Font Size: a A A

Dsp Hardware-based Jpeg2000 Algorithm Design And Implementation,

Posted on:2009-06-05Degree:MasterType:Thesis
Country:ChinaCandidate:H Z ZhangFull Text:PDF
GTID:2208360245461846Subject:Circuits and Systems
Abstract/Summary:PDF Full Text Request
It's no doubt that image is one of the most important means in multimedia information transmission. Image compression is the basis of image processing. Therefore, how important the image compression is! In the field of image compression, JPEG image compression standard is most widely used, however, it established ten years ago, it can't adapt to people's higher compression performance requirements at present. Thus JPEG and ISO committee drafted and formulated the JPEG2000 image compression standard in 2000. JPEG2000 standard not only inherits the advantages of the JPEG standard, but also has low bit-rate transmission, lossless compression, progressive transmission and the coding region of interest and so on. Because JPEG2000 has so many advantages, it has a wide range of applications, such as digital TV, network and multimedia communications, video conference, video telephony, image database and fingerprint identification system.JPEG2000 algorithm is complex and requires high hardware platforms, therefore, common embedded processors can't be competent. However, digital signal processor (DSP) is devoted to the real-time signal processing, with special hardware and software structure of the microprocessor. It provides a good platform for the achievement of JPEG2000 algorithm especially in the high-end TI's TMS320C64x digital signal processor.This paper studies the JPEG2000 image compression standard, and focuses on realizing it in the platform of image acquisition and processing card. To further improve the compression speed for satisfying the image acquisition and processing card requirements, the paper optimizes codes by the means of reasonable definition of data types, data packing, intrinsic functions, optimization options and library functions. Since the operation of discrete wavelet transform (DWT) and Embedded Block Coding with Optimal Truncation (EBCOT) account for 85 percent of the whole time, the paper optimizes discrete wavelet transform with unit of block and pipeline in EBCOT.Through MATLAB algorithm simulation, VC software implementation and DSP hardware simulation, the paper achieves all functions of the JPEG2000. Its compression quality and compression multiples can completely meet the need of real life, which life, which proves that it has better performance than JPEG compression standard, especially in the low bit-rate transmission circumstances.Experimental proved, with means of optimization, JPEG2000's speed in the TMS320DM642 meets requirement of the image acquisition and processing card.
Keywords/Search Tags:JPEG2000, DWT, EBCOT, Adaptive arithmetic coding, DSP, DSP Optimization
PDF Full Text Request
Related items