Font Size: a A A

Research Of Code Optimization Based On TMS320DM642

Posted on:2014-01-25Degree:MasterType:Thesis
Country:ChinaCandidate:H Y ShiFull Text:PDF
GTID:2248330398970530Subject:Communication and Information System
Abstract/Summary:PDF Full Text Request
Along with the swift progress of embedded technology, embedded chip, represented by ARM and DSP, owns a wide range of applications, such as vedio cassette recoder, automobile, microwave oven, etc. On the other hand, multimedia technology, including video, image compression and processing, plays more and more important role in people’s life. Due to the needs of miniaturization of terminal equipment, embedded platform and multimedia technique combine with each other increasingly close. As for its high operation speed and has strong multimedia data processing ability, DSP chips are widely used in intelligentmultimedia systems. This article discusses code optimization technology of DM642based on H.264vedio encoder and face detection and recognition system.This paper analyzes the most important and complex parts of H.264encoder-intra prediction and inter prediction, improves the initial algorithm on the basis of understanding them. With pre-judgment of intra44mode selection and the optimization of SKIP mode decision criterion, the encode speed improves a lot with little decrease in vedio quality.In addition, this paper also studies the Gabor transform in face recognition system thoroughly and apply real discrete Gabor transform to the system. The arithmetic speed increases a lot while the result is the same as initial.
Keywords/Search Tags:DSP DM642, H.264, Face-recognition, Gabor-transform, Optimization
PDF Full Text Request
Related items