AN EFFICIENT EDGE DETECTION APPROACH USING DWT

In today's digital world, image edge detection is widely used in all over the globe. Digital image edge detection is used to detect the sharp changes in brightness of the images and the edge of a digital image. There are many classical techniques used to detect image edge detection like Canny, Sobel, Prewitt and Robert edge detector method. The proposed research work mainly focuses on the study of different edge detection methods which were previously derived and to propose an efficient method of digital image edge detection technique using Discrete Wavelet Transform (DWT). The results were compared with one of the best edge detector method: Canny Edge Detector. In the proposed image edge detection method, the first analytical result showed that an input image after applying the thresholding technique represents a segmentation of image edges and the second analytical results showed the 8- connected components of pixel connectivity in two dimensional images related to their neighbor pixels producing the enhanced edges. To evaluate the performance of the proposed algorithm with the existing algorithms, the two thresholding values were applied as T_low and T_high using hysteresis thresholding method. PSNR and width σ were also computed to detect the efficiency of the proposed algorithm.