An enhancement of the novel cuckoo search algorithm applied in contrast enhancement of gray scale images

Image enhancement is a critical aspect of image processing, aimed at improving image quality for various applications. In this dynamic field, enhancing contrast in grayscale images is particularly significant across diverse domains such as autonomous driving, medical imaging, and pattern recognition. The Cuckoo Search Algorithm (CSA) has emerged as a promising optimization technique for image enhancement tasks due to its simplicity and efficacy. However, existing enhancements of CSA, notably the Novel Enhanced Cuckoo Search Algorithm, suffer from lengthy execution times, potential oversaturation in output, and challenges in convergence. This study proposes modifications to address these issues, focusing on reducing execution time, preserving image details, and improving convergence. Specifically, the modifications involve vectorization of the algorithm's existing code, fine-tuning of enhancement parameters, and replacing Levy flights with the Cauchy operator for better solution exploration. Experimental results demonstrate that the proposed modifications significantly enhance the algorithm's performance, leading to faster execution times, balanced enhancement, and improved overall performance based on the following five metrics: Fitness Value, Peak Signal-To-Noise Ratio (PSNR), Edge Detection, Entropy, and Feature Similarity Index (FSIM). The findings suggest that the Proposed Enhancement of the Novel Cuckoo Search Algorithm that employs Cauchy Operator yields superior results compared to Levy flights, making it a viable enhancement for image optimization tasks.