Enhanced object detection in videos using bio vision

Video surveillance systems enhance security and monitor various environments. They act as a deterrent to potential threats and provide vital surveillance to maintain safety. A key component of these systems is object detection, which identifies objects of interest and generates critical data for tracking and analysis. However, detecting objects in low-light conditions is particularly challenging due to reduced visibility, low contrast, and shadows, which can obscure objects or cause them to blend into the background. To address these challenges, this research introduces an enhanced object detection model that combines bio-inspired vision techniques with traditional object detection methods. The YOLO model is augmented with bio-vision principles, and specialized CNN models are developed to improve detection accuracy, particularly in low-light scenarios.