구조화된 광원을 갖는 단일비젼을 이용한 3차원 측정

Abstract

Industrial robots have been developed for several decades. Nowadays, robots equipped with 3D vision system are being widely deployed in increasing numbers into variety industries, including food, pharmaceutical, vehicle assembly, aerospace, etc. 3D vision system in indus-trial robots requires high levels of repeatability, modest in hardware, high accuracy and real-time performance. Toward this end, in this work, a 3D vision system using monocular camera and structured light is developed. The proposed 3D measurement sensor projects laser lines onto the target objects that are observed by the high-resolution camera. The measurement is initiated when the operator brings the device closer to the surface until it is within operating range. During the process, the line features are digitized by using proposed approach, the de-sired calculations are made, the non-conforming images are discarded, and the remaining im-ages are used to perform the 3D measurement. The measurement system can deal with com-plex surface in noisy industrial environment and achieve higher specifications compared with current 3D measurement sensors. The usefulness of the proposed system has been demon-strated using real tests with accurate know-size patterns and a real inline vehicle assembly sys-tem in Korea.