Comparative Analysis of Object Detection and Recognition in Low Light Image Techniques

Object detection in low-light conditions is crucial for many important applications, such as autonomous vehicles and visual surveillance. However, low light, noise and shadows pose major challenges. Modern machine learning models, including deep learning-based techniques, are analyzed, as well as traditional techniques such as graph equalization, Retinex based methods, and others. this paper provides an overview of developments in object detection methods, with a focus on cutting-edge frameworks such as YOLO, which detects objects by classifying them by object type. The enhanced images are sent to object detection algorithms, which contribute significantly to the development of carefully designed specialized techniques for accurate object detection. The paper consists of two main stages: the first is related to improving the image and overcoming all the problems that lead to inaccurate images, and the second stage is related to how to detect objects in this image, whether using traditional methods or modern methods. These developments are expected to significantly improve the effectiveness of computer vision systems in practical situations, helping to develop systems that work in difficult conditions such as lack of light in images or strong reflection that affects the vision of objects.