The Intra Prediction Unit Fast Selection Of HEVC And Disparity Map Generation

High Efficient Video Coding(HEVC)has significant improvements compared with the last generation H.264/AVC in compression performance and parallel processing.However,the high performance of HEVC is at the cost of the higher computation complexity which restricts its promotion in practical.Therefore,how to ensure the video quality and reduce coding complexity simultaneously in practical application is a significant research content of HEVC.In this thesis,based on the research of HEVC standard and its key technologies,the improved algorithm is proposed to reduce complexity from block partitioning in intra prediction.According to the complexity of the intra prediction coding unit,this paper presents a fast algorithm for CU block partitioning.This algorithm integrates K-means clustering algorithm into the process of coding unit.Firstly,from the characteristics of the complexity of image texture,the relationship between the complexity of image texture and the size of CU is obtained,and the variance of CU block partitioning is used to present the complexity of image texture.Then,we select the image texture complexity as the feature vector in order to obtain the clustering center of CU block partitioning.Finally,the original HEVC algorithm is combined with the K-means clustering method and then the CU block partitioning is divided.According to the problem of the edge and the occlusion region of the global matching algorithm,this paper proposes a global matching algorithm based on the guided filter and Mean Shift.Matching cost function was built according to Markov Random Field(MRF)and guided filter,the initial disparity value of each pixel was calculated by minimizing the corresponding matching cost function.Mean Shift algorithm and Super-resolution reconstruction algorithm were used to optimize the disparity map.The results of proposed method was tested on international standard Middlebury stereo benchmark and compared with some other methods,the compared results showed the high accuracy of the proposed algorithm and its superiority to some prevailing algorithm.