基于小波变换数据融合的图像边缘检测算法

本文提出了一种基于小波变换数据融合的图像边缘检测算法。该方法对原始图像分别采用小波变换、“Sobel”算子和“Log”算子三种方法提取边缘,再将三种算法结果通过数据融合的手段获得一幅新的边缘图像。仿真实验结果表明,该方案所得到的边缘优于单独采用“Sobel”或“Log”算子所获得的边缘图像,是一种有效的图像边缘提取算法。

A Mixed Scheme Based on Wavelet Transformfor Image Edge Detection

The image edge contains very important information for human visual system. The edge is usually the key feature to distinguish different objects. Therefore, edge detection is one of the most important issues in image processing. Wavelet analysis have good local resolution in both time domain and frequency domain. A new scheme for edge detection is presented in this paper. According to this scheme, wavelet transform is combined with traditional algrithms such as "Sobel"and "Log". Simulation result shows that the new scheme can be used effective to provide clearer and smoother edges than conventional methods such as "Sobel" or "Log" algorithms alone.