Dissertation > Excellent graduate degree dissertation topics show

Research of Multimedia Enhancement Unit in the Embedded Processor

Author: LiuXiaoDong
Tutor: ZhangPeiYong;DingYong;MengJianZuo
School: Zhejiang University
Course: Circuits and Systems
Keywords: multimedia enhancement unit complex media operation SIMD out of order de-interlacing motion detection wiener filtering motion adaptive
CLC: TP332
Type: Master's thesis
Year: 2012
Downloads: 13
Quote: 0
Read: Download Dissertation


Performance, cost, power, flexibility and many other factors are considered in the multimedia technology of embedded application. Of many solutions, the embedded processor with media enhancement unit is a balanced one and has been widely received attention and research. Based on this background, this paper firstly research a variety of media operations and typical extended instruction sets, and then implement the media enhancement unit of the embedded processor, also a novel de-interlacing algorithm is proposed. The major research and innovations are summarized as follows:1. Research a variety media operations and typical extended instruction sets. This paper studies two important aspects of media technology which are codec and post-processing. MPEG2, MPEG4, H.264, scaling, de-interlacing and frame rate up-conversion are analyzed here including algorithm research, operation extract and experimental analysis. Also the mainstream instruction sets, Amdahl law and many factors that affect performance of processor are absorbed here to develop the media enhancement unit. This paper focuses on the SIMD technology of addition, multiplication, SAD and so on.2. Design of the media enhancement unit. Based on the overall consideration of the existing processor, out of order execution and balanced pipeline division are introduced here. Verified multiplication IP is used to enhance the reliability; Decompose the multiplication from high-bit wide to the parallel low-bit wide is used to reduce the combinational logic delay; the internal reorder module is used to ensure the accuracy of interrupt and exception; clock gating and operand isolation are used to control the power.3. Research of de-interlacing. A novel classification-based motion adaptive de-interlacing algorithm is proposed in this paper. It mainly consists of motion classification, wiener-median and 3-tap median filtering. Experimental results show (?)both of the objective performance and subjective evaluation are satisfied. Also it is a case of media processing, the complex operation are analyzed and mapped to the media instruction set, and the result of media processor and algorithm simulation are comparative analysis.

Related Dissertations

  1. Architecture Design of Digital Video Process Chip and Core De-interlacing Algorithm,TP391.41
  2. The Stereo Matching and De-interlacing Algorithm Based on CUDA,TP391.41
  3. A Video Monitoring System for Anshan Power Supply Room,TP391.41
  4. DM6437 -based video motion detection system design pedestrian,TP391.41
  5. Multi-threaded fusion soccer video semantic analysis and event detection,TP391.41
  6. Moving Detection and Behavior Analyzing in Public,TP391.41
  7. Design of Low Power Pipelined ADC for Motion-Detection CMOS Image Sensor,TN792
  8. The Implementation of Image Detection and Target Tracking System Based on DSP,TP391.41
  9. The Research of Gait Recognition,TP391.41
  10. Research on the Detecting and Tracking of Moving Human Body Based on Monocular Vision Robot,TP391.41
  11. A Research on Motion Detection Algorithm Based on Background Subtraction,TP391.41
  12. FPGA-based digital image processing algorithms research,TP391.41
  13. Research on Image Denoising Approach Based on Wavelet and Its Statistical Characteristics,TP391.41
  14. System for Detecting and Diagnosis of the High Frequency Ventricular Reoplarization Waves and Prediction of Sudden Cardiac Death,R541.7
  15. Study on Channel Estimation of OFDM Systems Based on Combined Estimation in Time and Frequency Domain,TN919.3
  16. Wavelet and Partial Differential Equation Based Image Denoising,TP391.41
  17. Research of Denosing Algorithm Especial for Video Encoder,TP391.41
  18. Based on the Vide of Static Background for Moving Object Detection and Tracking Research,TP391.41
  19. Wavelet-based image denoising medical ultrasound and MRI study,TP391.41
  20. Study on Predicting Wave States Utilizing Ship Motion in Navigation,U661

CLC: > Industrial Technology > Automation technology,computer technology > Computing technology,computer technology > Electronic digital computer (not a continuous role in computer ) > Arithmetic unit and the controller (CPU)
© 2012 www.DissertationTopic.Net  Mobile