Method for informative description of image objects

Technical field

The invention relates to processes for informative description of image objects of Image comprising the steps of:<sl><li>(A) detecting and digitizing the image as digital pixel matrix,</li><li>(B) storing the image in a memory of a computer unit.</li></sl>

State of the art

Images can be recorded in various ways permanently. Thus, the Possibility to produce an image with conventional photography, ie with films. These images may be digitized using a scanner and a computer in are processed. It is now, due to the advanced CCD technology (= Charge-Coupled Device) possible with so-called digital cameras digital images to capture. In a further development of these digital cameras are equipped with mobile terminals combined. Thus, digital images directly, for example as an MMS (= Multimedia Messages Service) are sent over a cellular network to a third party. MMS is a standard for mobile terminals for the transmission of multimedia News.

The so acquired digital images are stored in a memory of the digital camera or Computer than before pixel graphics. Each pixel or pixels of such pixel graphics consisting of a digital number representing a particular color. The composite pixels finally give the image. The more pixels per Surface are present, the higher the resolution of the image.

Through ever faster computers and ever more sophisticated algorithms Objects can a digital image are detected. Such image recognition is performed For example, in the character recognition (OCR = Optical Character Recognition - opt. Character recognition). However, other objects as letters can be based on recognize significant points or lines. The image recognition is a discipline that in many areas protrudes. Digital image recognition is z. B. also in computerized machines (robots), especially in the automotive industry required which must hold different objects apart if necessary, or z. B. in the automatic flight navigation for aircraft and missiles to Landscapes of surface navigation. The application possibilities of the Image recognition are already very diverse. Neural networks help of image recognition the ability to learn, so that machines learn to recognize certain objects.

Often discover travelers on their journeys objects such. As buildings, squares, Landscapes etc., of which they do not know what it is for objects, and which history behind it. If you do not happen to have a guide here have, they will hardly be able to guess what it is the object. For the traveler, it is desirable as possible to without major complications To get information about these objects.

On the Internet (World Wide Web), such information on certain properties are easily available. The disadvantage here is not to access the Internet easily can be.

Disclosure of the Invention

The object of the invention is therefore to provide a method of the disadvantages Avoids prior art and allows someone to information from Objects of an image at any time and to get to any place.

The object is a method for informative description of image objects an image of the aforementioned type with the following further method step, dissolved:<sl><li>(C) image recognition to identify the image objects and assigning Terms from a database the image objects in the image.</li></sl>

The inventive method is based on the principle, based on digital images identify individual image objects with an image recognition and image these objects then subsequently assign terms. The necessary terms for the individual Objects are in a database that can be queried to.

In an advantageous embodiment of the inventive method is carried out, the Capture of the image by a camera, in particular digital camera. The pictures are especially in the digital camera in a storable form. let therefore easy to process by this measure for image recognition.

A preferred mode of the invention arises when the Image recognition to identify the image objects with the satellite-based Navigation System (GPS) is made. This measure is obtained additional information of an object in an image, making it easier to can be identified when the position of the object is already known.

Advantageously, the informative description on an output unit, in particular a display output. This can instantly collected Information is viewable via an object.

The Internet is a network with a vast inventory of data and Information in all variations on a variety of objects. A special Training of the method is obtained, therefore, when the informative Description of the objects, especially through search engines, from the Internet is based. In this way, this vast database, as otherwise in no database is found, for the informative description of the objects used.

Depending on requirements, it may also be advantageous if the informative description of a database of a computer unit is based. This is z. B. advantageous when not accurate enough are present, for example, from the Internet data. Then can also exact or shortened information from a database of a computer be used.

In a further preferred embodiment of the inventive method is carried out, the Image recognition by vectorization of digital pixel matrix to identify the Picture objects of the image. By Vektoriesierung the dataset is a significantly reduces image object and can be significantly in a computer unit easier to identify an object in an image process.

Mobile devices are now widely used. They can be within the Mobile Network cell of a cellular network locate. An advantageous embodiment The invention therefore provides that the image recognition to identify the image objects the mobile network cells is performed. Someone of the information about a wishes object, his mobile terminal has to basically be in addition this Bring object to how the GPS to track the location. These Location information helps considerably in the identification of an object.

Advantageously, in another embodiment of the invention, the camera provided acquisition of the image in a mobile terminal, or with a coupled mobile terminal. In this manner is firstly the image with the Objects which are to be recognized before and on the other hand the location information of the the mobile radio cells of the mobile network. Furthermore, can be over the air easily create a connection to a computer with an image recognition, the Computer work is done. The image must be only sent to the computer, which the Performs image recognition for identifying objects. The sending of Image information can be, for example, via standardized MMS (Multimedia = Messages Service) perform. This calculator is preferably provided with a database or connected to the Internet.

Preferably carried the image recognition using neural networks. This has the Advantage that once trained objects are always recognized. Moreover let also similar properties to determine precisely and assign an information better.

Further advantages arise from the subject matter of the claims and from the Drawing with the associated description.

Brief Description of Drawing

<dl tsize="6" compact="compact"><dt>Fig. 1</dt><dd>shows a schematic diagram of one embodiment of a principle Inventive method.</dd></dl>

Preferred Embodiment

In FIG. 1 is designated 10 a landscape in which hinted a city is seen. The landscape 10 contains different objects 12, 14, 16, 18, 20, 22, 24. In the present embodiment is in the objects 12, 14, 16, 18, 20, 22, 24 to buildings. Some buildings 18, 20, 22, 24 are somewhat closer together, other 12, 14, 16 slightly further apart. This landscape is a 10 Digital camera recorded 26, which is indicated with projection rays 27th The Landscape 10 is projected to in a lens 28 of the digital camera 26th Behind the Lens 28 is a CCD sensor 30, of the landscape 10 as a digital image 32 captured.

The digital image 32 is then present as pixel graphics and is in a data memory 34 stored. Each pixel or pixels of such pixel graphics consists of a digital number representing a particular color. The composite Pixels finally give the image. The more pixels per area are available, the higher the resolution of the digital image 32. From the data memory 34 from may be the digital image processed 32 or even deleted.

In the present embodiment, the digital camera 26 is provided with a Mobile terminal 36 is coupled, ie digital camera 26 and mobile terminal 36 can exchange data, in particular data of the digital image 32nd Arrows 37 indicate the data exchange between the digital camera 26 and the Mobile terminal to 36th In modern mobile stations 36, the digital camera is 26 integrated into the mobile station 36th The mobile station 36 has in addition a Operator panel 38 also a display 40. The display 40 serves with the Operator field 38 the mobile terminal 36 to control. This is usually at a suitable menu navigation performed. Further, on the display 40, the digital image represented the 32nd The menu is not shown, and the operation panel 38 also serve a particular image object 12, for example, the building 12 (a select church) from the digital image 32 to further information on this Image object to obtain 12a.

The digital image 32 is to means of mobile terminal 36 over a cellular network 42 transmitted to a server 44th The mobile network 42 is shown schematically as Funkmast illustrated and operates for example according to the GSM or UMTS standard. The Sending the digital image 32 by radio and landline is with arrows 46, 48 symbolizes. The server 44 represents virtually all of the administrative and Mediation of digital data that are incurred in the Mobile Network 42nd

The server 44 provides the digital image data 32 to an image recognition unit 50, arrow 51. The image recognition unit 50 consists of a computer, in which a suitable program for image recognition expires. To modify the program The image recognition, the digital image 32 is in an intermediate memory 53 of the Image recognition comply with 50 saved. The program selected for image recognition first the image objects 12a-24a, which recognizes it as such. can for image recognition be necessary that the objects that need to collect it, will vektoriesiert. In addition, can be useful in an image in the determination of an object, in which image is connected with other objects. Subsequently, the Data of the objects 12-24 is compared with the data of a database 52nd To the query to limit the data set be coordinates of the Mobile terminal 36 requested that the example of the mobile radio cell Cellular network 42 in which the mobile station is 36, give. It can also, if a corresponding GPS receiver in the digital camera 26 or in the Mobile terminal 36 is containing the exact coordinates are queried. when an object is detected by name, it is a term associated with what with box represented 54th In addition, this object named now conceptually can can be appended to, information. This information can the object describe. For this purpose, the largest existing data, namely the Internet 56 resorted. With sufficient capacity can also movies, explaining an object, be used instead of informative text.

Arrows 58, 60 indicate the bidirectional exchange of data between the database 52, and Internet and image recognition unit 50 at. Arrow 62 symbolizes how the Internet the detected object, arrow 64, a description or information in box 54 is assigned.

If an object can now be named with a phrase, a description or information has been assigned, the corresponding data on the be Server 44, arrow 66 and the mobile phone network 42 back Mobilfünkendgerät 36 transmitted. The display 40 can now be the desired information to the selected retrieve objects of the digital image 32nd