What question did this study set out to answer?

The aim is to develop a rapid identification system for soilborne fungi using automated imaging and machine learning techniques.

January 16, 2026Open Access

Machine Learning-Driven Soil Fungi Identification Using Automated Imaging Techniques

Key Points

The aim is to develop a rapid identification system for soilborne fungi using automated imaging and machine learning techniques.
Integrated robotic microscopy with deep learning algorithms.
Utilized the largest publicly available Soil Fungi Microscopic Images Dataset (SFMID) with over 20,000 images.
Evaluated four CNN architectures for accuracy and performance with transfer learning and majority voting.
Conducted statistical validation to compare classification methods.
Achieved optimal performance with ResNet152V2, showing precision of 0.6711 and AUC of 0.8031.
Validated real-time inference capability at 48–49 images per second.
Statistical analysis demonstrated significant improvement in classification accuracy with a three-shot method compared to single-image prediction.

Abstract

Soilborne fungi (Fusarium, Trichoderma, Verticillium, Purpureocillium) critically impact agricultural productivity, disease dynamics, and soil health, requiring rapid identification for precision agriculture. Current diagnostics require labor-intensive microscopy or expensive molecular assays (up to 10 days), while existing ML studies suffer from small datasets (<500 images), expert selection bias, and lack of public availability. A fully automated identification system integrating robotic microscopy (Keyence VHX-700) with deep learning was developed. The Soil Fungi Microscopic Images Dataset (SFMID) comprises 20,151 images (11,511 no-water, 8640 water-based)—the largest publicly available soil fungi dataset. Four CNN architectures (InceptionResNetV2, ResNet152V2, DenseNet121, DenseNet201) were evaluated with transfer learning and three-shot majority voting. Grad-CAM analysis validated biological relevance. ResNet152V2 conv2 achieved optimal SFMID-NW performance (precision: 0.6711; AUC: 0.8031), with real-time inference (20 ms, 48–49 images/second). Statistical validation (McNemar’s test: χ2=27.34,p<0.001) confirmed that three-shot classification significantly outperforms single-image prediction. Confusion analysis identified Fusarium–Trichoderma (no-water) and Fusarium–Verticillium (water-based) challenges, indicating morphological ambiguities. The publicly available SFMID provides a scalable foundation for AI-enhanced agricultural diagnostics.

Read Full Paperexternally

Ask AI

Helpful

Bookmark

View Full Paper