Robust Cabbage Recognition and Automatic Harvesting against Environmental Changes

Abstract

In Japanese agriculture, labor shortages are becoming increasingly serious due to the lack of farmers and aging. Therefore, research has been conducted on the automation of cabbage harvesting. There are problems that cabbage detection fails due to the backlight and controlling harvesting part does not work because of detecting cabbages in the back row. In order to solve these problems, we proposed new recognition methods in this paper. We worked on cabbage detection using lower half of the cabbage and cabbage selection using a depth camera. Moreover, sliding mode control was also incorporated to enable automatic harvesting in soft soil. The experimental results are shown to confirm the effectiveness of these proposed methods.