Automatic Onion Harvesting by Height Control of Harvesting Part with a Depth Camera

Abstract

Recently, the labor shortage is becoming more serious because of decrease and aging of farmers in Japan. As a solution，therefore, the automation by agricultural robots is trying to expand the scale of management and labor saving. In order to automate harvesting, we use semantic segmentation, a method of deep learnig, to discriminate onion, soil, and others in pixel units. Then, the height of only the soil part detected using a depth camera is measured. Harvesting with high precision is performed by using the information to automate height control of the harvesting part. The usefulness is confirmed by some experiments.