A review of key techniques of vision-based control for harvesting robot

Abstract

Although there is a rapid development of agricultural robotic technologies, a lack of access to robust fruit recognition and precision picking capabilities has limited the commercial application of harvesting robots. On the other hand, recent advances in key techniques in vision-based control have improved this situation. These techniques include vision information acquisition strategies, fruit recognition algorithms, and eye-hand coordination methods. In a fruit or vegetable harvesting robot, vision control is employed to solve two major problems in detecting objects in tree canopies and picking objects using visual information. This paper presents a review on these key vision control techniques and their potential applications in fruit or vegetable harvesting robots. The challenges and feature trends of applying these vision control techniques in harvesting robots are also described and discussed in the review.