The sweet pepper cultivation and plant protection robot developed includes three parts: the intelligent image identification and positioning system of sweet pepper diseases and insect pests is composed of YOLO4 deep learning model and image 3D vision integration. It can perform disease detection and spatial positioning, and guide the robotic arm spraying device to accurately spray the diseased. The automatic walking system can use GPS to plan the movement path of the plant protection robot in the field, search for plants infected with pests and diseases, and perform the tasks of pest control.
This technology establishes a database for the identification of pests and diseases of sweet pepper with planar images and stereo vision. In addition to the digital image data of the main hazards of the pests, it further establishes the early to late grade data of the hazards of the pests. The research results can be used for similar crops in the future. It is not only limited to the identification of pests and diseases of sweet peppers.
This project developed a multi-kernel sweet pepper disease and pest image identification system, using deep learning such as Faster R-CNN and YOLOv4 to identify plant pathology with different networks. The image database of sweet pepper pest symptoms contains the clear and severe symptoms; further establish the symptom severity from early to severe. The established database can be applied to the developed technology, but also to the general identification of sweet pepper pest, and assist in pest diagnosis on plant medicine.

線上展網址：
https://tievirtual.twtm.com.tw/iframe/b3718d3b-d86c-4f77-adee-e58a50754c96?group=23bfb1fa-dd5b-4836-81a1-4a1809b1bae5&lang=en