Low Quality CCTV Image and Video Enhancement using GANs and Audio Enhancement with Equalization Algorithms

In contemporary times, the proliferation of surveillance systems has witnessed exponential growth, resulting in an escalation of both data volume and computational demands for thorough video analysis. Compounding this challenge is the intentional recording of CCTV footage at lower resolutions to accommodate constrained storage capacities. While this practice enables the storage of more videos, it complicates the analysis process by compromising key investigative details. Furthermore, the adoption of space-saving recording modes can adversely impact audio quality, further hindering investigations. In response, our goal is to develop an application that addresses these issues, facilitating a more straightforward analysis of footage even under suboptimal conditions. The proposed application will employ deep learning techniques to upscale video resolution, enhancing feature extraction. Additionally, it will utilize YOLO for object detection and incorporate deep learning methods to reduce audio noise, ensuring clear audibility of the subject’s intent.