Compensation Scheme for Quantized Coefficients in Remote-Sensing-Image Compression

摘要：

For remote-sensing image compression, quantization causes texture distortion. This paper presents a new compensation scheme that improves decompressed images. The compensation model is designed to reconstruct the quantized coefficients in the frequency domain. Moreover, the power of redundancy in the (anti)symmetric tight frames is explored to establish the general principles of redundancy-based compensation. In addition, the compensation location judgement and displacement error estimation are designed for better quality. The experimental results show that the proposed algorithm is more efficient than JPEG2000, extremely for the remote-sensing image and the natural images with high compression ratio.

关键词： image coding remote-sensing image tight wavelet frames postprocessing

作者: Rui Yang Feifei Zhou Bo Li

作者单位: Beijing Key Laboratory of Digital Media, School of Computer Science and Engineering,Beihang Universi Beijing Key Laboratory of Digital Media, School of Computer Science and Engineering, Beihang Univers

会议类型: 国际会议

会议名称: 2010 Second Asia-Pacific Conference on Information Processing(2010年第二届亚太地区信息处理国际会议 APCIP 2010)

会议地点: 南昌

会议语种:英文

页码: 284-287

在线出版日期: 2010-09-17（万方平台首次上网日期，不代表论文的发表时间）