Advances in the field of computer vision, machine vision, and deep learning over the recent past have been noteworthy and have considerably influenced the solutions discovered to this important problem area, namely that of low-light image enhancement, within areas including surveillance, autonomous vehicles, and medical imaging. In this paper, a general review is presented using deep learning techniques for improving the quality of images captured with low-lighting conditions. Under such conditions, poor visibility, noise, and a loss of detail generally impair accurate analysis. A novel deep learning approach is proposed with a specific CNN architecture that has been trained on a large and diversified dataset of lowlight images. It shows outstanding capability in the restoration of image clarity and real-time improvement of visual quality. The computational efficiency and real-time feasibility of the proposed solution are evaluated, along with a promising application outlook for various practical scenarios. This model's behavior on samples taken under other low-light conditions and its generalization capability to diverse types of images is also discussed. In this sense, this research feeds into the perpetual discourse for efficiently overcoming poor imaging conditions with deep learning and promising avenues for further development within the domain of low-light image enhancement.