Evaluation of models generated from objects in video
Models are generated from objects identified in video. Each model is evaluated based on knowledge of the objects determined from video analysis, and preferred models are identified based on the evaluations. In some examples, each model could be evaluated by tracking a movement of each object in the video by using each model to track the object from which it was generated, evaluating an ability of each model to identify the objects in the video that are similar to the object from which it was generated, and determining an amount of false identifications made by each model of different objects in different video that does not include the object from which it was generated.
1. An image processing system comprising:
a processing system configured to generate a plurality of respective models from identified objects in a first video, perform evaluations on each of the plurality of respective models based on knowledge of the identified objects determined from video analysis, and identify at least one preferred model from the plurality of respective models based on the evaluations, wherein evaluating each model includes applying a set of video analytics in order to effectuate the identification of the at least one preferred model, and
tracking a movement of each identified object in the first video by using each of the plurality of respective models to track the respective identified object from which the plurality of respective models were generated, evaluating statistics from the tracking based on each of the plurality of respective models;
a storage system storing the statistics as a set of motion dynamics for the plurality of respective models in a nonvolatile computerized memory;
a processing system configured to identifying at least one scene in the first video by tracking the movement of a respective identified object in the video;
a storage system storing scene statistics in said nonvolatile computerized memory.
2. An image processing system according to claim 1 , wherein the scene statistics comprise at least one of average movement step size of the first identified object, maximum movement step size of the first identified object, a rate of speed of the first identified object, physical location of the first identified object as depicted in the identified scene, and most used traffic pattern in an identified scene of said first identified object.
3. An image processing system according to claim 1 , further comprising a video source transmitting a video stream of said first video, wherein the processing system identifies at least one scene in the first video by tracking the movement of an identified object in the video.
4. An image processing system according to claim 1 , further comprising a video source transmitting a video stream of said first video and a different video, wherein evaluating each model comprises evaluating a respective model for an ability to track a different object in the different video when the different object is similar to the identified object from the first video.
5. An image processing system according to claim 1 , wherein an absence of at least one object in additional portions of the first video is determined from the scene statistics based on the at least one preferred model.