This paper presents a procedure and the results of obstacle-avoidance control of a stationary strawberry-harvesting robot to be used with circulating-type movable bench cultivation. The stereovision unit detects mature and immature fruits from beneath the bench while being conveyed laterally. Based on the positional relation between them, the most appropriate direction for the end-effector approach is determined to avoid collision with immature fruits. The end-effector approaches along the calculated direction and controls its posture at 120 mm and 80 mm front of the target fruit, using visual feedback so as to match the roll angle of the end-effector with the peduncle angle, as estimated using a hand-eye camera. In the peduncle detection procedure, green LED lights are used to emphasize the calyx and peduncle. The proposed stereovision algorithm showed a success rate for detection of the appropriate approach angle of 89-93 . In the insertion test, the hand-eye camera was able to recognize the target peduncle at a success rate of 77-80 by facing the fruit from the most appropriate approach angle, and the end-effector was able to insert a peduncle at a success rate of 73-78 . The end-effector approach along the appropriate approach angle was confirmed to be the effective way to avoid collision with immature fruits, although several multiple insertions were observed. Our proposed procedure proved to work functionally together with the lateral movement of the movable bench.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.