배 과수원 대상 딥러닝 기반 지능형 방제 시스템: 예비실험 결과

Abstract

In this paper, we proposed a deep learning-based intelligent spraying system for a pear orchard. A fruit tree detection system was developed using the SegNet model, one of the semantic segmentation structures. The deep learning model performed with an accuracy of 75.39%. To operate the nozzle, each image captured form the camera was separated lengthwise into quarters and mapped to the nozzles. Then, the nozzle was opened when the area of fruit trees in each zone exceeded 20%. A field test in a pear orchard was performed to verify the effectiveness of our system. From the obtained results, we could achieve a satisfactory performance of the deep learning-based intelligent spraying system. The introduction of this system to actual farms is expected to significantly reduce the amount of pesticide use and provide a safer working environment for users.