Edge Detection Based on Global and Local Parameters of the Image

This paper presents an edge detection method based on global and local parameters of the image, which produces satisfactory results on the edge detection of complex images and has a simple structure for execution. The local and global parameters of the image are arithmetic means and standard deviations, the former acquired from a three sized window representing five pixels, the latter acquired from the entire row or column. We obtain the differences of grayscale intensities between two adjacent pixels and the sum of the modulus of these differences from the horizontal and vertical scans of the image. Using these obtained values, we calculate the local and global parameters. After the gathering of the local and global parameters, we compare each sum of the modulus of differences with its own local and global parameter. In the case of the comparison is true, the consecutive pixel to the modulus sum of differences index is marked as an edge. We present the results of the tests with grayscale images using different parameters and discuss the advantages and disadvantages of each parameter value and algorithm structure chosen on the edge processing. There is a comparison of results between this papers detector and Canny, where we evaluate the quality of the presented detector.