents.