Research On Image Assessment Method And Its Application In Video Coding

Image quality assessment plays an important role in the field of image processing. It can assist in the optimal design of bit assignment algorithms in image / video compression coding, and determine the quality of compressed images or videos; it can be used to benchmark image processing systems and algorithms, and determine the development of image / video processing algorithms. For that images are ultimately to be viewed by human beings, objective image quality assessment metric has to consistent with subjective assessment. However, the traditional objective image quality assessment algorithm-MSE does not correlate well with the HVS (Human Visual System). Therefore, designing a more reasonable image quality assessment metric, and applying it in the image / video processing, is not only the new direction for image / video processing technology, but also the inevitable trend of multimedia technology development.Currently, a new image quality evaluation method - structural similarity (SSIM) is proposed by Zhou Wang etc. This paper will introduce SSIM to H.264 interframe coding, and the SSIM based rate-distortion optimization algorithm (SSRDO) is proposed. After that, a new image quality assessment metric which is based on information classification is proposed. The main achievements involve the following aspects:1,In this paper, SSIM based rate-distortion optimation algorithm(SSRDO) for inter-frame coding in H.264 is proposed, and the results show that it has a better performance than the original rate distortion optimization used in H.264.2,Based on the observation that human will first classify the image information before acquire information from an image, a new image quality assessment which is based on information classification is proposed in this paper. Experimental results show that, after introducing the classification algorithm in MSE or MSSIM, the algorithm performance has greatly improved.