Assessing the performance of YOLOv5 algorithm for detecting volunteer cotton plants in corn fields at three different growth stages

• DOI: 10.1016/j.aiia.2022.11.005
• Publication Date: 2022-12-01T19:59:28Z
• Authors: Pappu Kumar Yadav; J. Alex Thomasson; Stephen W. Searcy; Robert G. Hardin; Ulisses Braga-Neto; Sorin C. Popescu; Daniel E. Martin; Roberto Rodriguez; Karem Meza; Juan Enciso; Jorge Solórzano Diaz; Tianyi Wang
• Keywords: Anthonomus

The feral or volunteer cotton (VC) plants when reach the pinhead squaring phase (5–6 leaf stage) can act as hosts for boll weevil (Anthonomus grandis L.) pests. Texas Boll Weevil Eradication Program (TBWEP) employs people to locate and eliminate VC growing by side of roads fields with rotation crops but ones in middle remain undetected. In this paper, we demonstrate application computer vision (CV) algorithm based on You Only Look Once version 5 (YOLOv5) detecting corn at three different growth stages (V3, V6 VT) using unmanned aircraft systems (UAS) remote sensing imagery. All four variants YOLOv5 (s, m, l, x) were used their performances compared classification accuracy, mean average precision (mAP) F1-score. It was found that YOLOv5s could detect maximum accuracy 98% mAP 96.3% stage while YOLOv5m resulted lowest 85% YOLOv5l had least 86.5% VT images size 416 × pixels. developed CV has potential effectively well expedite management aspects TBWEP.