The Design Of H.264Video Decoder On Main Profile

With the rapid development of the internet,the vedio has become a most important carrier for peaple to transfer information in life. In order to make the video information is universal, unified video code standard is indispensable.Now there are a lot of mature standards, where H.264/AVC is the current standard, which has a higher compression rate and higher video quality.It’s compression ratio is higher than MPEG-22-3times, and it has very good network compatibility. These new features make H.264/AVC save approximately50%bit rate to provide the same visual effect,so it is very suitable for video transmission and storage.In this paper a decoder is designed and programmed based on the H.264standard.After each module is analyzed,it is found that the time spent by the motion compensation is the largest proportion in each module because it has the highest complexity. If the video quality could not be affected when the coding complexity was reduced, the decoder performance would be improved greatly. So several motion compensation optimization methods to reduce Coding capacity are designed in this paper. In addition to motion compensation, transform and quantization also need a large of caculation, and it’s coding is also complex.According to the H.264,a simplified method is proposed,discussed and illustrated in this paper. And the inverse transform and quantization function module is achieved, at the same time, the algorithm is optimized;The main work is as followed:1. A decoder is designed and achieved based on pc according to H.264.2. Through anslysis of each module,it is found that the motion compensation program running time is the longest. Through analysis,the methods of the luminance edge block anticipation, interpolation look-up table, chrominance interpolation calculation for edge1/8pixel interpolation optimization and ordinary position interpolation optimization are optimized,and the the operation speed is improved. The algorithm is optimized by carrying out the full zero block anticipation and coefficient divided.3. The inverse transform and quantization calculation method is proposed based on the description of the transform and quantization process.The mount of caculation is reduced through the micro adjustment of coefficient and the adjustmentof the operation sequence. The algorithm is optimized by carrying out the full zero block anticipation and coefficient divided.After the module optimization, decoder performance improves significantly. Though image quality falls, but it does not affect the visual effects, and the specific situation is in text.