Methods, devices and computer programs for processing images in a system comprising a plurality of cameras

Patent metadata

Field	Value
Publication number	US-2016358032-A1
Application number	US-201615170647-A
Country	US
Kind code	A1
Filing date	Jun 1, 2016
Priority date	Jun 4, 2015
Publication date	Dec 8, 2016
Grant date	—

Abstract

The present disclosure relates to processing images representing a real scene, the images comprising at least two images obtained from at least two sources of images, a sub-image of each of the at least two images representing an overlapping area of the real scene. A role map comprising a plurality of parts, all the points of each part of the role map being associated with a single identifier of a source of images, is generated during a training phase. To process the images, parts of the role map are selected as a function of an identifier of a considered source of images, and sub-images of an image obtained from the considered source of images are processed, the processed sub-images corresponding to the selected parts of the role map.

First claim

1 . A method for optimizing processing of at least two images obtained from at least two sources of images, a sub-image of each of the at least two images representing an overlapping area of a real scene, the method comprising:
generating a map representing the overlapping area of the real scene; partitioning the generated map into a plurality of sub-areas as a function of an occlusion criterion determined for points of the overlapping area; associating an identifier of a source of images with points of the generated map, the associating step being based on the occlusion criterion or on a detection score depending on the sub-areas; and generating a role map based on the generated map and on the associated identifiers, the role map comprising a plurality of parts, all the points of each part of the role map being associated with a single identifier of a source of images so that only one of the at least two sources of images is to be selected to process a representation of a portion of the overlapping area.

2 . The method of claim 1 , wherein an identifier of a considered source of images is associated with a considered point of a sub-area of the overlapping area if no occlusion is detected for the considered point in an image obtained from the considered source of images.

3 . The method of claim 1 , wherein an identifier of a source of images is associated with:
a considered point of a sub-area of the overlapping area as a function of a detection score criterion associated with each source of images of a set of sources of images if no occlusion is detected for the considered point in an image obtained from each source of images of the set of sources of images; or a considered point of a sub-area of the overlapping area as a function of a detection score criterion if an occlusion is detected for the considered point in an image obtained from each of the sources of images providing images comprising a sub-image representing the overlapping area of the real scene.

4 . The method of claim 1 , further comprising preliminary steps of determining occlusion criterions, an occlusion criterion being associated with a point of the overlapping area and with a source of images.

5 . The method of claim 4 , wherein a preliminary step of determining an occlusion criterion is carried out for each point of the overlapping area.

6 . The method of claim 1 , further comprising a step of computing detection scores, a detection score being associated with a point of the overlapping area and with a source of images.

7 . The method of claim 1 , further comprising:
selecting parts of the role map as a function of an identifier of the considered source of images; and processing sub-images of an image obtained by the considered source of images, the processed sub-images corresponding to the selected parts of the role map.

8 . The method of claim 7 , further comprising:
a step of projecting the selected parts of the role map into a reference frame associated with the images obtained by the considered source of images; and a step of extracting the sub-images to be processed, the extracted sub-images corresponding to the projected parts of the role map.

9 . A computer-readable storage medium storing instructions of a computer program for causing a computer to implement the method according to claim 1 .

10 . A device for optimizing processing of at least two images obtained from at least two sources of images, a sub-image of each of the at least two images representing an overlapping area of a real scene, the device comprising a processor configured for carrying out the steps of:
generating a map representing the overlapping area of the real scene; partitioning the generated map into a plurality of sub-areas as a function of an occlusion criterion determined for each point of the overlapping area; associating an identifier of a source of images with each point of the generated map, the associating step being based on the occlusion criterion or on a detection score depending on the sub-areas; and generating a role map based on the generated map and on the associated identifiers, the role map comprising a plurality of parts, all the points of each part of the role map being associated with a single identifier of a source of images so that only one of the at least two sources of images is to be selected to process a representation of a portion of the overlapping area.

11 . The device of claim 10 , wherein the processor is further configured so that an identifier of a considered source of images is associated with:
a considered point of a sub-area of the overlapping area if no occlusion is detected for the considered point in an image obtained from the considered source of images; or a considered point of a sub-area of the overlapping area as a function of a detection score criterion associated with each source of images of a set of sources of images if no occlusion is detected for the considered point in an image obtained from each source of images of the set of sources of images.

12 . The device of claim 10 , wherein the processor is further configured so that an identifier of a source of images is associated with a considered point of a sub-area of the overlapping area as a function of a detection score criterion if an occlusion is detected for the considered point in an image obtained from each of the sources of images providing images comprising a sub-image representing the overlapping area of the real scene.

13 . The device of claim 10 , wherein the processor is further configured to carry out preliminary steps of determining occlusion criterions, an occlusion criterion being associated with a point of the overlapping area and with a source of images.

14 . The device of claim 13 , wherein the processor is further configured so that a preliminary step of determining an occlusion criterion is carried out for each point of the overlapping area.

15 . The device of claim 10 , wherein the processor is further configured to carry out a step of computing detection scores, a detection score being associated with a point of the overlapping area and with a source of images.

16 . The device of claim 10 , wherein the processor is further configured for carrying out the steps of:
selecting parts of the role map as a function of an identifier of the considered source of images; and processing sub-images of an image obtained by the considered source of images, the processed sub-images corresponding to the selected parts of the role map.

17 . The device of claim 16 , wherein the processor is further configured to carry out a step of projecting the selected parts of the role map into a reference frame associated with the images obtained by the considered source of images.

18 . The device of claim 17 , wherein the processor is further configured to carry out a step of extracting the sub-images to be processed, the extracted sub-images corresponding to the projected parts of the role map.

19 . A method for optimizing processing of at least two images obtained from at least two cameras, the method comprising:
generating an overlapping area of a real scene among at least two cameras; determining an occlusion criterion in the overlapping area; and generating a role map based on the occlusion criterion so that one of the at least two cameras is to be selected to process a representation of a portion of the overlapping area.

20 . A device for optimizing processing of at least two images obtained from at least two cameras, the de

Assignees

• Canon Kk

Inventors

• Sevin Julien
• Kergourlay Gérald

Classifications

• G06T7/0022

  Physics · mapped topic

• G06T11/60 Primary

  Creating or editing images; Combining images with text · CPC title

• G06T7/0097

  Physics · mapped topic

• G06K9/342 Primary

  Physics · mapped topic

• G06V20/52 Primary

  Surveillance or monitoring of activities, e.g. for recognising suspicious objects (recognising microscopic objects G06V20/69) · CPC title