Research On JPEG2000 Agorithm Bsed On GPU—Wavelet Algorithm Iplemented In The GPU

This paper introduces the JPEG2000 static image compression standards and related major technologies, as well as the working principle of GPU (Graphic Processing Unit), CUDA (Compute Unified Device Architecture) architecture, CUDA program optimization according to the basic steps, the CUDA platform for JPEG2000 algorithm inverse wavelet analysis module parallel design. And in the CPU and the GPU work model, the inverse wavelet transform and optimize experiments.After a number of experiments show that the optimization procedure greatly enhance the rate of decompression time, faster image processing and improved real-time decompression capacity to ensure that the image quality, Therefore, the use of this image decompression algorithm is very effective, this method is entirely feasible. Effective for further research to enhance the inverse wavelet transform, and high-performance image decompression and other aspects of the quality of the reconstructed image has laid a solid foundation. Realization of parallel computing design of GPU on CUDA platform, implemented to achieve a large number of image processing play the inspired role.