Multiscale edge detection

Fig. 24 Multiscale edge detection test image (top). size 256 x 256 pixels. Only the wavelet edge detector (middle) is able to detect all edges, while the Canny edge detector (bottom) misses some important features. The test image irai kindly provided by Prof. J.M.H. du Buf, Vision Laboratory, Department of Electronics and Computer Science,...

In this work a multiscale edge detection that is based on the wavelet transform modulus maxima (WTMM) is used to segment tumors in FFD mammograms. The ARS PCS with MDCS is developed and used as a platform to process the segmentation. Parallel technique and computation used in processing the mammograms adopted work done by My-lonas [5] and Srivinasan [4]. 

A local maximum is a value that is maximum within some specific neighborhood. The local maxima of a wavelet transform modulus is closely related with multiscale edge detection that concerns about contours or boundaries of objects. From the multiscale edges detected by the WTMM the original signal is reconstructed. 

According to Mallat et.al (1992), a multiscale edge detection in two dimensions can be formalized through a wavelet transform defined respectively by the following wavelets... 

---