Evaluating Histogram Matching for Robust Deep learning-Based Grapevine Disease Detection

TL;DR

This paper investigates Histogram Matching as a preprocessing and augmentation method to improve deep learning-based grapevine disease detection under variable lighting conditions, showing significant robustness gains especially in canopy images.

Contribution

It introduces a dual-stage Histogram Matching approach combining normalization and augmentation to enhance model robustness against lighting variability in plant disease detection.

Findings

Histogram Matching improves robustness in canopy images.

Dual-stage approach outperforms single techniques.

Significant gains in real-world, heterogeneous samples.

Abstract

Variability in illumination is a primary factor limiting deep learning robustness for field-based plant disease detection. This study evaluates Histogram Matching (HM), a technique that transforms the pixel intensity distribution of an image to match a reference profile, to mitigate this in grapevine classification, distinguishing among healthy leaves, downy mildew, and spider mite damage. We propose a dual-stage integration of HM: (i) as a preprocessing step for normalization, and (ii) as a data augmentation technique to introduce controlled training variability. Experiments using 1,469 RGB images (comprising homogeneous leaf-focused and heterogeneous canopy samples) to train ResNet-18 models demonstrate that this combination significantly enhances robustness on real-world canopy images. While leaf-focused samples showed marginal gains, the canopy subset improved markedly, indicating that balancing normalization with histogram-based diversification effectively bridges the domain gap caused by uncontrolled lighting.