摘要
提出了一种基于小波变换的图象累进编码传输方案,利用图象内在的自相似性对幅值较小的小波系数进行带间预测,对幅值较大的系数进行分层估值量化,辅以高效的自适应算术编码,形成按重要性顺序排列的图象编码,在很低的比特率下,重构图象仍可辨认;而在相同的比特率下,重构图象的质量一致优于JPEG标准。
Based on the wavelet transforming an image coding algorithm for progressive transmission is presented.By making use of images inherent self similarity,the inter band prediction of the low amplitude wavelet coefficient and the successive approximation quantization of high amplitude wavelet coefficient are completed.Finally,with the help of the efficient adaptive arithmetic coding the image coding in the order of importance is obtained.The reconstructed images are recognizable at low bit rate and their quality are better than JPEG at any bit rate.
出处
《重庆大学学报(自然科学版)》
EI
CAS
CSCD
1998年第1期66-70,共5页
Journal of Chongqing University
关键词
预测
编码
小波变换
累进传输
电视
图象处理
prediction
coding
wavelet transforming
progressive transmission