Motion Analysis Technology And Its Applications In Video Sequences

As a key technique of the digital image processing and computer vision, motion estimation of sequence images plays an important role in the compression of image data, the detection of the moving targets and the layering of video. It has wide application value in industry, national defense, medicine, internet technology, and etc. In this thesis, global motion estimation and its applications, which include the detection of moving dim point targets and layering of video sequences, are investigated. Its specific work includes:Making an analysis of the video motion models and deducing the parametric models which describe the projection of three-dimension rigid motion of a surface to the two-dimension image plane.Introducing the system structure and key technology modules of global motion estimation. Describing several possible details of each key technology module.Combining with the parametric model, proposing an improved global motion estimation algorithm based on the analysis of traditional algorithm. Comparing with the traditional algorithm, the experimental results show that the proposed algorithm is superior in computation and accuracy.Considering the concrete application in the detection of moving dim point targets of sequence images, a rapid algorithm of global motion estimation and compensation for engineering implementation is given. And then, making an analysis of several searching algorithms by straight line segment in three-dimension and giving their performances compared with the traditional searching techniques.Based on global motion estimation and compensation, making an analysis of the technique of the layering of video sequences. The analysis shows that the technique, dividing image space into the four independent parts, is very useful for drawing regions of interest (ROI).