基于B样条小波的图像边缘检测算法  

Research of image edge detection algorithm based on B-spline wavelet

作　　者：刘小豫[1] 韩丽娜[1] 赵蔷[1] 

机构地区：[1]咸阳师范学院信息工程学院,陕西咸阳712000

出　　处：《电子设计工程》2013年第14期178-180,共3页Electronic Design Engineering

基　　金：陕西省教育厅专项科研计划项目(09JK811);咸阳师范学院基金项目(12XSYK085;11XSYK210;12XSYK072)

摘　　要：边缘是图像最基本的特征,边缘检测是图像处理中的重要内容。传统的边缘检测方法只根据当前像素点和相邻像素点之间的关系进行边缘检测,抑制噪声效果不好,定位边缘精度较低。为了更好的抑制噪声和精确定位图像边缘,使用3次B样条小波对图像进行多级分解和重构,综合考虑不同尺度上的图像边缘的检测算法。与Canny算子进行了实验对比,结果表明,该算法能够精确定位图像边缘,具有更好的抗噪性能。Edge is the most basic feature of images,so edge detection is an important content of image processing.The traditional edge detection methods only consider the current pixels and the relationship between the adjacent pixels,restrain the noise effect is bad,edge positioning accuracy is low.In order to reduce the noise and the precise location of image edge,is proposed based on wavelet multiresolution image edge detection algorithm.Using three b-spline wavelet image to multistage decomposition and reconstruction,fusion of the different scale edge image.The experimental comparison of our method with Canny operator,the results show that this algorithm can accurate positioning image edge,has better resistance to noise performance.

关 键 词：边缘检测 小波变换 多分辨率分析 B样条小波 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]