Machine Vision System Development for Shake and Catch Cherry Harvesting
Sponsor Agency: CPAAS
Amatya, S.; Karkee, M.
This project is aimed at identifying branches of cherry trees for harvesting cherries using mechanical limb shakers. Automation in cherry harvesting using mechanical shakers requires sensors that can guide robotic arm to the branches. The machine vision system will help to identify branches for locating shaking points. Color cameras are used to acquire images of cherry trees. Image processing techniques are used to segment branches from the background. Processing images captured from different cameras by single algorithm is challenging due to variation in camera parameters. Therefore, branch identification was done with images captured from a single camera. In the future, the algorithm to identify cherry branches will be further improved.