摘要
提出了一种检测图像边缘的方法。该方法在满足Canny提出的边缘检测 3个准则的基础上 ,构造二维可分离滤波器组 ,利用B样条小波对图像函数进行小波变换 ,提取图像边缘 ,较好地解决了图像边缘的提取精度与图像噪声的抑制能力之间的矛盾 ,并采用了多孔算法防止图像信息丢失。实验证明了B样条小波检测图像边缘的有效性。
A method to detect image edges is presented, attempting to construct separable 2-dimensional filter banks on the basis of three criterions required in edge detection proposed by Canny. A wavelet transform is performed on the image function to detect an image edge using B-spline wavelets. The contradiction between the detection precision of the image edges and the restraining ability of an image noise is well resolved by this method. Algorithm a'trous is adopted to prevent the loss of image information. Experiments prove the veracity and effectiveness of this method.
出处
《武汉理工大学学报（信息与管理工程版）》
CAS
2002年第3期31-33,共3页
Journal of Wuhan University of Technology：Information & Management Engineering
基金
湖北省重点基金资助项目 (97C -10 4)
关键词
小波变换
滤波器组
边缘提取
B样条
wavelet transform
filter bank
edge detection
B-spline
