YOLOv5 spore detection algorithm for wheat powdery mildew based on attentional feature fusion

Chuangang Chong; Botao Wang; Haozhe Yuan

doi:10.1117/12.2686372

10 August 2023 YOLOv5 spore detection algorithm for wheat powdery mildew based on attentional feature fusion

Chuangang Chong, Botao Wang, Haozhe Yuan

Proceedings Volume 12759, International Conference on Automation Control, Algorithm, and Intelligent Bionics (ACAIB 2023); 127591O (2023) https://doi.org/10.1117/12.2686372
Event: 2023 3rd International Conference on Automation Control, Algorithm and Intelligent Bionics (ACAIB 2023), 2023, Xiamen, China

Abstract

Wheat powdery mildew spore detection is one of the most common scenarios in the field of agricultural pest and disease detection. Since wheat powdery mildew spore images possess the characteristics of small spore targets, easy adhesion, shallow feature dimension, and many background clutter, it leads to the increased difficulty of wheat powdery mildew spore detection. By analyzing the above detection difficulties, a YOLOv5 wheat powdery mildew spore detection algorithm based on attentional feature fusion is proposed in this paper. Improvements are made on the basis of YOLOv5s. Firstly, an attentional mechanism-based feature fusion module AM-FF (Attentional Mechanisms-Feature Fusion) is proposed to strengthen the fusion of features at different scales, which enhances the learning ability of network contextual association by autonomously learning and selecting the optimal features through the attentional mechanism. Secondly, in order to enhance the shallow features and small target detection accuracy, we modify the original network structure to enhance the acquisition of shallow features and add fused reinforced shallow features in the neck network FPN (Feature Pyramid Networks). Finally, the SIoU (SCYLLA-IoU) loss function is introduced in order to increase the consideration of vector angle in the boundary box regression calculation, which effectively improves the detection accuracy. Experiments were conducted on the expert-labeled wheat powdery mildew spore dataset, and the final result of the algorithm experiment AP (Average Precision) was 94.23%, which was a large improvement compared with the original YOLOv5 detection algorithm vertically and also compared with the algorithm made by previous authors horizontally, and the experimental results proved the superiority of the proposed method in this paper.

Citation Download Citation

Chuangang Chong, Botao Wang, and Haozhe Yuan "YOLOv5 spore detection algorithm for wheat powdery mildew based on attentional feature fusion", Proc. SPIE 12759, International Conference on Automation Control, Algorithm, and Intelligent Bionics (ACAIB 2023), 127591O (10 August 2023); https://doi.org/10.1117/12.2686372

ACCESS THE FULL ARTICLE

INSTITUTIONAL
Select your institution to access the SPIE Digital Library.

SELECT YOUR INSTITUTION

PERSONAL
Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.

PERSONAL SIGN IN

No SPIE Account? Create one

PURCHASE THIS CONTENT

SUBSCRIBE TO DIGITAL LIBRARY

50 downloads per 1-year subscription

Members: $195

Non-members: $335 ADD TO CART

25 downloads per 1 - year subscription

Members: $145

Non-members: $250 ADD TO CART

PURCHASE SINGLE ARTICLE

Includes PDF, HTML & Video, when available

Members: $17.00

Non-members: $21.00 ADD TO CART

PROCEEDINGS
7 PAGES

DOWNLOAD PAPER SAVE TO MY LIBRARY

GET CITATION

RIGHTS & PERMISSIONS

Get copyright permission Get copyright permission on Copyright Marketplace

KEYWORDS

Feature fusion

Detection and tracking algorithms

Target detection

Machine learning

Diseases and disorders

Small targets

Feature extraction

Show All Keywords

Keywords/Phrases

Search In:

Publication Years