带细节增强与降噪的真彩图像多尺度边缘检测  

True color image multi-scale edge detection with detail enhancement and noise reduction

作　　者：肖锋[1,2] 周明全[3] 耿国华[1] 

机构地区：[1]西北大学可视化技术研究所,西安710127 [2]西安工业大学计算机科学与工程学院,西安710032 [3]北京师范大学信息科学与技术学院,北京100875

出　　处：《计算机工程与应用》2011年第34期8-10,28,共4页Computer Engineering and Applications

基　　金：国家自然科学基金(No.60873094)~~

摘　　要：为了解决真彩图像降噪和边缘细节保留的问题,提出了使用一种软阈值方法进行细节增强与降噪的真彩图像多尺度边缘检测算法。使用小波多尺度真彩图像边缘检测算法得到不同尺度边缘图像,进而根据改进的软阈值滤波函数,对得到的边缘图像进行阈值选择,降噪的同时也增强了保留的边缘细节信息;最后对不同尺度边缘图像进行了加权二范数融合。实验证明,该算法能充分利用真彩图像的颜色和梯度信息,有效地抑制噪声,增强图像的细节信息。To tackle the problem of de-noising and edge detail preservation of true color images,this paper proposes a multi-scale edge detection algorithm which takes soft threshold method to implement detail enhancement and noise reduction of the true color image.Three steps are needed：obtaining the true color images at different scales through wavelet multi-scale edge detection algorithm;based on the improved soft threshold filter function,selecting appropriate threshold of the obtained image edges to perform noise reduction while enhance the edge details of the reservation;carrying out the weighted 2-norm fusion of edges of different-scale-image.Experimental results show that the algorithm can make the most of color and gradient information of true color images to effectively suppress noise,enhance the image edge details as well.

关 键 词：边缘检测 真彩图像 边缘融合 小波变换 增强细节 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]