Zernike亚像素图像矩边缘检测算子模型研究  

Research on Sub-Pixel Zernike Moments Edge Operator Based on Direction Angle Model

作　　者：姜迈[1] 梁晔[2] 梁炜[1] 张晓玲[1] 

机构地区：[1]中国科学院沈阳自动化研究所工业控制网络与系统实验室,沈阳110016 [2]北京联合大学北京市信息服务工程实验室重点实验室,北京100101

出　　处：《计算机测量与控制》2013年第4期874-876,共3页Computer Measurement &Control

基　　金：国家自然科学基金(61174026);国家863计划(2011AA040101);973计划前期研究专项课题(2010CB334705);中国科学院知识创新工程重要方向性项目(KGCX2-EW-104-2)

摘　　要：基于Zernike图像矩的理想边缘模型,利用Zernike矩定义及其旋转不变特性,提出一种新的基于4阶方向角的Zernike矩亚像素边缘检测算子;为了提高边缘算子定位速度,首先基于9×9尺寸模板对Zernike图像矩0到4阶正交复数多项式进行了计算,推导出基于4阶方向角的边缘检测算子参数模型;最后将边缘算子应用在理想图像与实际图像上,检测结果表明,相比于目前提出的算法,基于4阶方向角的边缘检测算子具有更高的检测精度。Based on the Zernike image moments edge model, provides a new edge operator method based on 4 orders direction angle by u-sing the Zernike moments definition and its rotation invariance. In order to improve the edge operator positioning speed, the orthogonal com- plex polynomial from 0 order to 4 orders of the Zernike moments based on 9 × 9 mask coefficient was given, the edge operator parameter mod-el was deduced. Finally, the edge operator is applied to the ideal binary image and a real image, the edge detection results showed that the 4 orders direction angle model has a higher accuracy compared to the existed methods.

关 键 词：ZERNIKE矩 亚像素 正交复数多项式 4阶方向角 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术]