Image Edge Detection Using Fuzzy Logic Controller

Abstract

Edge detection finds a greater significance in image processing and computer vision, as many machine learning models require images as input data. Edge detection can be used to extract important features to simplify the visual data. With the increased use of AI, latency can be reduced by processing the data locally which enhances the performance capabilities of the model. This paper reviews the effectiveness of the Fuzzy Inference System over traditional gradient-based approaches such as the Canny edge detection technique and presents a fuzzy logic-based approach for image edge detection. The fuzzy-based approach uses an open-loop fuzzy logic controller which comprises a series of steps instead of a simple thresholding techniques whose values are emperically determined. The performance is analysed for implementation in Python and MATLAB Platforms, with some variations in logic for algorithms in each software. The proposed model is applied to MRI images inorder to detect abnormalities such as tumours.