Computer vision for enhancing security and surveillance in a public space

One of the main issues in computer vision is object and action recognition. Object detection is to find anything in the scene, and action recognition tracks the subject’s movements. This paper proposes a method for improving real-time surveillance systems with OpenCV library and an upgraded YOLO v8 (You Only Look Once-version 8). The primary goal of the paper is surveillance and security. This study examines the training of computer systems to recognize digital patterns for supporting surveillance activities such as monitoring anomalous activity and criminal activity. It draws attention to the possibility of enhancing current object detection frameworks with cuttingedge artificial intelligence features to produce surveillance systems that are more perceptive and adaptable.