Abstract

Video compression without losing the quality of information is more complex and time consuming process in video communication. H.264/AVC is designed and developed to meet corresponding video compression. Motion estimation and motor vector are the two important key techniques considered during video compression. The encoder complexity and process time is increased as demand increases for better quality of video service. The proposed model mainly concentrated on bit rate reduction of H.264/AVC with reduced rate distortion optimization (RDO) computation. Based on the texture information each input frame slices into 16 × 16 and 4 × 4 Macro block (MB) divisions. For each MB a gradient bit cost and RDO value is calculated for the best mode selection. If the bit cost of the MB is less than predefined level then DC mode are directly chosen for frame prediction, similarly the process is repeated for all chroma samples. This process of mode selection minimizes the RDO calculation up to 36 modes, further the complexity of encoder is reduced by using a motion search algorithms. The proposed system is implemented using MATLAB tool with Hexagonal motion search and Binary Search methods to reduce the bit rate video frames. The system performance is analysed by using a PSNR value and bit rate of the Predicted frame.