On edge detection of images using ant colony optimization and fisher ratio

Abstract

Edge detection is one of the important parts of image processing. It is essentially involved in the re-processing stage of image analysis and computer vision. It generally
detects the contour of an image and thus provides important details about an image. So, it reduces the content to process for the high-level processing tasks like object recognition and image segmentation. The most important step in the edge detection, on which the success of generation of true edge map depends, lies on the determination of threshold. In this work, purpose of edge detection, inspired from Ant Colonies, is fulfilled by Ant Colony
Optimisation (ACO). For the determination of threshold calculation, a novel technique of Fisher ratio (F-ratio) is used. The success of the work done is tested visually with the help of test images and empirically tested on the basis of several statistical parameter of comparison. De-noising is the process of extracting the important features present in an image, keeping the unnecessary or unimportant information present in the form of noise out as much as possible. There are many Denoising methods that have been developed in these field, but the most trustworthy and used among them is the wavelet thresholding denoising method with hard thresholding. The proposed novel method presented in this thesis is tested on the denoised images. The Edge detected images obtained on the denoised images are showing better results than the other conventional edge detectors.