Dissertation > Excellent graduate degree dissertation topics show

AVS video encoder module optimization algorithm and its implementation on DSP

Author: WeiJianYun
Tutor: PengYuHuaï¼›YangMingQiang
School: Shandong University
Course: Signal and Information Processing
Keywords: AVS Sub-pixel motion estimation Inter-mode DM642
CLC: TN919.81
Type: Master's thesis
Year: 2010
Downloads: 74
Quote: 1
Read: Download Dissertation

Abstract


AVS is a country with independent intellectual property rights of the second generation source coding standard. The standard proposed only changed the first generation source coding technology behind the situation, but also to solve the deadlock AVC patent licensing issues. With the current popular international standard MPEG-4 and H.264/AVC, compared with high-performance, low computational complexity, patents and low cost. But AVS codec chip and software is not mature, its industrial implementation is still a long way to go. Therefore, the research of AVS encoder is of paramount importance. In this paper, the development of multimedia technology in the study and the current status of a variety of video compression technology, based on the AVS encoder sub-pixel motion estimation and inter mode selection module has been studied and proposed a fast algorithm to improve the encoding speed. Then the encoder and fast algorithms ported to the DM642 platform for real-time encoding and application of AVS foundation. The main thesis work includes: 1. Right to study sub-pixel motion estimation, and proposed two effective fast search algorithm. In order to obtain a better image quality and high coding efficiency, AVS using the 1/4 pixel accuracy motion estimation technique. But the standard used in the sub-pixel full search algorithm in motion estimation computation accounts for a very large proportion, thus reducing the search time subpixel is particularly important. As the video sequence has center-biased characteristics and strong temporal correlation, so these two aspects of the topic to study sub-pixel search to reduce time-consuming, we propose two fast algorithms. Consider an algorithm center-biased characteristic, according to sub-integer pixel advantages and benefits of the relationship between the orientation of sub-pixel motion vectors to predict the direction of the sub-pixel search points to be reduced from 16 to 4 to 8. The algorithm is simple, coding is better for the hardware and the application of real-time systems. Another algorithm uses the temporal correlation with the macro block of space and time of the neighboring block motion vectors to predict the motion vector of the current macroblock to reduce search time. The algorithm can be coded encoder save 21% to 36%, slightly better results on the coding method, but more complex implementation. In specific video applications, according to different needs, choose a suitable fast algorithm used. (2) Based on the statistics, we propose a threshold based on forecasts and judgments Fast inter mode selection algorithm. Standards for four kinds of inter mode and Skip. Intra other six modes have carried out the search. Are each inter mode for motion estimation, and for all of the search pattern rate distortion calculation should be carried out, such a search strategy is very time consuming. The statistics found that the best coding mode of the current block in time and space with the best coding mode of the adjacent block the same probability is very large, it can be the best coding mode of the adjacent block to predict the most likely current block encoding mode , thereby reducing the search mode, save the encoding time. Experiments show that, for different video sequences, the algorithm to ensure image quality and bit rate is almost constant, the encoding total time savings of 23% ~ 59%. 3. The AVS encoder and the topics proposed fast algorithms to embedded DM642 platform, and to achieve two video acquisition methods and data backhaul. One is based on UDP / IP protocol network transmission, DM642 received video sequence to be encoded. Second, the camera captures DM642 video information is encoded, ie monitoring system for coding. Encoded yards through the network transmission from the host computer program for storage. This is half AVS in embedded real-time encoding stage laid a good foundation.

Related Dissertations

  1. The Optimization of AVS Video Decoder on the PC Platform and Improvement of Field Decoding,TN919.81
  2. FPGA/DSP Image Co-processor Technology and Ethernet Data Transmission,TP391.41
  3. The Research and Application of H.264 Coding Technology in the Video Monitoring System,TP277
  4. High Efficiency 3D Video Representation and Coding,TN919.81
  5. The Study of the Background Modeling Algorithm for Object Detection and Implementation Based on DSP Technique,TP391.41
  6. Implementation and Optimization of H.264 Video Encoder Based on DSP,TN919.81
  7. Design and Implementation of H.264 Video Encolder and Storage System Based on DM642,TN919.81
  8. Automatic Exposure Technology Research for Airborne Visible Light Camera,TP391.41
  9. The Detection of Facial Fatigue by Infrared Imaging System Based on DM642,TP391.41
  10. Facial feature extraction system is a key technology research and its DSP,TP391.41
  11. Border infrared monitoring system design moving target detection warning,TP391.41
  12. DM642-based infrared moving target detection and tracking technology,TP391.41
  13. Low Complexity H.264/AVC Encoding and Transcoding,TN919.81
  14. Intelligent Vehicle Tracking System Based on DSP Platform,TP391.41
  15. Research on the Detection System of Eye Information Based on Embedded,TP391.41
  16. The H.264 Inter block matching algorithm with encoder DSP implementation,TN919.81
  17. The Research of H.264 Encoder and the Realization on DSP,TN919.81
  18. Research of Video Image Fusion Technology on Moving Vehicle,TP391.41
  19. Mean shift - based image tracking technology,TP391.41
  20. Hardware Design of Deblocking Loop Filter in AVS Video Decoder,TN919.81
  21. Fast Algorithm of the Inter Mode Decision for H. 264/AVC,TN919.81

CLC: > Industrial Technology > Radio electronics, telecommunications technology > Communicate > Image communication, multimedia communication > Image coding
© 2012 www.DissertationTopic.Net  Mobile