VR支援による果実収穫ロボットシステムの開発 (第2報)

ビジュアルフィードバック制御

抄録

A prototype system has been developed, for the purpose of realizing a fruit harvesting system, here by us. This prototype system consists of 1) a robot with an articulated arm that has a small CCD camera in its hand and 2) real-time image processing technology.
This single camera can measure the three-dimensional position of a target in space by means of moving the camera after catching the target in its sight. The parameters of the camera system can be determined automatically by using image processing to measure a standard mark put on a known position. The robot system has the ability to learn the visual features of a target, extracting them from an image for example that of a tomato or an apple. Using these features the target can be recognized.
Results obtained from harvesting trials using tomatoes and apples have shown that prospects of practical speed and precision of harvesting are evident.