CAJ | 학술논문

在分布式视频编解码系统中，由于细节变化丰富而生成的边信息与原始视频帧之间存在较大的差距，降低了解码帧质量。针对如何实现解码帧质量检测，并对低质量的解码帧进行优化以提高系统性能的问题，提出一种基于相关性的分布式解码检测优化算法。在编解码端分别计算相关性，得到置信度值并进行解码检测。对于低质量的像素块根据相关性及运动估计进行重构，最后通过滤波和双向运动补偿产生优化像素块。实验结果表明，所提算法对低质量边信息具有优化作用。
재분포식시빈편해마계통중，유우세절변화봉부이생성적변신식여원시시빈정지간존재교대적차거，강저료해마정질량。침대여하실현해마정질량검측，병대저질량적해마정진행우화이제고계통성능적문제，제출일충기우상관성적분포식해마검측우화산법。재편해마단분별계산상관성，득도치신도치병진행해마검측。대우저질량적상소괴근거상관성급운동고계진행중구，최후통과려파화쌍향운동보상산생우화상소괴。실험결과표명，소제산법대저질량변신식구유우화작용。
In distributed video coding/decoding system , there will be a big gap between the original video frame and the side information under the circumstance that the details of the changes are rich .It makes the quality of decoding frame reduced .In order to detect the quality of decoding frames and improve system performance by optimising the low-quality decoding frames , in this article we present a distributed decoding detection and optimisation algorithm which is based on the correlation .The correlations are calculated at the coding side and the decoding side respectively , the confidence values are then obtained to carry out the decoding detection .For low quality pixel blocks , they are reconstructed by correlations as well as motion estimation .Finally the optimised pixel blocks are generated by filtering and bi-directional motion compensation.Experimental results show that the proposed algorithm can optimise the low quality side information .