CCTV YOLOv3 De-identificarion Method and System using Simplified YOLOv3 Scheme in CCTV Environment

摘要

In the CCTV environment, a method and system for blindness using a simplified YOLOv3 technique are provided. In the video blindness method according to an embodiment of the present invention, the encoded video is decoded, and personal information is processed in the decoded video using a simplified deep learning model. This enables real-time video blindness with proper performance while applying the object detection technique of the deep learning technique in the CCTV environment.