Abstract
Real-time, rapid, accurate, and non-destructive batch testing of fruit growth state is crucial for improving economic benefits. However, for plums, environmental variability, multi-scale, occlusion, overlapping of leaves or fruits pose significant challenges to accurate and complete labeling using mainstream algorithms like YOLOv5. In this study, we established the first artificial dataset of plums and used deep learning to improve target detection. Our improved YOLOv5 algorithm achieved more accurate and rapid batch identification of immature plums, resulting in improved quality and economic benefits. The YOLOv5-plum algorithm showed 91.65% recognition accuracy for immature plums after our algorithmic improvements. Currently, the YOLOv5-plum algorithm has demonstrated significant advantages in detecting unripe plums and can potentially be applied to other unripe fruits in the future.
Publisher
Public Library of Science (PLoS)
Reference40 articles.
1. Research on multi-class fruits recognition based on machine vision and SVM[J];H. Peng;IFAC-PapersOnLine,2018
2. Yang Jiangping. Research on Fruit and vegetable recognition Method based on Computer vision [D]. Dalian: Dalian University of Technology, 2011.
3. Imm aure green citrus detection based on colour feature and sum of absolute transformed difference(SATD) using cobur images in the citrusg rove[J;Chuan yuan Zhao;Computers and Electronics in Agriculture,2016
4. The use of hyperspectral imaging in the visible and near infrared region to discriminate between table grapes harvested at different times[J];Francesca Piazzolla;Journal of Agricultural Engineering,2013
5. Machine vision color inspection of potatoes and apples[J];Y Tao;Transactions of theASAE,1995
Cited by
3 articles.
订阅此论文施引文献
订阅此论文施引文献,注册后可以免费订阅5篇论文的施引文献,订阅后可以查看论文全部施引文献