FHBDSR-Net: automated measurement of diseased spikelet rate of Fusarium Head Blight on wheat spikes

TL;DR

This paper introduces FHBDSR-Net, a deep learning framework for automatically measuring the diseased spikelet rate in wheat affected by Fusarium Head Blight, improving efficiency and accuracy over manual methods.

Contribution

The novel FHBDSR-Net framework addresses data scarcity and spatial encoding limitations in detecting densely arranged diseased wheat spikelets.

Findings

FHBDSR-Net achieved 93.8% average precision in detecting diseased spikelets.

The method showed strong correlation with expert evaluations (Pearson r = 0.901).

The framework is lightweight (7.2M parameters) and suitable for mobile deployment.

Abstract

Fusarium Head Blight (FHB), a fungal wheat (Triticum aestivum) disease that threatens global food security, requires precise quantification of diseased spikelet rate (DSR) as a phenotypic indicator for resistance breeding. Most techniques for measuring DSR rely on manual spikelet-by-spikelet observation and counting, which is inefficient and destructive. Although deep learning offers great promise for automated DSR measurement, existing intelligent detection algorithms are hampered by the lack of spikelet-level annotated data, insufficient feature representation for diseased spikelets, and weak spatial encoding of densely arranged spikelets. To address these challenges, we constructed a dataset of 620 high-resolution RGB images of wheat spikes with 5,222 spikelet-level annotations to systematically analyze spikelet size distributions to fill small-object detection data gaps in this field. We designed FHBDSR-Net, a light framework for automated DSR measurement centered on diseased spikelet detection, which features (1) multi-scale feature enhancement architecture that dynamically combines lesion textures, morphological features, and lesion-awn contrast through adaptive multi-scale kernels to suppress background noise; (2) the Inner-EfficiCIoU loss function to reduce small-target localization errors in dense contexts; and (3) a scale-aware attention module using dilated convolutions and self-attention to encode multi-scale pathological patterns and spatial distributions to enhance dense spikelet resolution. FHBDSR-Net detected diseased spikelets with an average precision of 93.8% with a lightweight design of 7.2 M parameters. The results were strongly correlated with expert evaluations, with a Pearson correlation coefficient of 0.901. Our method is suitable for deployment on resource-constrained mobile devices, facilitating portable plant phenotyping and smart breeding.

Figures