Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Quantifying Nematodes through Images: Datasets, Models, and Baselines of Deep Learning

Zhipeng Yuan, Nasamu Musa, Katarzyna Dybal, Matthew Back, Daniel Leybourne, Po Yang

TL;DR

The paper tackles the problem of quantifying nematodes in images by surveying deep-learning approaches, datasets, and baselines for nematode detection. It consolidates public DL-based methods into direct and indirect detection, reviews available datasets, and analyzes one- and two-stage object detectors, training tricks, and evaluation metrics. A practical baseline is built by evaluating seven state-of-the-art models on four datasets, including a newly constructed AgriNema dataset, to facilitate reproducible benchmarking and guide future research. The work highlights key challenges—data scarcity, tiny-object detection, and field-deployable, low-cost solutions—while proposing directions like lightweight architectures, data augmentation, and zero-/few-shot learning to advance agricultural nematode monitoring.

Abstract

Every year, plant parasitic nematodes, one of the major groups of plant pathogens, cause a significant loss of crops worldwide. To mitigate crop yield losses caused by nematodes, an efficient nematode monitoring method is essential for plant and crop disease management. In other respects, efficient nematode detection contributes to medical research and drug discovery, as nematodes are model organisms. With the rapid development of computer technology, computer vision techniques provide a feasible solution for quantifying nematodes or nematode infections. In this paper, we survey and categorise the studies and available datasets on nematode detection through deep-learning models. To stimulate progress in related research, this survey presents the potential state-of-the-art object detection models, training techniques, optimisation techniques, and evaluation metrics for deep learning beginners. Moreover, seven state-of-the-art object detection models are validated on three public datasets and the AgriNema dataset for plant parasitic nematodes to construct a baseline for nematode detection.

Quantifying Nematodes through Images: Datasets, Models, and Baselines of Deep Learning

TL;DR

The paper tackles the problem of quantifying nematodes in images by surveying deep-learning approaches, datasets, and baselines for nematode detection. It consolidates public DL-based methods into direct and indirect detection, reviews available datasets, and analyzes one- and two-stage object detectors, training tricks, and evaluation metrics. A practical baseline is built by evaluating seven state-of-the-art models on four datasets, including a newly constructed AgriNema dataset, to facilitate reproducible benchmarking and guide future research. The work highlights key challenges—data scarcity, tiny-object detection, and field-deployable, low-cost solutions—while proposing directions like lightweight architectures, data augmentation, and zero-/few-shot learning to advance agricultural nematode monitoring.

Abstract

Every year, plant parasitic nematodes, one of the major groups of plant pathogens, cause a significant loss of crops worldwide. To mitigate crop yield losses caused by nematodes, an efficient nematode monitoring method is essential for plant and crop disease management. In other respects, efficient nematode detection contributes to medical research and drug discovery, as nematodes are model organisms. With the rapid development of computer technology, computer vision techniques provide a feasible solution for quantifying nematodes or nematode infections. In this paper, we survey and categorise the studies and available datasets on nematode detection through deep-learning models. To stimulate progress in related research, this survey presents the potential state-of-the-art object detection models, training techniques, optimisation techniques, and evaluation metrics for deep learning beginners. Moreover, seven state-of-the-art object detection models are validated on three public datasets and the AgriNema dataset for plant parasitic nematodes to construct a baseline for nematode detection.
Paper Structure (13 sections, 3 equations, 3 figures, 3 tables)

This paper contains 13 sections, 3 equations, 3 figures, 3 tables.

Table of Contents

  1. Introduction
  2. Methodology
  3. Deep Learning in Nematode Detection
  4. SQ1 - Deep Learning in Nematode Detection
  5. SQ2 - Public Datasets
  6. Object Detection Model
  7. SQ3 - Object Detection Models Based on Deep Learning
  8. SQ4 - Model Training and Optimisation Technologies
  9. SQ5 - Evaluation Metrics
  10. Experiments and baseline
  11. Experiment Setup
  12. Baseline for Nematode Detection
  13. Discussions and Conclusions

Figures (3)

  • Figure 1: Examples from public datasets.
  • Figure 2: Model Structures of Object Detection Model. CBS, ELAN, MP1, MP2, ELAN-H, and SPPCSPC are the composite model structure in YOLOv7 wang2022yolov7. CONV, NMS, Reshape, and FC are convolutional network modules with pooling modules, non-maximum suppression, reshape modules, and full connection modules, respectively.
  • Figure 3: Examples from AgriNema Dataset.