System design of picking robot based on computer vision

Picking robots are widely used in the harvesting of various agricultural products, which can significantly reduce the labor cost in agricultural production and solve the problem of labor shortage in agricultural production. Computer vision technology can identify ripe melons and fruits from their natural growth environment, which is the key for picking robot to achieve target positioning and coordinate manipulator. In this paper, the method of designing robot picking experiment platform based on computer vision is proposed, and intelligent picking is realized by constructing robot picking experiment platform. The experimental results show that the robot picking experimental platform based on computer vision can solve the problem of low picking efficiency of the previous platform, and can realize the efficient picking of different fruits and melons, which provides a research basis for the picking robot to better meet the actual demand of agricultural production.