The principles and advantages of the RetinexNet model, built upon Retinex theory to enhance image quality under challenging lighting conditions are presented. The methodology involves decomposing the image into reflectance and illumination components, enabling flexible exposure and contrast control, even with limited computational resources.