Adaptive Image Enhancement Algorithm for Image Quality Improvement

Abstract

Images are often affected by factors such as lighting, environment, and equipment during daily collection, resulting in a decrease in quality, low contrast, insufficient brightness, and blurry details. To improve the visual quality of images, this paper proposes an adaptive image enhancement algorithm based on image quality. This method first performs grayscale conversion and preprocessing on the input image, and get image parameters from three dimensions: contrast, brightness, and clarity. Dynamically adjust the enhancement parameters of the algorithm based on the read image parameters, including contrast limited adaptive histogram equalization, gamma correction, Retinex simulation enhancement, and sharpening processing. This system can be applied to both color and grayscale images, and the enhancement strategy has high adaptability. The experimental part comprehensively evaluates the enhancement effect using indicators such as structural similarity, peak signal-to-noise ratio, natural image quality assessment, blind reference image quality assessment, image entropy, and color richness. The research results indicate that the method provided in this paper can enrich the image details. However, there are still issues such as limited color enhancement and significant structural changes during the enhancement process. Further research will focus on optimizing enhancement strategies and introducing deep learning methods.