AI CCTV CCTV-based missing person search method using face prediction according to age change and AI

• Main Authors: KANG WOO SUK, JEONG JAE HONG
• Format: Patent
• Language: English; Korean
• Published: 15.04.2022
• Subjects: CALCULATING; COMPUTING; COUNTING; ELECTRIC COMMUNICATION TECHNIQUE; ELECTRICITY; HANDLING RECORD CARRIERS; PHYSICS; PICTORIAL COMMUNICATION, e.g. TELEVISION; PRESENTATION OF DATA; RECOGNITION OF DATA; RECORD CARRIERS

The present invention relates to a closed cable television (CCTV)-based missing person search method using face prediction according to aging and AI. According to the present invention, the method predicts a face of a missing person who has changed over time with a third party of video management system (VMS) recognizing video signals captured through a CCTV and compares the predicted face with database data to identify a face, thereby finding out the missing person. In the case of a montage, a missing person will be found within one minute so that a montage prediction time can be significantly reduced and face prediction according to age change of an existing montage prediction program and similar faces are searched through a CCTV, thereby providing an effect of being used for a practical search for a missing person. The method comprises a photo input step, a face detection step, a face prediction step, a recorded video input step, a face recognition step, and a face matching step. 나이변화에 따른 얼굴예측 및 AI를 활용한 CCTV 기반의 실종자 검색방법에 관한 것으로, CCTV(Closed Circuit Television)를 통해 촬영된 영상신호를 기록하는 VMS(Video Management System)의 3rd-party로 세월의 흐름에 따라 변화한 실종자의 얼굴을 예측하고, DB데이터와 비교 후 인지해 내는 얼굴 예측 및 얼굴인식하여 DB와의 매칭시켜 실종자를 찾을 수 있도록 한 것을 특징으로 하며, 실종자 몽타주 예상을 할 경우 1분 안에 이뤄지며 이는 몽타주 예측시간을 획기적으로 줄이면서 기존의 몽타주 예측 프로그램의 나이 변화에 따른 얼굴예측 뿐만 아니라 CCTV를 통해 유사 얼굴에 대해 검색함으로써 실종자에 대한 실질적인 수색에 활용 가능한 효과가 있다.