Video object tracking cutting method using Snake profile model

Abstract

The invention relates to a video object tracking cutting method using a Snake profile model, comprising the following steps: based on a space-time fusion method, roughly locating a Snake profile at a time field passes through a sectional frame core vector prediction manner, and then evolving from the initial profile by using a modified Snake greedy method in an air field to obtain a precise profile of the video object. The method comprises the following specific steps: dividing a video sequence into cut units with every four frames as a unit in the time field; and selecting the two front frames in one unit as key frames, wherein the initial profile is an external rectangle of the movement area obtained through the detection of movement change, and the initial profiles of the third and fourth frames are obtained by mapping the previous frame of the premise profile and the two previous frames of movement vector reflections. In the air field, during profile point iteration updating, large errors are considered, the impossible profile point is eliminated in real time. Compared with the prior art, the method has the advantages that the disadvantages of manually drawing the initial profile are overcome and high precision and rapid speed are achieved.

Description

technical field [0001] The invention relates to a processing method in video object extraction, in particular to a video tracking and segmentation method using a Snake contour model. Background technique [0002] The method of generating video object plane (VOP) by using video object tracking technology can not only improve the accuracy of segmentation results, but also conform to the content-based image representation of MPEG-4. Its essence is: use the segmentation results of the previous frame to find the best matching position of the object in the current frame. [0003] At present, some video segmentation methods based on object tracking are also emerging at home and abroad. Video object tracking usually uses model matching, and the main methods are as follows: project the object contour segmented in the current frame to the next frame according to the motion information, as the initial contour of the moving object segmentation in the next frame, and then combine other ...

AI Technical Summary

Problems solved by technology

[0004] Zhang Xiaobo adopts the video segmentation method based on block affine classification and HD tracking, automatically obtains the binary model of the moving object and uses the Hausdorff distance for tracking in subsequent frames, and divides the motion of the video object into two parts: slow-changing and fast-changing. The background edge model is matched and updated, and the segmentation effect is better, but the calculation is more complicated

Li Bing (see Li Bing, Xu De, Wang Fangshi. A video segmentation algorithm based on object tracking [J]. Journal of Beijing Jiaotong University. 2005, 29(5): 89-91) uses motion information to track the target, through the initial The accurate template of the frame can automatically segment the subsequent image, but this method needs to manually draw the initial rough outline of the key frame, and it will fail when the object deformation is large

Song Lifeng (see Song Lifeng, Wei Gang, Wang Qunsheng. A semi-automatic method for segmenting video objects [J]. Journal of South China University of Technology, 2002, 30(8): 49-54) Template matching forms a closed loop to limit the segmentation of subsequent frames As a result, it avoids the continuous transmission and expansion of the segmentation error in the process of object tracking, and the accuracy is high. However, it is still necessary to manually outline the first frame of the initial video sequence, and automatic and fast segmentation cannot be achieved.

Images