The Research and Design of Related Technology for Video Post-Processing Chip
|School||Hefei University of Technology|
|Course||Microelectronics and Solid State Electronics|
|Keywords||video processing chip de-interlacing improved ELA contrastenhancement image sharpening|
With the rapid development of digital TV, digital video post-processing chipgain a wide prospects for development. In order to effectively solve compatibilityissues of different digital video format, and improve the video image quality, thedigital video post-processing chip is used. In this paper, the basic knowledge ofdigital video is introduced, then the de-interlacing and image enhancementtechnology has been discussed.In order to solve the interlaced video signal defect，then improve the TVdefinition, de-interlacing has become an indispensable technology in current digitalvideo post-processing chip. We analyses all kinds of de-interlacing technique,including de-interlacing algorithm, de-interlacing effect, computational complexity.According to the actual requirement, an efficient motion adaptive de-interlacing isproposed in this paper. The mixing pixels is classified to motion or static region bymotion detection of the same parity four field, then intra-field interpolation methodis used in motion region, inter-field interpolation method is used in statics region.The smooth region adopts the cubic curve fitting interpolation, and the edge regionadopts the improved ELA. The experiment shows that the proposed de-interlacingtechnique not only improve the image definition, but also preserve the edgeeffectively, and achieve better de-interlacing effect.At the same time, in order to further enhance the image quality, we focus onanalysing the video image contrast enhancement and the image sharpening. Andconsidering the characteristic of hardware presents the basic structure of hardwareimplementation. A piecewise linear transformation is used for image contrastadjustment. Owing to traditional sharpening algorithm is sensitive to noise andprone to excessive enhancement, this paper adopts an adaptive un-sharp maskingtechnique to achieve image sharpening. The application of improved Sobel operatorwhich enhance the reliability of edge detection. The experiment shows that theproposed contrast enhancement technique and sharpening technique gain goodeffect. And it’s suitable for our application requirements.