Robust plant disease segmentation in complex field environments: an in-depth analysis and validation with STAR-Net

TL;DR

This paper introduces STAR-Net, a new method for accurately identifying plant diseases in real-world agricultural settings.

Contribution

The paper introduces STAR-Net with a novel HBAA module and DPW-Loss for robust plant disease segmentation.

Findings

STAR-Net achieves a state-of-the-art mIoU of 93.36% on the NLB dataset for elongated disease segmentation.

The model obtains a 41.13% mIoU on the challenging PlantSeg dataset, showing robustness in complex environments.

Abstract

Plant disease segmentation in real-world agricultural environments poses significant technical challenges, including complex backgrounds, diverse lesion morphologies, and extreme class imbalance.

In this paper, we propose an integrated solution, STAR-Net, which combines a novel network architecture with a dynamic training strategy. The architecture features an innovative Heterogeneous Branch Attention Aggregation (HBAA) module to robustly represent multi-scale and multi-morphology features. The training strategy employs a Dynamic Phase-Weighted Loss (DPW-Loss) to navigate the complexities of imbalanced data.

Our method achieves a state-of-the-art average mIoU of 93.36% on the NLB dataset. This result demonstrates its superior ability to precisely segment diseases with specific elongated morphologies. Furthermore, the model obtains a competitive average mIoU of 41.13% on the highly challenging PlantSeg dataset. This result validates its robustness in complex 'in-the-wild' scenarios.

Our work presents a powerful, well validated, and synergistic solution for plant disease segmentation. It also paves the way for practical applications in precision agriculture.

Figures