Lightweight weed detection using re-parameterized partial convolution and collection-distribution feature fusion DOI

Yan Kun-yu,

Wenbin Zheng, Yujie Yang

et al.

The Visual Computer, Journal Year: 2024, Volume and Issue: unknown

Published: Dec. 9, 2024

Language: Английский

YOLO-CWD: A novel model for crop and weed detection based on improved YOLOv8 DOI
Chong Ma,

Ge Chi,

Xueping Ju

et al.

Crop Protection, Journal Year: 2025, Volume and Issue: unknown, P. 107169 - 107169

Published: Feb. 1, 2025

Language: Английский

Citations

1
SFL-MobileNetV3: A lightweight network for weed recognition in tropical cassava fields in China DOI Creative Commons

Jiali Zi,

W. Hu, Guangpeng Fan

et al.

Expert Systems with Applications, Journal Year: 2025, Volume and Issue: unknown, P. 127196 - 127196

Published: March 1, 2025

Language: Английский

Citations

0
RLCFE-Net: A reparameterization large convolutional kernel feature extraction network for weed detection in multiple scenarios DOI

Ao Guo,

Zhenhong Jia,

Baoquan Ge

et al.

Expert Systems with Applications, Journal Year: 2025, Volume and Issue: unknown, P. 126941 - 126941

Published: Feb. 1, 2025

Language: Английский

Citations

0
Improved you only look once for weed detection in soybean field under complex background DOI

W. Zhang,

Xiaowei Shi, Minlan Jiang

et al.

Engineering Applications of Artificial Intelligence, Journal Year: 2025, Volume and Issue: 151, P. 110762 - 110762

Published: April 8, 2025

Language: Английский

Citations

0
MKD8: An Enhanced YOLOv8 Model for High-Precision Weed Detection DOI Creative Commons

Wenxuan Su,

Wenzhong Yang, Jiajia Wang

et al.

Agriculture, Journal Year: 2025, Volume and Issue: 15(8), P. 807 - 807

Published: April 8, 2025

Weeds are an inevitable element in agricultural production, and their significant negative impacts on crop growth make weed detection a crucial task precision agriculture. The diversity of species the substantial background noise images pose considerable challenges for detection. To address these challenges, constructing high-quality dataset designing effective artificial intelligence model essential solutions. We captured 2002 containing 10 types weeds from cotton corn fields, establishing CornCottonWeed dataset, which provides rich data support weed-detection tasks. Based this we developed MKD8 enhance model’s feature extraction capabilities, designed CVM CKN modules, effectively alleviate issues deep-feature information loss difficulty capturing fine-grained features, enabling to more accurately distinguish between different species. suppress interference noise, ASDW module, combines dynamic convolution attention mechanisms further improve ability differentiate detect weeds. Experimental results show that achieved mAP50 mAP[50:95] 88.6% 78.4%, respectively, representing improvements 9.9% 8.5% over baseline model. On public CottoWeedDet12, reached 95.3% 90.5%, 1.0% 1.4%

Language: Английский

Citations

0
Research on Weed Reverse Detection Methods Based on Improved You Only Look Once (YOLO) v8: Preliminary Results DOI Creative Commons
Hui Liu,

Yushuo Hou,

Jicheng Zhang

et al.

Agronomy, Journal Year: 2024, Volume and Issue: 14(8), P. 1667 - 1667

Published: July 29, 2024

The rapid and accurate detection of weeds is the prerequisite foundation for precision weeding, automation, intelligent field operations. Due to wide variety in their significant morphological differences, most existing methods can only recognize major crops weeds, with a pressing need enhance accuracy. This study introduces novel weed approach that integrates GFPN (Green Feature Pyramid Network), Slide Loss, multi-SEAM (Spatial Enhancement Attention Modules) accuracy improve efficiency. recognizes crop seedlings utilizing an improved YOLO v8 algorithm, followed by reverse through graphics processing technology. experimental results demonstrated model achieved remarkable performance, 92.9%, recall rate 87.0%, F1 score 90%. speed was approximately 22.47 ms per image. And when shooting from height ranging 80 cm 100 test, effect best. method addresses challenges posed diversity complexities image recognition modeling, thereby contributing enhancement automated weeding efficiency quality. It also provides valuable technical support farmland

Language: Английский

Citations

1
Lightweight weed detection using re-parameterized partial convolution and collection-distribution feature fusion DOI

Yan Kun-yu,

Wenbin Zheng, Yujie Yang

et al.

The Visual Computer, Journal Year: 2024, Volume and Issue: unknown

Published: Dec. 9, 2024

Language: Английский

Citations

0