Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2024 May 10:15:1360419.
doi: 10.3389/fpls.2024.1360419. eCollection 2024.

An improved algorithm based on YOLOv5 for detecting Ambrosia trifida in UAV images

Chen Xiaoming  1 , Chen Tianzeng  1 , Meng Haomin  1 , Zhang Ziqi  1 , Wang Dehua  1 , Sun Jianchao  1 , Wang Jun  1
Affiliations

An improved algorithm based on YOLOv5 for detecting Ambrosia trifida in UAV images

Chen Xiaoming et al. Front Plant Sci. .

Abstract

A YOLOv5-based YOLOv5-KE unmanned aerial vehicle (UAV) image detection algorithm is proposed to address the low detection accuracy caused by the small size, high density, and overlapping leaves of Ambrosia trifida targets in UAV images. The YOLOv5-KE algorithm builds upon the YOLOv5 algorithm by adding a micro-scale detection layer, adjusting the hierarchical detection settings based on k-Means for Anchor Box, improving the loss function of CIoU, reselecting and improving the detection box fusion algorithm. Comparative validation experiments of the YOLOv5-KE algorithm for Ambrosia trifida recognition were conducted using a self-built dataset. The experimental results show that the best detection accuracy of Ambrosia trifida in UAV images is 93.9%, which is 15.2% higher than the original YOLOv5. Furthermore, this algorithm also outperforms other existing object detection algorithms such as YOLOv7, DC-YOLOv8, YOLO-NAS, RT-DETR, Faster RCNN, SSD, and Retina Net. Therefore, YOLOv5-KE is a practical algorithm for detecting Ambrosia trifida under complex field conditions. This algorithm shows good potential in detecting weeds of small, high-density, and overlapping leafy targets in UAV images, it could provide technical reference for the detection of similar plants.

Keywords: YOLOv5; deep learning; invasive plant; small object detection; unmanned aerial vehicle.

PubMed Disclaimer

Conflict of interest statement

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Figures

Figure 1
Figure 1
Network architecture diagram of YOLOv5.
Figure 2
Figure 2
Schematic diagram of YOLOv5-KE application.
Figure 3
Figure 3
YOLOv5 original network structure.
Figure 4
Figure 4
YOLOv5 network structure with the addition of a microscale detection layer.
Figure 5
Figure 5
Schematic diagram of non-maximum suppression (NMS) and weighted box fusion (WBF).
Figure 6
Figure 6
WBF algorithm fusion process.
Figure 7
Figure 7
Images of Ambrosia trifida taken at data collection sites and by drones (A) Cropped image (B) Laplace transformed image (C-F) Artificially labeled image.
Figure 8
Figure 8
Data enhancement (A) original image, (B) rotated by 90°, (C) rotated by 180°, (D) flipped horizontally, (E) brightness-enhanced, (F) brightness-darkened. (G) pretzel noise, (H) Gaussian noise.
Figure 9
Figure 9
Resampling of Ambrosia trifida pictures.
Figure 10
Figure 10
Adding a comparison between before and after learning rate decay.
Figure 11
Figure 11
P-R curves of Ambrosia trifida images with different resolutions for YOLOv5-KE and standard YOLOv5 inputs.
Figure 12
Figure 12
YOLOv5-KE detection anchor box (blue). Ambrosia trifida labeling position (green) (A) 20x20 large-target detection layer, (B) 40x40 medium-target labeling detection layer, (C) 80x80 small-target detection layer, (D) 160x160 microscale detection layer.
Figure 13
Figure 13
Detection results using standard Anchor Box and k-means Anchor Box (A) Standard Anchor Box detection results; (B) k-means based Anchor Box hierarchical detection results.
See this image and copyright information in PMC

References

    1. Betti A. (2022). A lightweight and accurate YOLO-like network for small target detection in aerial imagery. Sensors 23, 1865. doi: 10.3390/s23041865 - DOI - PMC - PubMed
    1. Bhatt D., Patel C., Talsania H., Patel J., Vaghela R., Pandya S., et al. . (2021). CNN variants for computer vision: History, architecture, application, challenges and future scope. Electronics. 10, 2470. doi: 10.3390/electronics10202470 - DOI
    1. Bochkovskiy A., Wang C.-Y., Liao H.-Y. (2020). Yolov4: Optimal speed and accuracy of object detection. arXiv 2004, 10934. doi: 10.48550/arXiv.2004.10934 - DOI
    1. Ding K., Li X., Guo W., Wu L. (2022). “Improved object detection algorithm for drone-captured dataset based on yolov5,” in In Proceedings of the 2022 2nd International Conference on Consumer Electronics and Computer Engineering (ICCECE). 895–899.
    1. Dong S., Wang P., Abbas K. (2021). A survey on deep learning and its applications. Comput. Sci. Rev. 40, 100379. doi: 10.1016/j.cosrev.2021.100379 - DOI

LinkOut - more resources