基于DRNAMC表示的快速矩计算算法  

Fast algorithm for computing lower order moments based on DRNAMC representation

作　　者：郑运平[1] 李祖嘉[1] 张佳婧[1] 廖秀秀[1] 

机构地区：[1]华南理工大学计算机科学与工程学院,广东广州510006

出　　处：《华中科技大学学报（自然科学版）》2012年第6期69-72,共4页Journal of Huazhong University of Science and Technology(Natural Science Edition)

基　　金：国家自然科学基金资助项目(60973085);教育部博士点基金资助项目(20120172120036);广东省自然科学基金资助项目(S2011040005815);广东省高校优秀青年创新人才培养计划资助项目(LYM11015);中央高校基本科研业务费专项资金资助项目(2011ZM0074);国家级大学生创新训练项目(111056154)

摘　　要：为了进一步提高低阶矩的计算速度,分析了基于对角优先的矩形非对称逆布局模型编码(DRNAMC)表示的同类块的矩计算的复杂度,给出了基于DRNAMC表示的矩计算的2个重要定理,提出了一种DR-NAMC基于表示的快速矩计算算法.实验结果表明:以Peppers,Boat和Goldhill等灰度图像作为典型测试对象,与流行的基于S树编码(STC)表示和传统表示的矩计算算法相比,基于DRNAMC表示方法的矩计算平均执行速度比基于STC表示方法的矩计算平均执行速度提高了32.92%,比传统表示方法的平均执行速度提高了91.07%,是一种良好的矩计算方法.To improve the speed of computing lower order moments, the time complexity of the moment computation of a homogeneous block was analyzed, based on the diagonal-first rectangular nonsymmetry and anti-packing model coding （D1RNAMC）. Two important theorems of the moment computation based on the DRNAMC were presented. A fast algorithm for computing the lower order moments based on the DRNAMC representation was proposed. By taking three gray images Peppers, Boat, and Goldhill as typical test objects, and by comparing our proposed algorithm for computing lower order moments based on DRNAMC representation with those based on the conventional representation and the popular S-tree coding （STC） representation for computing the lower order moments, the experimental results presented in this paper show that the average executing time improvement ratio of the former over the latter is 32.92 % and 91.07 %, respectively whereas remaining the image quality, and therefore it is an effective algorithm for computing the lower order moments.

关 键 词：图像表示 几何矩 矩计算 矩形非对称逆布局模型 对角优先 S树编码 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]