Research On H.264 Motion Estimation Algorithm In Video Processing Of Angiography

With the development of medical imaging technology and computer technology, digital imaging has been widely used in clinical diagnosis. As large amount of data, it needs to be compressed frequently when storage and network transmission. Medical imaging requires high compression rate while the reconstructed image has better rate-distortion. Therefore, research about image compression techniques for medical research has an important practical significance.This paper introduces the compression principle, methods of medical imaging and H.264; then analyzes the motion estimation algorithm which is the impact on compression efficiency and the rate-distortion of reconstructed image in H.264; and then the improved motion estimation algorithm is proposed for the characteristics of cardiac imaging. Experiment with the Standard test sequence and the clinical trial data, the results show that:The improved algorithm improves the compression efficiency significantly and ensures the quality of cardiac imaging at the same time. The main tasks of this paper include as follows:(1)This paper juxtaposes and analyzes the performance and features about the four motion estimation algorithms (FS, FFS, UMHexagonS, EPZS) applying to the video compressing and transmitting of angiography, bring to light:UMHexagonS algorithm reaching equilibrium at high compression ratio and better rate distortion and is more suitable for the video compressing and transmitting of angiography.(2)The paper proposes an optimization on UMHexagonS algorithm and its improved algorithm FAMRS algorithm:Add a control condition in FAMRS algorithm to determine the search direction and reduce search points, and combine with cross search algorithm.(3)This paper does the experiment using the standard test sequences and the actual cardiac imaging image, the results show that:the improved algorithm reduces the time of the motion estimation significantly and improves the compression efficiency compared with the UMHexagonS algorithm and FAMRS algorithm in the similar reconstruction video quality and bit rate.