Dissertation > Excellent graduate degree dissertation topics show

Compressed video image interpolation based technology research

Author: GaoXinZuo
Tutor: ZhaoDeBin
School: Harbin Institute of Technology
Course: Computer Science and Technology
Keywords: Image Interpolation Compressed video Model guidance Wiener filter Low Complexity
CLC: TP391.41
Type: Master's thesis
Year: 2011
Downloads: 36
Quote: 0
Read: Download Dissertation

Abstract


Image interpolation is the most basic image research topics, many of image interpolation methods have been proposed in the literature to solve the problem of non-compressed image interpolation. However, the large number of video sequences are stored in compressed format, or by the requirements of video bandwidth limitation in compressed form for transmission. Some non-compressed video-based image interpolation method, when applied directly to the compressed video image interpolation, often not better results. This is because, on the one hand, these methods are not already in the use of video information stream; other hand, they do not consider the quantization error of a compressed video, the quantization error in some cases is obvious. On these issues, we study H.264/AVC and MPEG-2 compressed video image interpolation has made some effort. We propose a pair of H.264/AVC video compression, pattern-based instruction video image frame interpolation method. We designed interpolation filter direction when considering the intra prediction mode information. Direction for each intra-prediction mode, the video sequence in classical trained on training sets corresponding to a set of optimum interpolation filters, so that each interpolation filter can be automatically adapted to the corresponding one of the frame with his direction prediction model, further, quantitative factors are also involved as context reference to the interpolation filter design and selection. Experimental results show that the model is compared to the other conventional models Bicubic, Bilinear, LAZA and NEDI, interpolation can improve performance while maintaining low computational complexity. Further on we present a model based on the H.264/AVC video compression interframe video guide interpolation method, for each inter-frame (P frames and B-frames), the inter prediction mode is considered to obtain the motion information, such as the motion vector (motion vector). Each of the inter-frame pixel to be interpolated, the interpolation filter which is based on the motion information by its corresponding reference point interpolation filter copied, so the way the reference pixel to be interpolated, the interpolation filtering whichever way the mean pixel values ??obtained. This design does not destroy the structure of the compressed video. Experimental results show that the model is compared to the other conventional models Bicubic, Bilinear, LAZA and NEDI, interpolation can improve performance while maintaining low computational complexity. Reference H.264/AVC intra prediction mode and the edge of the image interpolation technique guidance successful experience, we propose a MPEG-2 compressed video based on the direction of interpolation method in the model, 8x8 intra prediction mode frame rules block in the transform domain is divided into nine kinds of directions, and the interpolation in this piece is considered to be along the direction of the pixel block. Each predetermined direction, the video sequence in the classical trained on training sets a set of optimal Wiener interpolation filter and the interpolation filter with these groups. Using a similar method, we have for each rule blocks, the inter-frame, along the direction of the corresponding block, as the interpolation reference block, the experimental results show that the model is compared to other conventional linear Bicubic interpolation model and Bilinear model LAZA guidance and direction and NEDI, can improve the interpolation performance while maintaining low computational complexity to meet the practical application.

Related Dissertations

  1. Of speckle noise removal in medical ultrasound image,TP391.41
  2. Research and Implementation of Algorithm for Restoration of Motion Blurred Images,TP391.41
  3. Research of Underwater Image Restoration Algorithm Based on Backward Scattering Noise Model,TP391.41
  4. Common types blurred image restoration Research and Implementation,TP391.41
  5. Study on the Algorithms of Adaptive Optical Images Restoration,TP391.41
  6. The Fast Method for DOA Estimation of Wideband Signals and Its Implementation on DSP,TN911.7
  7. The Frontend Noise Reduce of Isolated Word Speech Recognition,TN912.34
  8. The Study on Image Processing and Faults Location of Transmission Line Based on Aerial Image,TM76;TM755
  9. Research on Channel Estimation Methods in MIMO-OFDM Systems,TN919.3
  10. Research on the Algorithm of Super-resolution Reconstruction,TP391.41
  11. MAP-based super-resolution reconstruction algorithm for compressed video research,TP391.41
  12. Algorithm Research on CBCT Images Denosing and Registration,TP391.41
  13. A Study on Image Restoration Method for Magnetic Particle Inspection System,TP391.41
  14. Based on the time domain filter for speech enhancement algorithm,TN912.35
  15. Sparse aperture optical imaging system algorithm for image restoration,TP391.41
  16. Optical synthetic aperture image reconstruction algorithm for high-definition,TP391.41
  17. Allocation and Recognition of the Anchor Point in PCB Board,TP391.41
  18. Reasearch on the Methods of Sound Dereverberation,O422.8
  19. Pilot-based frequency domain channel estimation algorithm and VLSI Implementation of,TN949.2
  20. The Research & Realization of the Learning System for Chinese Calligraphy,TP391.41

CLC: > Industrial Technology > Automation technology,computer technology > Computing technology,computer technology > Computer applications > Information processing (information processing) > Pattern Recognition and devices > Image recognition device
© 2012 www.DissertationTopic.Net  Mobile