Video target dividing method based on motion detection

Abstract

The invention belongs to technique area of video monitoring and video processing in bottom layer. The method includes following steps: (1) reading out original video frame, initializing background model; (2) reading out current video frame, extracting characteristic value of video, matching old background model by using current characteristic value, analyzing statistical characterization; based on model, setting up parameters, and updating background model; (3) in current background model, accounting background distribution with largest occurrence probability; after operation of erosion and clearing up shadow, the current background is picked up. The invention segments reliable background and foreground in real time. The invention also improves segmenting speed in a certain extent and avoids error expand in order to meet requirement in real time and stability of video monitoring.

Description

technical field [0001] The invention relates to a video object segmentation method, in particular to a video object segmentation method based on motion detection. The invention belongs to the technical field of video surveillance and underlying video processing. Background technique [0002] Video signal is the most important one in multimedia information, and the traditional monitoring system can no longer meet the monitoring of complex and changing environments. The monitoring system must be able to analyze the video signal and understand its content, so as to further understand the suspicious behavior in a specific place, issue real-time alarms, and mark and transmit relevant video content. This kind of high-level video content understanding, the first thing to realize is the segmentation of video objects based on motion detection, that is, to separate the moving foreground objects and background in the video. [0003] So far, video segmentation technology based on moti...

AI Technical Summary

Problems solved by technology

Its iterative algorithm can get better segmentation results, but there are two problems: one is the large amount of calculation, and the other is that the convergence speed depends on the scene, noise and motion

The basis of the motion tracking method is the feature matching or optical flow estimation of the image frames of the video sequence and the dynamic model describing its real-time motion process. However, due to the selection of features, the amount of data to be processed is greatly reduced, which affects the accuracy of video segmentation.

[0004] After literature search, Bouthemy P and Francois E found in the article "Motion segmentation and qualitative dynamic scene analysis from an image sequence" Int′l Journal of Computer Vision ("Computer Vision") , 1993, 10(2): 157~182) proposed a space-time method based on change region detection, which does not require the estimation of the optical flow field and the correspondence of any feature points, but relies on the brightness gradient of the space-time image, and the segmentation accuracy is easily affected by the observation noise effect

Images