Abstract
A cherry tomato harvesting robot equipped with a 3-D vision sensor on the end effector was made and fundamental experiments were carried out. The end effector picked fruits by the suction force of an electric cleaner. To avoid the obstacles in it's harvesting locus, the end effector was designed to be bent right and left. To get the image of the objects in the dead angle from the fixed sensor, the 3-D vision sensor was miniaturized and mounted on the end effector so that it can also scan the side field from the space among the stalks and leaves. The experimental results showed that the position of red fruits could be measured accurately. Moreover, it was also possible to recognize the space for the side field scan by processing the 3-D image.
