Year: 2015
Type: article
Abstract: In the JPEG2000 standard for image compression, quantization is a necessary lossy tech-nique to significantly reduce bit number in all sub-bands. In this paper, mantissa and expo- nent coefficients have been found for the best solution for the 8-bit and 16-bit images in the JPEG2000 image compression standard. Furthermore, the method also divides image data into code-blocks sizing 16x16, 32x32, and 64x64. The code-block process supports the real time tier-2 encoding implementation. Experiment has been done on the FPGA to demonstrate the effectiveness of this proposed method. Matlab, Modelsim VCS tools and Quartus II tools have also been used for testing.
Language: Vietnamese
Subfield: Computer Vision and Pattern Recognition
Field: Computer Science
Domain: Physical Sciences
Open Access status: diamond
OpenAlex ID: W7217560948