Image Quality Assessment: Edge Based Entropy features estimation using Soft Computing Techniques

Abstract

With the always expanding interest for image handling based applications, the requirement for Image Quality Assessment (IQA) techniques for proficient and dependable assessment of picture quality in concurrence with human quality decisions has similarly taken off at each phase of image processing. The limits of customary IQA strategies as far as having the option to work just under a pre-characterized set of filtering algorithms confines the effectiveness of methods in explicit degradations has roused the investigation of IQA dependent on Soft Computing metaheuristic techniques, which are equipped for choosing the features shrewdly. A swarm based meta heuristic technique known as Grey Wolf Optimization (GWO) has been utilized, which impersonates the conduct of wolf packs to encompass and chase their prey. This paper presents a basic and direct strategy for artificially assessing computerized images by a solitary parameter, entropy. The benchmark dataset Colourlab Image Database: Image Quality (CID:IQ) is employed for the extensive Matlab investigations. Outcomes are addressed graphically to approve the viability of GWO algorithm over others. GWO evidently surmounts Artificial Neural Networks (ANN) based optimization algorithm in terms of entropy. These meliorations are likely to propel the future researchers to prefer the soft computing techniques for use in different image handling areas including horticultural sciences, medication, cryptography, and so on, for evaluating their images in pre-processing phases for improved outcomes.