Enhanced Detection of Soybean Leaf Diseases Using an Improved Yolov5 Model

Journal Title: International Journal of Knowledge and Innovation Studies - Year 2024, Vol 2, Issue 1

Abstract

To facilitate early intervention and control efforts, this study proposes a soybean leaf disease detection method based on an improved Yolov5 model. Initially, image preprocessing is applied to two datasets of diseased soybean leaf images. Subsequently, the original Yolov5s network model is modified by replacing the Spatial Pyramid Pooling (SPP) module with a simplified SimSPPF for more efficient and precise feature extraction. The backbone Convolutional Neural Network (CNN) is enhanced with the Bottleneck transformer (BotNet) self-attention mechanism to accelerate detection speed. The Complete Intersection over Union (CIoU) loss function is replaced by EIoU-Loss to increase the model's inference speed, and Enhanced Intersection over Union (EIoU)-Non-Maximum Suppression (NMS) is used instead of traditional NMS to optimize the handling of prediction boxes. Experimental results demonstrate that the modified Yolov5s model increases the mean Average Precision (mAP) value by 4.5% compared to the original Yolov5 network model for the detection and identification of soybean leaf diseases. Therefore, the proposed method effectively detects and identifies soybean leaf diseases and can be validated for practicality in actual production environments.

Authors and Affiliations

Shiqin Peng, Guiqing Xi, Yongshun Wei, Ling Yu

Keywords

Related Articles

Advanced Logarithmic Aggregation Operators for Enhanced Decision-Making in Uncertain Environments

This study introduces logarithmic operations tailored to intuitionistic fuzzy sets (IFSs) aimed at mitigating uncertainty in decision-making processes. Through logarithmic transformations, the membership and non-membersh...

An ISM-Based Framework for Open Innovation Enablers Driving the Adoption of Industry 4.0 in Small Manufacturing Firms in Caribbean Small Island Developing States

This study examines the role of Open Innovation (OI) in facilitating the adoption of Industry 4.0 (I4.0) technologies by small manufacturing enterprises in the non-energy sector of Caribbean Small Island Developing State...

Selection of CRM Systems Using Objective Criteria for the Needs of Small Companies

This research examines customer relationship management (CRM) systems using multi-criteria decision-making (MCDM) methods, with the intention of selecting the most suitable solution for small companies. The main goal of...

Integration of Fuzzy Inference Systems and Linear Regression for Enhanced Height Prediction of Deodar Cedar Trees in Kumrat Valley

Accurate estimation of tree height is fundamental to sustainable forest management, particularly in regions such as Kumrat Valley, Pakistan, where Deodar Cedar (Cedrus deodara) serves as a vital ecological and economic r...

Enhanced Fault Diagnosis in Motor Bearings: Leveraging Optimized Wavelet Transform and Non-Local Attention

Recent advancements in non-destructive testing methodologies have significantly propelled the efficiency of bearing defect detection, vital for maintaining optimal final quality standards. This study introduces a novel a...

Download PDF file
  • EP ID EP744651
  • DOI https://doi.org/10.56578/ijkis020105
  • Views 50
  • Downloads 0

How To Cite

Shiqin Peng, Guiqing Xi, Yongshun Wei, Ling Yu (2024). Enhanced Detection of Soybean Leaf Diseases Using an Improved Yolov5 Model. International Journal of Knowledge and Innovation Studies, 2(1), -. https://europub.co.uk/articles/-A-744651