TLDDM: An Enhanced Tea Leaf Pest and Disease Detection Model Based on YOLOv8 DOI Creative Commons
Jun Song, Y. Zhang, Shuo Lin

et al.

Agronomy, Journal Year: 2025, Volume and Issue: 15(3), P. 727 - 727

Published: March 18, 2025

The detection and identification of tea leaf diseases pests play a crucial role in determining the yield quality tea. However, high similarity between different difficulty balancing model accuracy complexity pose significant challenges during process. This study proposes an enhanced Tea Leaf Disease Detection Model (TLDDM), improved based on YOLOv8 to tackle challenges. Initially, C2f-Faster-EMA module is employed reduce number parameters while enhancing image feature extraction capabilities. Furthermore, Deformable Attention mechanism integrated improve model’s adaptability spatial transformations irregular data structures. Moreover, Slimneck structure incorporated scale. Finally, novel head structure, termed EfficientPHead, proposed maintain performance improving computational efficiency reducing which leads inference speed acceleration. Experimental results demonstrate that TLDDM achieves AP 98.0%, demonstrates enhancement compared SSD Faster R-CNN algorithm. not only great significance accuracy, but also can provide remarkable advantages real-time applications with FPS (frames per second) 98.2.

Language: Английский

Estimation of weed distribution for site-specific weed management—can Gaussian copula reduce the smoothing effect? DOI Creative Commons
Mona Schatke, Lena Ulber, Christoph Kämpfer

et al.

Precision Agriculture, Journal Year: 2025, Volume and Issue: 26(2)

Published: Feb. 28, 2025

Language: Английский

Citations

0
TLDDM: An Enhanced Tea Leaf Pest and Disease Detection Model Based on YOLOv8 DOI Creative Commons
Jun Song, Y. Zhang, Shuo Lin

et al.

Agronomy, Journal Year: 2025, Volume and Issue: 15(3), P. 727 - 727

Published: March 18, 2025

The detection and identification of tea leaf diseases pests play a crucial role in determining the yield quality tea. However, high similarity between different difficulty balancing model accuracy complexity pose significant challenges during process. This study proposes an enhanced Tea Leaf Disease Detection Model (TLDDM), improved based on YOLOv8 to tackle challenges. Initially, C2f-Faster-EMA module is employed reduce number parameters while enhancing image feature extraction capabilities. Furthermore, Deformable Attention mechanism integrated improve model’s adaptability spatial transformations irregular data structures. Moreover, Slimneck structure incorporated scale. Finally, novel head structure, termed EfficientPHead, proposed maintain performance improving computational efficiency reducing which leads inference speed acceleration. Experimental results demonstrate that TLDDM achieves AP 98.0%, demonstrates enhancement compared SSD Faster R-CNN algorithm. not only great significance accuracy, but also can provide remarkable advantages real-time applications with FPS (frames per second) 98.2.

Language: Английский

Citations

0