The Proceedings of Conference of Kanto Branch
Online ISSN : 2424-2691
ISSN-L : 2424-2691
2023.29
Session ID : 17G22
Conference information

Pattern recognition of stinkbugs by the robot using AI
*Yang YIWEIHiromu HIGUREGai KANEDAYuki ENOMOTOYuichi HASUDA
Author information
CONFERENCE PROCEEDINGS RESTRICTED ACCESS

Details
Abstract

Stink bugs are pests that damage crops such as beans, cruciferous vegetables, and eggplant. When fruit is sucked by stink bugs, the fruit may deform and fall. In addition, when it sucks on shoots, deformation of new leaves, stem bending, and perforation of leaves can occur. Damage to beans and fruit trees caused by stink bugs occurs every year. Farmers are therefore uncomfortable with stink bugs, and there is a need for early detection of damage caused by stink bugs and early countermeasures. In this study, a recognition model was generated using YOLOv5, a type of deep learning software, and successfully detected stink bugs individually by body part. Once the robot recognizes the target pest, it sprays chemicals and tries to prevent damage caused by stink bugs.

Content from these authors
© 2023 The Japan Society of Mechanical Engineers
Previous article Next article
feedback
Top