An innovative approach to multi-camera frame and video concatenation is introduced, tackling challenges in real-time person detection, anomaly identification, and efficient data visualization in surveillance systems. A custom-trained YOLOv8 model optimized for diverse surveillance scenarios enhances person detection capabilities. A novel algorithm efficiently extracts and stitches relevant frames and clips from multiple cameras. The approach achieves a mAP50 of 0.951 for person detection, improving accuracy by 62.8% over 25 epochs. The method optimizes computational resources while providing a scalable solution for large-scale surveillance networks, significantly improving situational awareness and anomaly detection efficiency.