除草作業自動化のための5G通信およびAI画像認識を用いたトラクタの自動操舵に関する研究

抄録

Most of the current autopilot technologies for agricultural tractors use Real Time Kinematic-Global Navigation Satellite System (RTK-GNSS) that provides positioning accuracy of several centimeters. However, the positioning accuracy of GNSS can be degraded in the area around near windbreaks or buildings. This paper proposes an automatic steering system based on Artificial Intelligence (AI) image recognition using a camera. In this paper, since it is difficult to install mount a large high performance PC for AI image recognition on a tractor, the high performance PC for AI processing is set at a remote data processing room. In addition, the image data obtained by the camera on the tractor are send to the PC for AI processing through the 5th Generation (5G) mobile communication system. This study realizes automatic steering of an agricultural tractor by using AI image recognition method. The results obtained by AI image processing are transferred back to the small PC on the tractor through the 5G communication network. The results show that crop row can be correctly recognized by the AI model based on Semantic Segmentation when applying organic dent corn and organic onion. The guidance line are calculated by the least squares method from the crop rows which are recognized by the AI image recognition. By using a 5G connection, images are able to send and receive stably at 10 frame per second (fps).