Modeling of temporarily static objects in surveillance video data

摘要

A foreground object blob having a bounding box detected in frame image data is classified by a finite state machine as a background, moving foreground, or temporally static object, namely as the temporally static object when the detected bounding box is distinguished from a background model of a scene image of the video data input and remains static in the scene image for a threshold period. The bounding box is tracked through matching masks in subsequent frame data of the video data input, and the object sub-classified within a visible sub-state, an occluded sub-state, or another sub-state that is not visible and not occluded as a function of a static value ratio. The ratio is a number of pixels determined to be static by tracking in a foreground region of the background model corresponding to the tracked object bounding box over a total number of pixels of the foreground region.

主权项

1. A method for using region-level adaptive background modeling, the method comprising:
classifying via a finite state machine device an object blob having a bounding box detected in frame image data of a video data input as a background object, as a moving foreground object, or as a temporally static object, by classifying the object blob as the temporally static object when the detected bounding box is distinguished from a background model of a scene image of the video data input and remains static in the scene image for at least a first threshold period; tracking the bounding box of the object blob classified as the temporally static object via the finite state machine device by matching masks of the bounding box in subsequent frame data of the video data input; and sub-classifying the tracked temporally static object as within a visible sub-state, as within an occluded sub-state, or as within another sub-state that is not visible and not occluded, as a function of a static value ratio of a total number of pixels overlapped by both the tracked bounding box and a foreground region of the background model over a total number of pixels contained within the tracked bounding box, wherein the foreground region corresponds to the tracked bounding box.