Efficient H.264 Intra Frame CODEC with Best Prediction Matrix Mode Algorithm

Abstract

The continuous growth of smart communities and ever-increasing demand of sending or storing videos, have led to consumption of huge amount of data. The video compression techniques are solving this emerging challenge. However, H.264 standard can be considered most notable, and it has proven to meet problematic requirements. The authors present (BPMM) as a novel efficient Intra prediction scheme. We can say that the creation of our proposed technique was in a phased manner; it's emerged as a proposal and achieved impressive results in the performance parameters as compression ratios, bit rates, and PSNR. Then in the second stage, we solved the challenges of overcoming the obstacle of encoding bits overhead. In this research, we try to address the final phase of the (BPMM) codec and to introduce our approach in a global manner through realization of decoding mechanism. For evaluation of our scheme, we utilized VHDL as a platform. Final results have proven our success to pass bottleneck of this phase, since the decoded videos have the same PSNR that our encoder tells us, while preserving steady compression ratio treating the overhead. We aspire our BPMM algorithm will be adopted as reference design of H.264 in the ITU.

Keywords

• best prediction matrix mode (bpmm)
• compression efficiency
• h.264/avc
• huffman coding
• intra prediction