Abstract
Three-D vision sensor to determine the harvest maturity of cabbage heads was developed for use in selective cabbage harvester. The sensor measured a 3-D image of the entire head formation including the outside leaf of the cabbage and the condition of curled leaf. For Early ball, using the sensor based on head size and curled leaf, 77% of the cabbage heads were judged as the best cabbage to harvest in rank 3 while 8% were misjudged in rank 2 (immature cabbage). However, when the firmness parameter of the compression method was combined, the determination accuracy and error of 95% and 8%, respectively, was obtained, when the only the head size factor was considered, the accuracy and error of 91% and 7%, respectively, was obtained for Kinkei 201.