Font Size: a A A

The Research Of H.264 Decoder Error Concealment Technology

Posted on:2009-04-10Degree:MasterType:Thesis
Country:ChinaCandidate:B Q RongFull Text:PDF
GTID:2178360242976856Subject:Communication and Information System
Abstract/Summary:PDF Full Text Request
H.264/AVC is the most recent international standard of video compression, designed by ITU-T VCEG and ISO/IEC MPEG. In the latest years, thanks to the great efforts paid by professors and engineers, H.264 has been introduced into practical applications. This paper is focus on the development of video compressed standard and the adaptation of lost package network. According to the requirement of next generate network, we provide a new algorithm on H.264 error concealment.First we introduce the lost packet network, including the Internet and wireless network system architecture, the problem which we need to solve urgently. We talk about the error condition in transmission, enumerate some error concealment algorithm in decoder. Then we make extensive analyze on the spatial concealment in Intra frame and temporal concealment in Inter frame, discuss about the error detective way for the lost pixels.Secondly, we make the research of Xin.Li's LMS filter error concealment. It is based on the local stationary Gaussian process. It can recover the sequence error pixels by the MAP estimation. According to the mutil-scanning approach and local stationary in image, we refer to effective edge detection LS error concealment. Focusing on the Intra frame in video streams, we use the Sobel algorithm and mean square algorithm to detective the edge of error pixels. It is helpful to judge whether there are edges pass though the error pixel. Sobel algorithm is used in low bit rate. Mean square algorithm is used in high bit rate.After we get the edge information, the error pixels may be divided from two parts, stable pixels and unstable pixels. The errors which are stable pixels can be concealed by linear interpolation. For the missing pixels, which are unstable, the initial value of the missing pixel can be obtained by the lease square method; the optimization value can be built up by the mutil-direct scanning. Our algorithm has lower time complexity than Xin.Li's, but has higher quality than before. It is a good way for the Intra frame concealment.Finally, introduce the architecture of H264 decoder. Aiming at Intra Frame and Inter Frame, we use SI and BMA algorithm. Adding the SI, BMA and edge detection LS algorithm into the x264 decoder, we suppose an improvement error concealment solution. Though the complementation of these three algorithms, we get the new way for error concealment. This solution is not depending on the encoder information, no time complexity or space complexity will be added. Especially in fast movement and sense transsion, the simulator result is much better than before.
Keywords/Search Tags:H.264, Spatial domain, Error concealment, Sobel, Least square
PDF Full Text Request
Related items