Real-time video analysis for security surveillance
First Claim
1. In a camera, a method for identifying moving objects of interest in a field of view of the camera, comprising:
- capturing two or more images successively over a time period, each image being associated with a different time during the time period;
obtaining a binary image from each successive pair of images, the binary image comprising a binary value at each pixel indicating whether or not a change in pixel values of at least a predetermined magnitude has occurred at that pixel between the time associated with the first image of the successive pair of images and time associated with the second image of the successive pair of images;
deriving one or more motion boxes each encapsulating one or more nearby pixels in the binary image;
processing the motion boxes of each binary image to obtain refined motion boxes, wherein the motion boxes are processed using a plurality of modular filters each receiving one or more motion boxes as input and providing one or more processed motion boxes as output, the processed motion boxes of one of the modular filters being suitable to be provided as input to another one of the modular filters, the modular filters comprising one or more tracking filters, each tracking filter tracking motion boxes of the binary images through the time period, and the tracking filter rejects motion boxes that show a displacement over the time period that is less than a predetermined value; and
classifying portions of the successive pair of images corresponding to the refined motion boxes each into a category representative of one of the moving objects of interest.
6 Assignments
0 Petitions
Accused Products
Abstract
A camera and method identify moving objects of interest in a field of view of the camera. The method includes: capturing two or more images successively over a time period, each image being associated with different times during the time period; obtaining binary image from each successive pair of images, the binary image comprising a binary value at each pixel indicating whether a change in pixel values of at least a predetermined magnitude has occurred at that pixel between the time associated with the first image of the success pair of images and time associated with the second image of the successive pair of images; deriving one or more motion boxes each encapsulating one or more nearby pixels in binary image; processing motion boxes of each binary image to obtain refined motion boxes; and classifying refined motion boxes each into categories representative of one moving object of interest.
56 Citations
36 Claims
-
1. In a camera, a method for identifying moving objects of interest in a field of view of the camera, comprising:
-
capturing two or more images successively over a time period, each image being associated with a different time during the time period; obtaining a binary image from each successive pair of images, the binary image comprising a binary value at each pixel indicating whether or not a change in pixel values of at least a predetermined magnitude has occurred at that pixel between the time associated with the first image of the successive pair of images and time associated with the second image of the successive pair of images; deriving one or more motion boxes each encapsulating one or more nearby pixels in the binary image; processing the motion boxes of each binary image to obtain refined motion boxes, wherein the motion boxes are processed using a plurality of modular filters each receiving one or more motion boxes as input and providing one or more processed motion boxes as output, the processed motion boxes of one of the modular filters being suitable to be provided as input to another one of the modular filters, the modular filters comprising one or more tracking filters, each tracking filter tracking motion boxes of the binary images through the time period, and the tracking filter rejects motion boxes that show a displacement over the time period that is less than a predetermined value; and classifying portions of the successive pair of images corresponding to the refined motion boxes each into a category representative of one of the moving objects of interest. - View Dependent Claims (2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17)
-
-
18. A camera comprising:
-
an optical and actuation system; a communication interface; non-volatile storage; a processor; and a memory module, wherein the memory module, upon initialization, loads from the non-volatile storage instructions to be executed by the processor, the instructions comprising instructions for carrying out a method for identifying moving objects of interest in a field of view of the camera, the method comprising; capturing two or more images successively over a time period, each image being associated with a different time during the time period; obtaining a binary image from each successive pair of images, the binary image comprising a binary value at each pixel indicating whether or not a change in pixel values of at least a predetermined magnitude has occurred at that pixel between the time associated with the first image of the successive pair of images and time associated with the second image of the successive pair of images; deriving one or more motion boxes each encapsulating one or more nearby pixels in the binary image; processing the motion boxes of each binary image to obtain refined motion boxes, wherein the motion boxes are processed using a plurality of modular filters implemented as instructions executable by the processor, each modular filter receiving one or more motion boxes as input and providing one or more processed motion boxes as output, the processed motion boxes of one of the modular filters being suitable to be provided as input to another one of the modular filters, the modular filters comprising one or more tracking filters, each tracking filter tracking motion boxes of the binary images through the time period, and the tracking filter rejects motion boxes that shows a total distance traversed over the time period that is less than a predetermined value; and classifying portions of the successive pair of images corresponding to the refined motion boxes each into a category representative of one of the moving objects of interest. - View Dependent Claims (19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36)
-
Specification