Detection of Tomato Spotted Wilt Virus Disease Piggybacked on SC-171 Edge Device Based on Improved YOLOv5s Modeling
Real-time status monitoring during crop growth is crucial to safeguard crop production,especially for tomato spotted wilt virus disease,which is prevalent in tomato crops.The aim of this study is to propose a detection method for tomato spotted wilt virus disease onboard SC-171 edge devices based on the improved YOLOv5s model.Under the same test conditions,our method improves the detection accuracy P,detection completeness R,F1 score,and average recognition accuracy mAP0.5 and mAP0.5:0.95 up to 1.40%,2.80%,1.86%,1.30%,and 3.40%,respectively,compared with the generalized YOLOv5 model.The algorithm maintains a high computing speed while improving the recognition accuracy,which satisfies the detection of tomato spotted wilt virus disease by edge detection equipment.
万方数据
