Crowd separation
A method for crowd separation, the method may include receiving or generating a signature of an image of a first plurality of pedestrians, wherein the image comprises regions, wherein the signature of the image comprises descriptors of the regions, wherein each descriptor of a region is associated with a region and comprises a set of identifiers that identify content included in the region and in a vicinity of the region; detecting, within the descriptors of the regions, unique combinations of identifiers that are indicative of spatial relationships between the regions and bounding boxes that surround pedestrians of the first plurality of pedestrians; wherein the unique combinations are learnt during a supervised machine learning process that is fed with test images of densely positioned pedestrians; and locating the pedestrians of the first plurality of pedestrians based, at least in part, on the spatial relationships related to detected unique combinations and to locations of the regions.
1. A method for crowd separation, the method comprises:
receiving or generating a signature of an image of a first plurality of pedestrians, wherein the image comprises regions, wherein the signature of the image comprises descriptors of the regions, wherein each descriptor of a region is associated with a region and comprises a set of identifiers that identify content included in the region and in a vicinity of the region;
detecting, within the descriptors of the regions, unique combinations of identifiers that are indicative of spatial relationships between the regions and bounding boxes that surround pedestrians of the first plurality of pedestrians; wherein the unique combinations are learnt during a supervised machine learning process that is fed with test images of densely positioned pedestrians; and
locating the pedestrians of the first plurality of pedestrians based, at least in part, on the spatial relationships related to detected unique combinations and to locations of the regions.
2. The method according to claim 1 comprising ignoring unique combinations of identifiers that have less than a predefined number of identifiers.
3. The method according to claim 1 wherein the locating of the pedestrians of the first plurality of pedestrians comprises locating the bounding boxes and locating the pedestrians within the bounding boxes.
4. The method according to claim 1 comprising distinguishing between the pedestrians of the first plurality of pedestrians.
5. The method according to claim 1 comprising receiving or generating signatures of a group of images of the first plurality of pedestrians that were acquired at different points of time and estimating movement patterns for at least some of the pedestrians of the first plurality of pedestrians.
6. The method according to claim 1 wherein different test images represents pedestrians at different spatial relationships from one or more regions of the test images.
7. The method according to claim 1 comprising performing the supervised machine learning process.
8. The method according to claim 7 wherein the performing of the supervised machine learning process comprises learning unique combination that appear in regions that belong to a bounding box and learning unique combinations that appear in regions that are located outside the bounding boxes.
9. A non-transitory computer readable medium that stores instructions for:
receiving or generating a signature of an image of a first plurality of pedestrians, wherein the image comprises regions, wherein the signature of the image comprises descriptors of the regions, wherein each descriptor of a region is associated with a region and comprises a set of identifiers that identify content included in the region and in a vicinity of the region;
detecting, within the descriptors of the regions, unique combinations of identifiers that are indicative of spatial relationships between the regions and bounding boxes that surround pedestrians of the first plurality of pedestrians; wherein the unique combinations are learnt during a supervised machine learning process that is fed with test images of densely positioned pedestrians; and
locating the pedestrians of the first plurality of pedestrians based, at least in part, on the spatial relationships related to detected unique combinations and to locations of the regions.
10. The non-transitory computer readable medium according to claim 9 that stores instructions for ignoring unique combinations of identifiers that have less than a predefined number of identifiers.
11. The non-transitory computer readable medium according to claim 9 wherein the locating of the pedestrians of the first plurality of pedestrians comprises locating the bounding boxes and locating the pedestrians within the bounding boxes.
12. The non-transitory computer readable medium according to claim 9 that stores instructions for distinguishing between the pedestrians of the first plurality of pedestrians.
13. The non-transitory computer readable medium according to claim 9 that stores instructions for receiving or generating signatures of a group of images of the first plurality of pedestrians that were acquired at different points of time and estimating movement patterns for at least some of the pedestrians of the first plurality of pedestrians.
14. The non-transitory computer readable medium according to claim 9 wherein different test images represents pedestrians at different spatial relationships from one or more regions of the test images.
15. The non-transitory computer readable medium according to claim 9 that stores instructions for performing the supervised machine learning process.
16. The non-transitory computer readable medium according to claim 15 wherein the performing of the supervised machine learning process comprises learning unique combination that appear in regions that belong to a bounding box and learning unique combinations that appear in regions that are located outside the bounding boxes.
17. A computerized system that comprises:
an input, a memory and a processor that comprises at least one processing circuits;
wherein the processor is configured to receive or generate a signature of an image of a first plurality of pedestrians, wherein the image may include regions, wherein the signature of the image may include descriptors of the regions, wherein each descriptor of a region may be associated with a region and may include a set of identifiers that identify content included in the region and in a vicinity of the region;
wherein the processor is configured to:
detect, within the descriptors of the regions, unique combinations of identifiers that may be indicative of spatial relationships between the regions and bounding boxes that surround pedestrians of the first plurality of pedestrians; wherein the unique combinations may be learnt during a supervised machine learning process that may be fed with test images of densely positioned pedestrians;
and locate the pedestrians of the first plurality of pedestrians based, at least in part, on the spatial relationships related to detected unique combinations and to locations of the regions.
18. The computerized system according to claim 17 wherein the input is configured to receive the signature of the image of the first plurality of pedestrians.
19. The computerized system according to claim 17 wherein the input is a sensor.
20. The computerized system according to claim 17 wherein the processor is configured to generate the signature of the image of the first plurality of pedestrians.