System Research of Pear Picking Robot

Abstract

China is a big producer and consumer of pears, and Chinese pear industry has long been based on labor as the main productive force. With the decline of the demographic dividend, labor costs have gradually increased, and labor efficiency has gradually become difficult to meet the needs of modern production. The development of a pear picking robot and its application to actual picking operations is an urgent research. Based on the investigation of the planting environment of pear orchards, this paper designs and builds a pear picking robot. The picking robot consists of a right-angle mechanical arm, a moving system, a control system, and a visual recognition system. On account of the deep learning algorithm, it can accurately identify and locate the pears, and then pick the fruit in a way that the external force is released, so as to realize the damage free picking of the pears. The Outdoor picking experiment shows that the picking robot designed in this paper has a good picking rate in different time periods.