Implementation and Optimization of H.264 Encoder Based on DSP Platform
Xiaozhi Liu
Abstract
Xiaozhi Liu
Abstract
Compared with other previous standards,H.264 standard has great improvement in aspects of hierarchical encoding,intra/inter prediction coding,multi-frame reference and prediction accuracy.It′s practical and necessary that the optimization and implementation of H.264 baseline profile encoder on the platform of TMS320DM642.Based on the architectural features of DSP and the computational complexity analysis of H.264 encoder,the H.264 encoder is optimized from the following three aspects:core algorithms,data transfer and memory/Cache use.The experimental results demonstrate that the optimized H.264 encoder can achieve the encoding speed of more than 24 frames/s for the video sequences with CIF format,which can meet the real-time requirements of the applications.
A significance statement is not available in the OpenAlex record.
A contribution statement is not available in the OpenAlex record.
Method details are not available in the OpenAlex metadata.
Findings are not separately available in the OpenAlex metadata.
Limitations are not available in the OpenAlex metadata.
Application details are not available in the OpenAlex metadata.
Compared with other previous standards,H.264 standard has great improvement in aspects of hierarchical encoding,intra/inter prediction coding,multi-frame reference and prediction accuracy.It′s practical and necessary that the optimization and implementation of H.264 baseline profile encoder on the platform of TMS320DM642.Based on the architectural features of DSP and the computational complexity analysis of H.264 encoder,the H.264 encoder is optimized from the following three aspects:core algorithms,data transfer and memory/Cache use.The experimental results demonstrate that the optimized H.264 encoder can achieve the encoding speed of more than 24 frames/s for the video sequences with CIF format,which can meet the real-time requirements of the applications.
Key concepts: Encoder, Computer science, Digital signal processing, Encoding (memory), Coding (social sciences), Computer hardware, Cache, Frame (networking)