Skip to Main Content
A simple and efficient edge detection algorithm based on wavelet transform and morphology is proposed. Combined with dyadic wavelet transform and image gradient calculation, the smoothing filter operators are introduced. Then to the non-uniform illumination problem, a local thresholding, which determines the edges, can be constructed by the algorithm of maximization of between-class variance (OtsuÂ¿s method) and morphological opening. Through the synthetic image experiment, the proposed scheme has the better performance than MallatÂ¿s and CannyÂ¿s algorithms. Experiments on Lena and Peppers benchmark images show that it not only detects the image edge correctly but also has the good suppressing noise ability. Further it is effective to the Bank image darkened gradually.