Method for finding paths in video
First Claim
1. A video processing system comprising:
- an up-stream video processing device to accept an input video sequence and to output target information derived from at least one target in said input video sequence, wherein said target information comprises at least one target property or at least one function of said at least one target property;
a target property map builder, coupled to said up-stream video processing device to receive at least a portion of said target information and to build at least one target property map for at least one target type, wherein said target property map comprises a statistical model based on a record of said at least one target property from instances of at least one target type or said at least one function of said at least one target property from instances of said at least one target type in said input video sequence, wherein the statistical model corresponding to an image location, target property or function of target properties and target type is a record of said property of instances of targets of said type that pass through said location, and wherein the statistical model represents at least one of a normal behavior or a normal property of instances of targets of said type that pass through said location; and
the target property map builder being further configured to, for a given target;
consider at least one instance of the target;
determine if said at least one instance of the target is a bad instance, andif said at least one instance of the target is a bad instance, filter said at least one instance of the target by ignoring said at least one instance of the target.
7 Assignments
0 Petitions
Accused Products
Abstract
A system for detecting behavior of a target may include: a target detection engine, adapted to detect at least one target from one or more objects from a video surveillance system recording a scene; a path builder, adapted to create at least one mature path model from analysis of the behavior of a plurality of targets in the scene, wherein the at least one mature path model includes a model of expected target behavior with respect to the at least one path model; and a target behavior analyzer, adapted to analyze and identify target behavior with respect to the at least one mature path model. The system may further include an alert generator, adapted to generate an alert based on the identified behavior.
-
Citations
23 Claims
-
1. A video processing system comprising:
-
an up-stream video processing device to accept an input video sequence and to output target information derived from at least one target in said input video sequence, wherein said target information comprises at least one target property or at least one function of said at least one target property; a target property map builder, coupled to said up-stream video processing device to receive at least a portion of said target information and to build at least one target property map for at least one target type, wherein said target property map comprises a statistical model based on a record of said at least one target property from instances of at least one target type or said at least one function of said at least one target property from instances of said at least one target type in said input video sequence, wherein the statistical model corresponding to an image location, target property or function of target properties and target type is a record of said property of instances of targets of said type that pass through said location, and wherein the statistical model represents at least one of a normal behavior or a normal property of instances of targets of said type that pass through said location; and the target property map builder being further configured to, for a given target; consider at least one instance of the target; determine if said at least one instance of the target is a bad instance, and if said at least one instance of the target is a bad instance, filter said at least one instance of the target by ignoring said at least one instance of the target. - View Dependent Claims (2, 3, 4, 5, 6, 7, 8, 9, 10)
-
-
11. A method of video processing, comprising:
-
processing an input video sequence to obtain target information derived from at least one target in said input video sequence, wherein said target information comprises at least one target property or at least one function of said at least one target property; building at least one target property map that comprises a statistical model based on a record of said at least one target property or said at least one function of said at least one target property from said at least one target in said input video sequence, wherein the statistical model corresponding to an image location, target property or function of target properties and target type is a record of said property of instances of targets of said type that pass through said location, and wherein the statistical model represents at least one of a normal behavior or a normal property of instances of targets of said type that pass through said location; and wherein said building at least one target property map further includes, considering at least one instance of the target; determining if said at least one instance of the target is a bad instance, and if said at least one instance of the target is a bad instance, filtering said at least one instance of the target by ignoring said at least one instance of the target. - View Dependent Claims (12, 13, 14, 15, 16, 17, 18, 22, 23)
-
-
19. A non-transitory computer-readable medium containing instructions that, when executed by a processor, cause the processor to:
-
process an input video sequence to obtain target information derived from at least one target in said input video sequence, wherein said target information comprises at least one target property or at least one function of said at least one target property; and build at least one target property map that comprises a statistical model based on a record of said at least one target property or said at least one function of said at least one target property from said at least one target in said input video sequence, wherein the statistical model corresponding to an image location, target property or function of target properties and target type is a record of said property of instances of targets of said type that pass through said location, and wherein the statistical model represents at least one of a normal behavior or a normal property of instances of targets of said type that pass through said location; and wherein said building at least one target property map further includes, considering at least one instance of the target; determining if said at least one instance of the target is a bad instance, and if said at least one instance of the target is a bad instance, filtering said at least one instance of the target by ignoring said at least one instance of the target. - View Dependent Claims (20, 21)
-
Specification