Симуляция Распознавания Лиц на CCTV

A high-angle CCTV security camera view based on the reference photo, maintaining the original environment, layout, and perspective. White rectangular bounding boxes frame the face of every visible person in the scene. For the most prominent subject, a large floating rectangular inset displays a sharp, high-resolution zoomed-in close-up of their face, linked to their face box by a thin white line. The overall surveillance feed has realistic camera artifacts, including soft film grain, slight wide-angle lens distortion, and muted, desaturated colors, while the zoomed inset is clear, bright, and finely detailed. No timestamps, text, or interface overlays other than the bounding boxes and connecting line.