Huisong Gao, Jinlin Xue, and Weihao Gu
Recognition, sorting, image segmentation, manipulator, vegetable harvesting
An automatic sorting system was designed to improve vegetable production efficiency and reduce labour force. Machine vision was adopted to detect dead yellow vegetables that were transported on the conveyor belt, mixed among green vegetables. And a sorting manipulator was developed with three steering gears and some mechanical arms. Dead yellow vegetables were detected and judged by processing and analysis of the captured images. If dead yellow vegetables existed, their positions were calculated, and then the manipulator was controlled to grab and remove them according to the calculated positions after image processing and the motion model and Kalman filter–based position prediction of the dead yellow vegetables. Finally, a prototype of an automatic sorting system was built to assess the performances of recognition and grabbing of the system. Experimental results showed that the recognition and the grabbing rates were up to 98% and 96%, respectively, for 50 experiments. The designed system can recognize, locate and remove dead yellow vegetables effectively.
Important Links:
Go Back
