基于融合技术的小波变换和形态学边缘检测算法

Fusion-Based Algorithm of Edge Detection Combining Wavelet Transform with Morphology

提出了一种图像边缘检测的方法.本算法将小波变换和数学形态学相结合,并采用简单的融合策略,确定图像边缘位置.在小波域中,对高频子图像用小波模极大法进行边缘检测,对低频子图像用数学形态学法进行边缘检测,然后采用一定的融合规则分别对高、低频边缘子图像进行融合,最后进行小波逆变换重构融合图像.实验结果表明,该算法融合规则简单,泛化能力强,能有效地抑制噪声,较好地再现图像的边缘信息,是一种有效的图像边缘检测算法.

An algorithm of edge detection is proposed combining wavelet transform with mathematical morphology and using a simple fusion strategy to locate image edges. In the wavelet domain the high-frequency sub-image edges are detected by solving the maximum points of local wavelet coefficient model to restore edges, while the low-frequency sub-image edges are detected by mathematical morphology. Then, both sub-image edges are fused according to certain rules and restructured through inverse wavelet transform. Experimental results showed that the fusion rules are simple with high generlizability, thus making the algorithm effective in noise control to regenerate well the image edges. So, it is an effective algorithm to detect image edges.