An improved morphological gradient edge detection algorithm

Morphological gradient is a nonlinear edge detector that can overcome some drawbacks of the linear edge detectors. Wide edges in edge image detected by morphological gradient cause the low edge resolution. According to the inherent properties of the impulsive noise and reasons that wide edges produce, an edge detection algorithm is presented that the gradient image is segmented in two orthogonal orientations and local maxima are derived from the section curves. The algorithm´s steps are presented in detail. The experimental results show our algorithm can improve the edge resolution and insensitivity to noise.