OGIF: A novel Optimized Guided Image Filter for image enhancement

Image processing techniques are applied on various kinds of images such as medical images, microscopy images and remote sensing images for improving the visual qualities of these images as their credibility is questioned because of poor visibility, contrast, acquired noise etc. Image enhancement is necessary for removing the visual ailment of these images. Guided image filter (GIF) has been a suitable filter for image contrast enhancement, but the controlling parameters are taken as constant values in the conventional GIF. These control parameters vary for different kind of images; therefore, require manual tuning and adjustment which is often cumbersome, time-consuming. The proposed work modifies the existing GIF by incorporating performance improvement in terms of optimal selection of control parameters. Thus, a novel optimized guided image filter (OGIF) is developed with the help of particle Swarm optimization (PSO) algorithm. The design methodology involves contrast and edge enhancement of the images which helps in accurate detection of region of interest (ROI) as well as noise removal. OGIF automatically adjusts the value of the controlling parameters in an optimal manner on which the detection of ROI is the most dependants. Image quality assessment (IQA) of enhanced images has been carried out using measurement of enhancement (EME), peak signal-to-noise ratio (PSNR in dB), measure of enhancement error using entropy (EMEE), standard deviation (SD) and entropy. The incremental change in the values of EME and EMEE advocates the suitability of OGIF for contrast enhancement.