A Comparative Study of Image Enhancement Methods: A Literature Review

Abstract

Image enhancement in computer vision and image processing problems is routinely used to improve performance by sharpening the image content, increasing contrast or rectifying aesthetic failures. The current paper examines some common techniques used, including black balance correction, dehazing, sharpening, histogram equalization, and CLAHE (Contrast Limited Adaptive Histogram Equalization) considering their applicability in both underwater and medical imagery. The comparison discussion entraps the main literature, using the standard measures like: Peak Signal-to-Noise Ratio (PSNR) and Structural Similarity Index (SSIM). Also, the white balance always ranks the best, providing the 54.8672 dB and SSIM value 0.9991, which significantly underlines the ability to maintain structural integrity and color true-to-life. New real-time-potential deep-learning models, e.g. U-Net, require significant computational power, though. The review was summarized by stating the advantages and drawbacks of each of the methods and concluding on the worthiness of the hybrid designs that integrate multiple methods, the need to adhere to the requirements in the specific domains and how AI features are becoming increasingly important in shaping versatility and automation. Future work must thus be directed to real-time and adaptive enhancement architecture so as to meet the requirements of underwater robotics, medical imaging, and autonomous navigation.