Publications by authors named "Danlan Zhai"
Background: Machine vision-based precision weed management is a promising solution to substantially reduce herbicide input and weed control cost. The objective of this research was to compare two different deep learning-based approaches for detecting weeds in cabbage: (1) detecting weeds directly, and (2) detecting crops by generating the bounding boxes covering the crops and any green pixels outside the bounding boxes were deemed as weeds.
Results: The precision, recall, F1-score, mAP, mAP0 of You Only Look Once (YOLO) v5 for detecting cabbage were 0.
Article Synopsis
- This research focuses on improving weed detection for automatic herbicide application using a novel semi-supervised learning (SSL) approach, reducing the reliance on large labeled image datasets.
- SSL methods (like FixMatch) showed increased classification accuracy compared to fully supervised learning (FSL) models like ResNet50, with FixMatch being the most effective at achieving high accuracy with fewer labeled images.
- The findings indicate that SSL deep neural networks are not only accurate but also more time- and labor-efficient, making them a viable alternative to traditional FSL methods in weed detection.
The study is aimed to investigate the effects of light intensities on growth,photosynthetic physiology,antioxidant systems and chemical composition of Viola yedoensis and provide cultivation references for V.yedoensis.Five groups of V.
View Article and Find Full Text PDF