High-Resolution Low-Light Image Enhancement Technique Using Edge and Color Information

The paper addresses the challenges in enhancing Low-light images, which are crucial for various computer vision applications like semantic segmentation, object detection, and autonomous driving. Low-light image enhancement is vital because poor image quality in such conditions affects the accuracy and performance of algorithms. Traditional enhancement techniques, such as histogram equalization and gamma correction, improve contrast and brightness but often fail to maintain image details and produce unnatural colors. In contrast, newer deep learning methods like LLNet and EnlightenGAN have made progress in this area. However, these methods still struggle to handle high-resolution Low-light images effectively.