현장의 3차원 형상을 이용한 CCTV의 내부 및 외부 파라메터 예측

초록

The CCTV camera images obtained in a crime scene provide clues for the police investigation. Recently, a technique for predicting a three-dimensional gait shape of a criminal from a single CCTV camera has been developed. To accurately predict the gait shape, the internal and external parameters of the CCTV camera should be accurately calculated. In this study, the three-dimensional space of the crime scene was reconstructed using multiple view images. The internal and external parameters of the CCTV camera were estimated from the three-dimensional scene geometry. The corresponding points in the three-dimensional scene geometry and CCTV camera image were manually matched. The matching pairs were used to calculate the optimal values of the CCTV camera parameters. The objective of this study was to develop a method to estimate the internal and external parameters of the CCTV camera using the matching points between the three-dimensional scene geometry and CCTV camera.

키워드