摘要
讨论一种利用图像重要性测度实现静态图像压缩标准JPEG2000编码方法。JPEG2000编码器的技术核心是离散小波变换和EBCOT,而EBCOT是由两个编码引擎T1,T2组成的,其中T2编码主要是完成码流的组织,其方法是灵活多变的,用户可以根据特殊要求组织码流,只要码流的格式符合JPEG2000的格式即可。本文根据编码图像小波域的系数和时域的像素值计算该系数的重要性,并利用重要性组织完全符合JPEG2000格式的码流。可以利用标准JPEG2000解码器对本算法形成的码流进行解码。
A method is proposed for using Image Importance Measure in the new international standard JPEG2000 for still image coding. The kernel algorithms in JPEG2000 are DWT and EBCOT. EBCOT is comprised of Tier1 encoder and Tier2 encoder. Tier2 is responsible for the organization of the code stream. According to specific requirement, the code-stream can be organized in different ways. As for the JPEG2000 standard, importance measure is a good choice. The importance measure is computed using the coefficients in the wavelet domain and the pixel values. Importance measure is then used to organize the code stream to the JPEG2000 format. Such encoded code stream can be decoded by means of JPEG2000 decoder.
出处
《电路与系统学报》
CSCD
2003年第5期145-148,共4页
Journal of Circuits and Systems
基金
"十五"军事通信技术预研基金资助项目(410010302)
国防科技预研基金资助项目(J20.12.1.DZ01)