Optimizing Grayscale Image Quality Through Genetic Algorithm Over Quad Histogram Division

In medical images, poor-quality images such as X-rays, MRI scans, CT scans, and ultrasound are common. These images can make it difficult to accurately interpret the affected areas in the body. To improve image clarity and computational accuracy, the proposed method employs an eight-degree polynomial-based histogram clipping method for controlling global contrast enhancement. Additionally, the method determines exposure threshold value and Genetic Algorithm (GA) to optimize histogram divisions, ensuring consistent brightness and minimizing information loss through a single objective fitness function. To evaluate the quality of the enhanced images, several metrics are taken into consideration such as Peak Signal-to-Noise Ratio (PSNR), Entropy, Feature Similarity Indexing Method (FSIM), and Structural Similarity Index (SSIM).