提升小波图像编码技术实验研究  

Study on Lifting Wavelet Image Coding Technology

作　　者：方志军[1] 潘健[1] 

机构地区：[1]江西财经大学信息管理学院,江西南昌330013

出　　处：《电视技术》2008年第10期8-10,共3页Video Engineering

基　　金：国家自然科学基金项目(60662003;60462003);江西省青年科学家(井冈之星)培养对象计划资助

摘　　要：主要对基于邻域预测、基于线性预测和基于双线性预测的提升算法进行了研究,并结合集合分裂零树(SPIHT)算法,对标准图像进行嵌入式编码。实验表明,在同样的比特率和同样分解级数的情况下,对于纹理丰富的图像,基于双线性预测提升的小波图像编码算法,具有更好的图像主客观质量,最差的是基于邻域的预测方法,而线性预测方法则介于两者之间。Based on neighbor prediction, linear prediction and the double linear prediction on the lifting wavelet are mainly studied, and embedded coding on standard images are processed combined with SPIHT algorithm. Under the same bit rate and the same level of the decomposition, the results of experiment show a better image of the objective and subjective quality which is based on the double linear prediction for the rich texture of the image. And the experiment also shows that the worst is the lifting wavelet based on neighborhood prediction, and the lifting wavelet based on linear prediction method is something in between.

关 键 词：提升小波 邻域预测 线性预测 双线性预测 SPIHT算法 

分 类 号：TN919.81[电子电信—通信与信息系统]