Pointer tracking across multiple overlapping coordinate input sub-regions defining a generally contiguous input region.

<b>Field of the Invention</b>

The present invention relates generally to interactive input systems and in particular to a system and method for tracking a pointer across multiple overlapping coordinate input sub-regions defining a generally contiguous input region and to an interactive touch system incorporating the same.

<b>Background of the Invention</b>

Touch systems are well known in the art and typically include a touch screen having a touch surface on which contacts are made using a pointer in order to generate user input. Pointer contacts with the touch surface are detected and are used to generate corresponding output depending on areas of the touch surface where the contacts are made. Common touch systems utilize analog resistive, electromagnetic, capacitive, acoustic or machine vision to identify pointer contacts with the touch surface.

For example, International PCT Application No. PCT/CA01/00980 filed on July 5, 2001 and published under No. WO 02/03316 on January 10, 2002, assigned to SMART Technologies Inc., assignee of the present invention, discloses a camera-based touch system comprising a touch screen that includes a passive touch surface on which a computer-generated image is presented. A rectangular bezel or frame surrounds the touch surface and supports digital cameras at its corners. The digital cameras have overlapping fields of view that encompass and look across the touch surface. The digital cameras acquire images looking across the touch surface from different locations and generate image data. Image data acquired by the digital cameras is processed by digital signal processors to determine if a pointer exists in the captured image data. When it is determined that a pointer exists in the captured image data, the digital signal processors convey pointer characteristic data to a master controller, which in turn processes the pointer characteristic data to determine the location of the pointer in (x,y)-coordinates relative to the touch surface using triangulation. The pointer coordinate data is conveyed to a computer executing one or more applications programs. The computer uses the pointer coordinate data to update the computer-generated image that is presented on the touch surface. Pointer contacts on the touch surface can therefore be recorded as writing or drawing or used to control execution of applications programs executed by the computer.

Although the above touch system works extremely well, since the fields of view of the cameras are arranged to encompass the entire touch surface, camera resolution has placed a limit on the size of the touch system that can be made.

In many environments such as in teaching institutions, very large scale touch systems are desired so that visible presentations can be made to large groups. A very large scale touch system created from a series of side-by-side mounted touch panels has been considered. Although this touch system provides a larger touch surface, the touch surface is not continuous due to the individual frames surrounding the touch surfaces. Also, tracking pointer movements from one touch surface to another is cumbersome and user unfriendly. As will be appreciated, improvements in very large scale touch systems are desired.

It is therefore an object of the present invention to provide a system and method for tracking a pointer across multiple overlapping coordinate input sub-regions defining a generally contiguous input region and to an interactive touch system incorporating the same.

<b>Summary of the Invention</b>

According to one aspect of the present invention there is provided in a pointer tracking system including at least two overlapping coordinate input sub-regions defining a generally contiguous input region, each coordinate input sub-region generating pointer coordinate data in response to pointer movement therein, a method for tracking a pointer across overlapping portions of said coordinate input sub-regions comprising: <ul id="ul0001" list-style="none" compact="compact"><li>detecting pointer movements within overlapping portions of said coordinate input sub-regions; and</li><li>processing the pointer coordinate data generated by each of said coordinate input sub-regions as a result of pointer movement within said overlapping portions in accordance with defined logic to yield a single set of pointer coordinate data representing the pointer movement.</li></ul>

In one embodiment, the pointer coordinate data is combined in accordance with the defined logic. The defined logic may be a weighted averaging technique. The coordinate input sub-regions may only partially overlap. Each coordinate input sub-region generates pointer coordinate data by capturing overlapping images looking across the coordinate input sub-region, detecting the presence of a pointer in each of the captured images and triangulating the detected pointers to determine (x,y)-coordinates of the pointer.

An image generally spanning the contiguous input region may be displayed with the image being updated to reflect pointer activity. The image may include image segments with each segment being associated with a respective coordinate input sub-region. The image segments associated with adjacent coordinate input sub-regions are joined substantially seamlessly within the overlapping portions of the coordinate input sub-regions.

According to another aspect of the present invention there is provided in a touch system including a plurality of coordinate input sub-regions that overlap defining a generally contiguous touch surface, each coordinate input sub-region generating pointer coordinate data in response to pointer contacts thereon, said pointer coordinate data being processed to update image data presented on said input surface, a method of detecting the position of a pointer contact relative to said touch surface comprising: <ul id="ul0002" list-style="none" compact="compact"><li>acquiring overlapping images of each coordinate input sub-region;</li><li>when a pointer contact is made on a portion of a coordinate input sub-region that does not overlap with an adjacent coordinate input sub-region, processing acquired images to derive pointer data and triangulating the position of the pointer using the derived pointer data thereby to determine the position of the pointer contact relative to the touch surface; and</li><li>when a pointer contact is made on a portion of a coordinate input sub-region that overlaps with an adjacent coordinate input sub-region, for each coordinate input sub-region processing acquired images to derive pointer data, and triangulating positions of the pointer using the derived pointer data, and thereafter processing the triangulated positions in accordance with defined logic thereby to determine the position of the pointer contact relative to the touch surface.</li></ul>

According to yet another aspect of the present invention there is provided a touch system comprising: <ul id="ul0003" list-style="none" compact="compact"><li>a plurality of coordinate input sub-regions, said input sub-regions overlapping to define a generally contiguous touch surface, each coordinate input sub-region acquiring overlapping images thereof and generating pointer coordinate data in response to pointer contacts thereon, said pointer coordinate data being processed to update image data presented on said input surface, wherein: <ul id="ul0004" list-style="none" compact="compact"><li>when a pointer contact is made on a portion of a coordinate input sub-region that does not overlap with an adjacent coordinate input sub-region, said coordinate input sub-region processes acquired images to derive pointer data and triangulates the position of the pointer using the derived pointer data thereby to determine the position of the pointer contact relative to the touch surface; and</li><li>when a pointer contact is made on a portion of a coordinate input sub-region that overlaps with an adjacent coordinate input sub-region, each overlapping coordinate input sub-region processes acquired images to derive pointer data and triangulates the position of the pointer using the derived pointer data, the triangulated positions generated by the overlapping coordinate input sub-regions being processed in accordance with defined logic thereby to determine the position of the pointer contact relative to the touch surface.</li></ul></li></ul>

According to still yet another aspect of the present invention there is provided a touch system comprising: <ul id="ul0005" list-style="none" compact="compact"><li>a large-scale touch surface;</li><li>at least three imaging devices positioned along at least one side of said touch surface at spaced locations, each of said imaging devices looking across at least a portion of said touch surface, fields of view of said imaging devices overlapping in a manner so that each location on the touch surface falls within the fields of view of at least two imaging devices; and</li><li>processing means communicating with said imaging devices, said processing means processing image data generated by selected imaging devices capturing images of a pointer contacting said touch surface to calculate the position of the pointer contact using triangulation.</li></ul>

The imaging devices are selected based on proximity of the pointer to the imaging devices. The fields of view of the imaging devices may overlap in a manner so that many locations on the touch surface fall within the fields of view of at least three imaging devices. The processing means averages triangulation results when the pointer contact is within the fields of view of at least three imaging devices. In one embodiment, the imaging devices are laterally spaced from the touch surface and are disposed above the plane of the touch surface.

The present invention provides advantages in that pointer contacts over an input region defined by multiple overlapping coordinate input sub-regions can be tracked effectively in a user friendly manner. Also, since a transition zone is provided between adjacent overlapping coordinate input sub-regions, coordinate input events can be transferred smoothly between coordinate input sub-regions. The transition zone also increases alignment tolerances between adjacent coordinate input sub-regions.

<b>Brief Description of the Drawings</b>

Embodiments of the present invention will now be described more fully with reference to the accompanying drawings in which: <ul id="ul0006" list-style="none" compact="compact"><li>Figure 1 is a front plan view of a very large scale touch system in accordance with the present invention including an elongate rectangular, generally contiguous touch surface divided into a series of coordinate input sub-regions;</li><li>Figure 2 is a schematic block diagram of the touch system of Figure 1;</li><li>Figures 3a and 3b are front plan views of the touch surface showing the positions and orientations of cameras used to capture overlapping images looking across the touch surface;</li><li>Figures 4a and 4b are front plan views of portions of an illuminated bezel surrounding the touch surface;</li><li>Figure 5 is a flow chart showing the steps performed during orientation of the coordinate input sub-regions;</li><li>Figure 6 is a flow chart showing the steps performed during handling of pointer contacts on the touch surface;</li><li>Figure 7 is a front plan view of an alternative camera arrangement for capturing overlapping images looking across the touch surface;</li><li>Figure 8 is a front plan view of yet another alternative camera arrangement for capturing overlapping images looking across the touch surface; and</li><li>Figures 9a to 9c show different pointer contacts on the touch surface of Figure 8.</li></ul>

<b>Detailed Description of the Embodiments</b>

Turning now to Figures 1 and 2, a very large scale interactive touch system is shown and is generally identified by reference numeral 100. Touch system 100 includes an elongate generally rectangular and contiguous touch surface 102 surrounded by an illuminated bezel 104. Illuminated bezel 104 provides infrared backlighting across the touch surface 102 and is of the type described in U.S. Patent Application No. 10/354,168 to Akitt et al. filed on January 30, 2002, assigned to SMART Technologies, Inc., assignee of the present invention.

A plurality of sets 106 of cameras, in this example three sets 106 of cameras, is associated with the touch surface 102. The sets 106 of cameras are positioned along the length of the touch surface 102. Each set 106 of cameras includes four cameras arranged to define the corners of a rectangle. The cameras of each set 106 have overlapping fields of view oriented to look across a portion of the touch surface 102 thereby to define a touch or coordinate input sub-region. Thus, in the present embodiment, the sets 106 of cameras define a series of three side-by-side coordinate input sub-regions CIR₁, CIR₂ and CIR₃ respectively. The fields of view of the cameras are also oriented so that the coordinate input sub-regions defined by adjacent sets of cameras overlap to define two transition zones TZ_1/2 and TZ_2/3 respectively.

Each set 106 of cameras communicates with a master controller 108. Each master controller 108 processes pointer characteristic data received from its associated set 106 of cameras to determine the positions of pointers appearing in images captured by the cameras in (x,y)-coordinates using triangulation. The master controllers 108 transmit the pointer coordinate data to a computer 110 allowing the computer 110 either to record the pointer coordinate data as writing or drawing or use the pointer coordinate data as a mouse event to control execution of an applications program executed by the computer 110. The computer 110 provides image data to a series of projectors P₁ to P₃, which in turn project images onto the touch surface 102. The image data is updated by the computer 110 in response to received pointer coordinate data so that the images projected onto the touch surface 102 reflect the pointer activity.

Each projector is associated with a different coordinate input sub-region and projects an image thereon. As can be seen, projector P₁ projects an image I₁ onto coordinate input sub-region CIR₁, projector P₂ projects an image I₂ onto coordinate input sub-region CIR₂ and projector P₃ projects an image I₃ onto coordinate input sub-region CIR₃. The projected images I₁, I₂ and I₃ are aligned and joined seamlessly along vertical lines generally at the mid-points of the transition zones TZ_1/2 and TZ_2/3 to provide a smooth and continuous image spanning the touch surface 102. The computer 110 in this example executes a desktop application. Each coordinate input sub-region is associated with and tied to a specific section of the desktop during an orientation procedure as will be described. As a result, the computer 110 provides image data to the projectors P₁ to P₃ so that the appropriate desktop sections are displayed on the coordinate input sub-regions.

A tool tray 112 is also associated with each coordinate input sub-region. Each tool tray 112 holds a number of pointers or tools (not shown) having different assigned attributes. In this case, each tool tray holds a number of colored pens as well as an eraser. When a tool is lifted from the tool tray, the tool tray 112 provides a signal to the associated master controller 108 which in turn conveys the signal to the computer 110 to identify the selected tool. In this manner, when a colored pen is used to contact the touch surface 102 writing in the appropriate color tracking the pen movement is projected onto the touch surface. When an eraser is used to contact the touch surface 102 writing projected onto the touch surface over which the eraser is moved is erased. As is known, the desktop application can be conditioned to assign properties to pointers used to contact each coordinate input sub-region.

Turning now to Figures 1, 3a and 3b, the positions and the orientations of the cameras in the sets 106 will be further described. The set 106 of cameras associated with coordinate input sub-region CIR₁ includes cameras C₁ and C₀ positioned at the top left and bottom left corners of the touch surface 102 and cameras C₂ and C₃ positioned at the top and bottom of the touch surface 102 intermediate its length. The camera C₁ is oriented so that its optical axis is aimed generally towards the intermediate camera C₃ and forms a 45 degree angle with respect to the vertical. The camera C₀ is oriented so that its optical axis is aimed at generally towards intermediate camera C₂ and forms a 45 degree angle with respect to the vertical. The intermediate camera C₂ is oriented so that an edge of its field of view is aimed slightly towards camera C₀ and forms a 10 degree angle with respect to the vertical. The intermediate camera C₃ is oriented so that an edge of its field of view is aimed slightly towards camera C₁ and forms a 10 degree angle with respect to the vertical.

The set 106 of cameras associated with coordinate input sub-region CIR₃ includes cameras C₂" and C₃" positioned at the top right and bottom right corners of the touch surface 102 and cameras C₁" and C₀" positioned at the top and bottom of the touch surface 102 intermediate its length. The camera C₂" is oriented so that its optical axis is aimed generally towards the intermediate camera C₀" and forms a 45 degree angle with respect to the vertical. The camera C₃" is oriented so that its optical axis is aimed generally towards the intermediate camera C₁" and forms a 45 degree angle with respect to the vertical. The intermediate camera C₁" is oriented so that an edge of its field of view is aimed slightly towards camera C₃" and forms a 10 degree angle with respect to the vertical. The intermediate camera C₀" is oriented so that an edge of its field of view is aimed slightly towards camera C₂" and forms a 10 degree angle with respect to the vertical.

The set 106 of cameras associated with the coordinate input sub-region CIR₂ includes laterally spaced cameras C₁' and C₂' positioned along the top of the touch surface 102 and laterally spaced cameras C₀' and C₃' positioned along the bottom of the touch surface 102. The top left camera C₁' is located between cameras C₁ and C₂ and is oriented so that an edge of its field of view is aimed slightly towards camera C₃' and forms a 10 degree angle with respect to the vertical. The top right camera C₂' is located between cameras C₁" and C₂" and is oriented so that an edge of its field of view is aimed slightly towards camera C₀' and forms a 10 degree angle with respect to the vertical. The bottom left camera C₀' is located between cameras C₀ and C₃ and is oriented so that an edge of its field of view is aimed slightly towards camera C₂' and forms a 10 degree angle with respect to the vertical. The bottom right camera C₃' is located between cameras C₀" and C₃" and is oriented so that an edge of its field of view is aimed slightly towards the camera C₁' and forms a 10 degree angle with respect to the vertical.

The cameras at the corners and along the top and bottom of the touch surface 102 are accommodated by the illuminated bezel 104 as shown in Figures 4a and 4b. Each camera is of the type disclosed in U.S. Patent Application No. 10/384,796 to Morrison et al. filed on March 11, 2003, assigned to SMART Technologies Inc, assignee of the present invention. Each camera has a field of view slightly greater than 90° and is operable to capture images looking across the touch surface 102 at a very high frame rate. Images captured by each camera are processed on-board to determine if a pointer exists in the captured images. If a pointer exists in an image captured by a camera, pointer characteristic data is generated by the camera and is conveyed to the associated master controller 108. When the master controller 108 receives pointer characteristic data from a pair of cameras having overlapping fields of view, the master controller triangulates the pointer characteristic data to calculate the position of the pointer in (x,y)-coordinates.

Each coordinate input sub-region is divided into four quadrants using diagonal lines extending between the cameras at opposite corners of the coordinate input sub-region. Image capture to permit pointer tracking within each quadrant is the responsibility of a different pair of cameras in the set. The top quadrant Q_T is the responsibility of the bottom left and bottom right cameras in the set, the bottom quadrant Q_B is the responsibility of the top left and top right cameras in the set, the left quadrant Q_L is the responsibility of the top left and bottom left cameras in the set and the right quadrant O_R is the responsibility of the top right and bottom right cameras in the set.

Since each camera is only responsible for capturing images looking across two quadrants of the coordinate input sub-region, the field of view of each camera need only cover one half of the coordinate input sub-region. Using cameras with fields of view extending well beyond this requirement provides great flexibility with respect to orienting the cameras. For example, by tilting the intermediate cameras so that edges of their fields of view form 10° angles with respect to the vertical, the cameras provide coverage over the entire touch surface 102 while maintaining a 4:3 ratio and without requiring the cameras to intrude onto the touch surface thereby avoiding blind spots being created by the cameras. As will be appreciated, if the intermediate cameras intrude over the touch surface 102, they will block each others' view of portions of the touch surface. Also, by arranging the intermediate cameras so that edges of their fields of view form small angles with respect to the vertical, infrared lighting can be provided at the cameras that is generally in line with the infrared lighting provided by the illuminated bezels 104. This avoids dark spots from being introduced along the top and bottom of the touch surface at the intermediate camera locations.

The general operation of the touch system 100 will now be described. Initially an orientation procedure is performed by the computer 110 to calibrate the touch system 100 so that the coordinate systems of the coordinate input sub-regions can be mapped to the display coordinate systems and so that the overlapping portions of the coordinate input sub-regions within the transition zones can be mapped to one another.

With the touch system 100 calibrated, when a pointer contacts a quadrant within one of the coordinate input sub-regions outside of a transition zone, the images captured by the pair of cameras assigned to that quadrant are processed by the cameras and the associated master controller 108 in the manner described in U.S. Patent Application No. 10/294,917 to Morrison et al., assigned to SMART Technologies Inc., assignee of the present invention. In this manner, a bounding box or region surrounding the pointer contact is determined allowing the location of the pointer in (x,y)-coordinates with respect to the coordinate input sub-region to be calculated. Thus, in this case only one master controller 108 reports pointer coordinate data to the computer 110. The computer 110 in turn records the pointer coordinate data as writing or drawing if the pointer contact is a write event or injects the pointer coordinate data into the active applications program being run by the computer 110 if the pointer contact is a mouse event.

In general to determine if a pointer contact is a write or mouse event, the tool type and point of first contact is examined. If a drawing tool is used to make the contact and the contact is within a designated writing area within the projected desktop section, the pointer contact is treated as a write event; otherwise the pointer contact is treated as a mouse event. At initial contact, the pointer is given a pointer identification (ID) and the pointer ID along with any assigned pointer attributes (i.e. shape, color, width etc.) are stored.

When a pointer contacts the touch surface 102 within a transition zone, the master controllers 108 associated with the two sets of cameras that observe the transition zone generate pointer coordinates in the same manner referenced above and convey the generated pointer coordinates to the computer 110. Upon receipt of the two reported pointer coordinates, the computer 110 uses defined logic, in this case a weighted averaging technique, to yield a single (x,y)-coordinate pair representing the position of the pointer contact. The computer 110 in turn records the pointer coordinate data as writing or drawing if the pointer contact is a write event or injects the pointer coordinate data into the active applications program being run by the computer 110 if the pointer contact is a mouse event.

By using a weighted averaging technique to determine pointer positions within the transition zones, a smooth pointer transition is achieved as a pointer is moved from one coordinate input sub-region to an adjacent coordinate input sub-region. When a pointer moves from one coordinate input sub-region to an adjacent coordinate input sub-region, since the attributes assigned to the pointer are stored with the pointer ID, the attributes of the pointer are maintained by the computer 110. Alternatively, when a pointer moves from one coordinate input sub-region to an adjacent coordinate input sub-region, the properties of the pointer can be changed to properties established by the desktop section associated with the new coordinate input sub-region after a pointer-up event or a certain period of pointer inactivity occurs.

Further specifics of the orientation procedure and pointer tracking will now be described with reference to Figures 5 and 6.

<b>Orientation Procedure</b>

The orientation procedure performed by the computer 110 is similar to that described in U.S. Patent No. 5,448,263 to Martin, assigned to SMART Technologies, Inc., assignee of the present invention. Turning now to Figure 5, the steps performed by the computer 110 during orientation are shown. When the orientation procedure is launched (step 200), an orientation desktop section with targets is projected by one of the projectors onto its associated coordinate input sub-region (step 202). If the orientation desktop section is presented on the correct coordinate input sub-region, the desktop section is tied to the coordinate input sub-region and the user is prompted to contact the coordinate input sub-region at the target locations using a pointer (step 204 and 206). The pointer coordinate data generated as a result of each contact is collected allowing the coordinate system of the coordinate input sub-region to be mapped to the display coordinate system (step 208). When pointer coordinate data for each of the targets has been generated (step 210), the computer 110 checks to determine if the appropriate flag has been set to signify that the coordinate input sub-region is part of a multiple overlapping coordinate input sub-region configuration (step 212).

If the coordinate input sub-region is part of a multiple overlapping coordinate input sub-region configuration, the computer 110 checks to determine whether an oriented coordinate input sub-region exists to the right of the coordinate input sub-region being oriented on which the desktop section is displayed (step 214). If such an oriented coordinate input sub-region exists, the object representing the coordinate input sub-region being oriented is updated to include the address of the right neighbour coordinate input sub-region (step 216). The computer 110 then updates the object representing the right neighbour coordinate input sub-region to include the address of the coordinate input sub-region currently being oriented i.e. its left neighbour (step 218).

Upon completion of step 218 or if an oriented coordinate input sub-region does not exist to the right as determined at step 214, the computer 110 checks to determine whether an oriented coordinate input sub-region exists to the left of the coordinate input sub-region being oriented on which the desktop section is displayed (step 220). If such an oriented coordinate input sub-region exists, the object representing the coordinate input sub-region being oriented is updated to include the address of the left coordinate input sub-region (step 222). The computer 110 then updates the object representing the left neighbour coordinate input sub-region to include the address of the coordinate input sub-region currently being oriented i.e. its right neighbour (step 224).

Upon completion of step 224 or if an oriented coordinate input sub-region does not exist to the left as determined at step 220 or if the coordinate input sub-region is not part of a multiple overlapping coordinate input sub-region configuration as determined at step 212, the pointer coordinate data determined at step 208 is stored to registry (step 226). If the orientation procedure has not been performed with respect to the other coordinate input sub-regions, another section of the desktop with targets is presented on an adjacent coordinate input sub-region and the orientation procedure reverts to step 204 (step 228). If the orientation procedure has been performed with respect to the other coordinate input sub-regions or if the coordinate input sub-region is not part of a multiple overlapping coordinate input sub-region configuration as determined at step 212, the orientation procedure is exited.

At step 204, if the desktop section is not displayed on the proper coordinate input sub-region, the desktop section can be moved to the next coordinate input sub-region by hitting the space bar (step 230). When the space bar is hit, the computer 110 provides the output image data to the next projector so that the desktop section is presented on the adjacent coordinate input sub-region (step 232).

The pointer coordinate data stored in the registry at step 226 allows pointer contacts over the entire touch surface 102 to be mapped to the display coordinates and register the overlapping portions of the coordinate input sub-regions.

<b>Pointer Tracking</b>

When pointer coordinate data is output by a master controller 108 in response to a pointer contact on the associated coordinate input sub-region, the pointer coordinate data is conveyed to the computer 110 (step 300). In response, the computer 110 orients the pointer coordinate data to the display coordinates using the results of the orientation procedure (step 302) and then examines the pointer coordinate data to determine whether the pointer contact represents a mouse event or a write event (step 304). If the pointer contact represents a mouse event, the mouse event is examined to determine its type (step 306). If the mouse event represents a first contact with the coordinate input sub-region, a check is made to determine whether a mouse down event exists on another coordinate input sub-region within a transition zone shared by the coordinate input sub-regions (step 308). If such a mouse down event does not exist on another coordinate input sub-region, a mouse event is created (step 310) and the created mouse event for the coordinate input sub-region is stored (step 312). The stored mouse event is then injected into the active applications program running on the computer 110 (step 314). At step 308, if such a mouse down event exists, a new mouse event is created by averaging the locations of all mouse down events with the location of the current mouse event (step 316). The created new mouse event similarly is stored (step 312) and is then injected into the active applications program running on the computer 110 (step 314).

At step 306 if the mouse event represents a move contact within the coordinate input sub-region, a new mouse event is created by averaging the locations of all mouse down events with the location of the current mouse event (step 316). The created new mouse event similarly is stored (step 312) and is then injected into the active applications program running on the computer 110 (step 314).

At step 306, if the mouse event represents a remove contact from the coordinate input sub-region, the mouse data associated with the coordinate input sub-region is removed (step 320). A check is then made to determine if a mouse down event exists on another coordinate input sub-region (step 322). If not a remove mouse event is created (step 324) and the created remove mouse event is injected into the active applications program running on the computer 110 (step 314). If a mouse down condition exists on another coordinate input sub-region, the mouse event is ignored (step 326).

At step 304, if the pointer contact on the coordinate input sub-region represents a write event, the write event is examined to determine its type (step 340). If the write event represents a first contact with the coordinate input sub-region, contact down information together with the pointer (x,y)-coordinate data is stored (step 342). A check is then made to determine whether the stored information matches a pointer contact on a neighbour coordinate input sub-region (step 344). If the stored information does not match a pointer contact on a neighbour coordinate input sub-region, contact identification for the write event is created (step 346). Pointer information for the contact identification is stored (step 348). The pointer (x,y)-coordinate data in turn is recorded by the computer 110 as writing or drawing (step 350).

At step 344, if the stored information matches a pointer contact on a neighbour coordinate input sub-region signifying a pointer contact in a transition zone, contact identification and pointer information from the neighbour coordinate input sub-region is obtained (step 352). A check is then made to determine whether the pointer contact was made on the projected desktop section associated with the current coordinate input sub-region or on the projected desktop section associated with the neighbour coordinate input sub-region (step 354). If the pointer contact was made on the neighbour coordinate input sub-region, the pointer coordinate data is ignored (step 356) since the recorded pointer (x,y)-coordinate data will be processed by the neighbour coordinate input sub-region. Otherwise, the pointer coordinate data is averaged with the pointer (x,y)-coordinate data recorded by the neighbour coordinate input sub-region (step 358) prior to being recorded by the computer 110 as writing or drawing (step 350).

During averaging of the pointer coordinate data, the computer 110 processes the reported y-coordinates in accordance with defined logic to yield a single y-coordinate. Specifically, the computer 110 calculates the y-coordinate using a weighted averaging technique. The weighted averaging technique used by the computer 110 to calculate the y-coordinate from the pair of reported y-coordinates is in accordance with the following rule assuming that the pointer enters the transition zone from coordinate input sub-region CIR_x and is travelling in a direction towards coordinate input sub-region CIR_x+1:<maths id="math0001" num=""><math display="block"><mrow><msub><mrow><mtext>y-coordinate = (100 - P%) * y-coordinate of CIR</mtext></mrow><mrow><mtext>x</mtext></mrow></msub><msub><mrow><mtext> + P% * y-coordinate of CIR</mtext></mrow><mrow><mtext>x+1</mtext></mrow></msub></mrow></math><img file="EP1550940A2_D0001.tif" /></maths> where:    P% is the distance travelled through the transition zone in the x-direction expressed a percentage. As will be appreciated in this case at the first transition border i.e. the border between the coordinate input sub-region CIR_x and the transition zone, the y-coordinate is equal to the y-coordinate reported by the master controller 108 associated with coordinate input sub-region CIR_x. At the second transition border i.e. the border between the transition zone and the coordinate input sub-region CIR_x+1, the y-coordinate is equal to the y-coordinate reported by the master controller 108 associated with coordinate input sub-region CIR_x+1. At the midpoint of the transition zone, the y-coordinate is equal to the average of the y-coordinates reported by the master controllers 108 associated with the coordinate input sub-regions CIR_x and CIR_x+1.

At step 340, if the write event represents a move contact over the coordinate input sub-region, the pointer (x,y)-coordinate data is stored (step 360). A check is then made to determine if the pointer (x,y)-coordinate data matches pointer (x,y)-coordinate data generated by a neighbour coordinate input sub-region (step 362). If the pointer (x,y)-coordinate data does not match pointer (x,y)-coordinate data generated by a neighbour coordinate input sub-region, the pointer (x,y)-coordinate data is recorded by the computer 110 as writing or drawing (step 350). If the pointer (x,y)-coordinate data matches pointer (x,y)-coordinate data generated by a neighbour coordinate input sub-region, a check is made to determine whether the pointer contact was made on the projected desktop section associated with the current coordinate input sub-region or on the projected desktop section associated with the neighbour coordinate input sub-region (step 354). If the pointer contact was made on the neighbour coordinate input sub-region, the pointer coordinate data is ignored (step 356) since the recorded pointer (x,y)-coordinate data will be processed by the neighbour coordinate input sub-region. Otherwise, the pointer coordinate data is averaged with the pointer (x,y)-coordinate data recorded by the neighbour coordinate input sub-region in the manner previously described (step 358) prior to being recorded by the computer 110 as writing or drawing (step 350).

At step 340, if the write event represents a lost contact with the coordinate input sub-region representing a clear contact state (step 380), a check is made to determine whether a neighbour coordinate input sub-region has generated a matching write event (step 382). If not, a clear contact state event is generated and recorded by the computer 110 at step 350. If a neighbour coordinate input sub-region has generated a matching write event, the write event is ignored (step 384).

Although each set 106 of cameras is shown communicating with an associated master controller 108, other processing configurations to triangulate pointer data derived from captured images may be used. For example, a single master controller 108 may be used to triangulate the pointer data generated by the cameras of each set 106. In this case, it is preferred that a synchronization signal be conveyed to each set of cameras simultaneously so that images looking across the coordinate input sub-regions are captured at the same time.

Also, although the touch system 100 is described as including a projector associated with each coordinate input sub-region, it will be appreciated that other projector configurations may be used. For example, a single projector projecting an image that spans the touch surface 102 or other combinations of multiple projectors projecting overlapping images that span the touch surface 102 are suitable. In cases where multiple projectors are used, the projected images may overlap. For example, if desired, each projector may project an image spanning the entire touch surface 102, in which case each projected image fully overlaps. Alternatively, the projected images may only overlap in the transition zones. In this case, the projected images are blended at their edges within the transition zones TZ_1/2 and TZ_2/3 to provide a smooth and continuous image spanning the touch surface 102.

As will be appreciated by those skilled in the art, the overlapping coordinate input sub-regions need not be arranged in a horizontal row as shown. The coordinate input sub-regions may be arranged in a vertical column or arranged in an NxM matrix.

In the above-described example, the touch surface 102 and camera orientations are configured to maintain a 4:3 aspect ratio to conform to the shape of the coordinate input sub-regions. If the aspect ratio of the coordinate input sub-regions is equal to or less than 1, the intermediate cameras can be oriented so that the edges of their fields of view are vertical. Alternatively, in this case the orientation of the cameras can be maintained and their fields of view increased. In fact, the cameras can be oriented at virtually any angle provided their fields of view observe their assigned quadrants of the coordinate input sub-regions.

If desired, global attributes can be assigned to each pointer used to contact the touch surface 102 thereby obviating the need for tool trays associated with each coordinate input sub-region. Also, the computer 110 need not execute a desktop application that is presented in sections over the touch surface 102. Rather the computer 110 can execute and present a separate applications program on each coordinate input sub-region, while still permitting pointer properties to be carried over from one coordinate input sub-region to another.

Since the touch system 100 uses machine vision to detect pointers over the touch surface 102, the touch system 100 can support multiple users each using a pointer to contact the touch surface at different locations simultaneously. In this case writing input by different users using pointers is presented in the projected images typically using different colors or other visual differentiators although visually differentiating the user input is not necessary. When a pointer transition between coordinate input sub-regions is made resulting in multiple users contacting the same coordinate input sub-region, a decision algorithm based on proximity of last contact is used to differentiate the pointer contacts.

Turning now to Figure 7, an alternative camera arrangement for a large scale touch system is shown. In this embodiment, cameras C₁ to C₈ are only provided along the top of the touch surface 402 and look down across the touch surface. In particular, cameras are located at the top left and top right corners of the touch surface 402. Intermediate pairs of cameras are located at spaced locations along the top of the touch surface 402. The fields of view of the cameras are shown by the dotted lines. As can be seen, the fields of view of the cameras overlap so that each location on the touch surface 402 falls within the fields of view of at least two cameras. This of course allows a pointer to be tracked across the entire touch surface 402 using triangulation

Figure 8 shows yet another camera arrangement for a large scale touch system. In this embodiment, evenly spaced cameras C₁ to C₇ are positioned above the top edge of the touch surface 502 and look down across the touch surface. The fields of view of the cameras are shown by the dotted lines and as can be seen, the fields of view of the cameras overlap so that each location on the touch surface 502 falls within the fields of view of at least two cameras. Again this allows a pointer to be tracked across the entire touch surface 502 using triangulation. In fact in this embodiment, most locations on the touch surface 502 fall within the fields of view of more than two cameras allowing multiple triangulation results to be generated for each pointer contact on the touch surface. Depending on the pointer contact locations, different logic can be used to select the triangulation results to be used to determine the pointer contact location.

For example, as shown in Figure 9a, the position of pointer P on touch surface 502 can be calculated by triangulating pointer information derived from images captured by cameras C₁ and C₂ and possibly by triangulating pointer information derived from images captured by camera C₃. In this latter case pointer information derived from images captured by cameras C₁ and C₃ and cameras C₂ and C₃ can be triangulated resulting in multiple triangulation results. The multiple triangulation results can be averaged or processed according to other logic to yield a single pointer contact location. If camera C₃ is deemed to be too far from the pointer P, the camera C₃ can be ignored. Alternatively, pointer information derived from images captured by camera C₃ can be used to track the pointer to determine when the pointer reaches a certain proximity to the camera C₃. When the pointer P reaches a certain proximity to the camera C₃, the pointer information derived from images captured by camera C₃ can be triangulated to determine the position of the pointer on the touch surface 502.

Figures 9b and 9c show other positions of pointers on the touch surface 502 and the various triangulation results that can be derived from images captured by the cameras. In the example of Figure 9b, the position of pointer P on the touch surface 502 is derived from images captured by cameras C₁, C₂ and C₃. In particular, the triangulation results derived from the images captured by camera pair C₁ and C₂, camera pair C₁ and C₃ and camera pair C₂ and C₃ are averaged or processed according to other logic to yield a single pointer contact location. Camera C₄ may also be used if the pointer P is captured in images acquired thereby. If camera C₄ is deemed to be too far from the pointer P, the camera C₄ can be ignored. Alternately, pointer information derived from images captured by camera C₄ can be used to track the pointer to determine when the pointer reaches a certain proximity to camera C₄. When the pointer P reaches a certain proximity to camera C₄, the pointer information derived from images captured by camera C₄ can be triangulated to determine the position of the pointer on the touch surface 502. In this case, the triangulation results derived from the images captured by camera pair C₁ and C₄, camera pair C₂ and C₄ and camera pair C₃ and C₄ can be used in conjunction with the other triangulation results.

In the example of Figure 9c, the position of the pointer P on the touch surface 502 is derived from images captured by cameras C₂ and C₃. Cameras C₁ and C₄ may also be used if the pointer P is captured in images acquired thereby. If cameras C₁ and C₄ are deemed to be too far from the pointer P, the cameras C₁ and C₄ can be ignored. Alternatively, pointer information derived from images captured by cameras C₁ and C₄ can be used to track the pointer to determine when the pointer reaches a certain proximity to the cameras C₁ and C₄. When the pointer P reaches a certain proximity to one or more of the cameras C₁ and C₄, the pointer information derived from images captured by the cameras can be triangulated to determine the position of the pointer on the touch surface 502. In particular, the triangulation results derived from the images captured by camera pair C₁ and C₂, camera pair C₁ and C₃, camera pair C₁ and C₄, camera pair C₂ and C₃, camera pair C₂ and C₄ and camera pair C₃ and C₄ may be averaged or processed according to other logic to yield a single pointer contact location. As will be appreciated, the areas on the touch surface 502 falling within the fields of view of different pairs of cameras are similar to the transition zones described with reference to the embodiment of Figures 1 to 6.

Those of skill in the art will appreciate that although the above embodiments show vision-based touch systems, the present method of tracking a pointer can be used in other types of touch systems having overlapping input regions defining transition zones.

Although embodiments of the present invention have been described with reference to the figures, those of skill in the art will appreciate that variations and modifications may be made without departing from the spirit and scope thereof as defined by the appended claims.