STUDY OF EDGE DETECTION IN COLOURED IMAGES USING ZERO-CROSSING TECHNIQUE

Abstract

The "Zero—Crossing Technique" for edge detection is a proven powerful technique for black and white images. Laplacian of Gaussian filter applied on gray scale images gives a better edge detection quality as compared to other classical edge detection operators. An effort has been made to use this technique for coloured imageries. Edge detection in coloured images is the need of the day. In this dissertation, Laplacian of Gaussian filter is. applied on variety of coloured images containing mainly edges. Later three bands of colours (Red, Green and Blue) were separately observed on Silicon Graphics work station from edge detection point of view. In addition, various image enhancement tools have been developed such as smoothening; sharpening and diffusing of images using Visual Basic. A custom oriented filter is also developed which can be used for improving image quality as user's choice. All the processed images have been studied using silicon graphics work station & results are presented in chap # 5. The packages for these operators, so developed perform relatively better than those available on Silicon Graphics /MATLAB. Ultimately it was found that the "Zero Crossing Technique" is quite suitable for the detection of edges in the coloured images.