| 主权项 |
1. A method for real time processing of a sequence of video frames, said method comprising:
analyzing, by a processor of a computer system, a current frame in the sequence and at least one frame in the sequence occurring prior to the current frame, each frame comprising a two-dimensional array of pixels and a frame-dependent color intensity at each pixel, said array of pixels in each frame being a totality of pixels in each frame in the sequence of video frames received in synchronization with a recording of the video frames in real time, said analyzing comprising performing a background subtraction on the at least one frame, said performing the background subtraction determining a background image and also determining a static region mask associated with a static region consisting of a contiguous distribution of pixels in the current frame, said static region mask identifying each pixel in the static region upon the static region mask being superimposed on the current frame, said background image comprising the array of pixels and a background model of the at least one frame and not comprising any moving object, wherein said performing the background subtraction comprises executing a mixture of Gaussians algorithm to generate the background model, a foreground image, and the static region, wherein the mixture of Gaussian algorithm utilizes 3 to 5 Gaussian functions coupled together in a linear combination by Gaussian weight coefficients to define a pixel color intensity probability function, and wherein the background subtraction constrains a sum of the Gaussian weight coefficients (a) for identifying the foreground image to exceed a specified high weight threshold that is sufficiently high to ensure that the foreground image is adequately identified and (b) for identifying the static region to exceed a specified low weight threshold that is sufficiently low to ensure that the static region is not excessively fragmented. |
