摘要
首先提出图像粒的概念,然后将所提出的概念应用于QaR树进行图像分解,最后对基于QaR树分解的结果和其他算法进行评估和对比。实验结果表明,在冗余度和精度两个指标上,基于QaR树的图像粒分解结果均优于已有方法,能够提供更为贴近图像数据和图像空间的图像局部区域。
The novel definition of image particle is proposed firstly. Then procedure of image particle decomposing is accomplished via QaR tree. Finally, the optimal comparative evalua- tion from QaR tree to other classic algorithms is given. Experimental results reveal that on the basis of accuracy and redundancy, QaR tree method holds better effects than other exist- ing methods, and its results are closer to local image regions consisting of image space and image data.
出处
《武汉大学学报(信息科学版)》
EI
CSCD
北大核心
2013年第2期204-207,共4页
Geomatics and Information Science of Wuhan University
基金
国家973计划资助项目(2010CB731801)
国家自然科学基金资助项目(61172174
N010978003)
国家科技专项资助项目(2012YQ16018505
2013BAH42F03)