Lightweight Transformer-Driven Segmentation of Hotspots and Snail Trails in Solar PV Thermal Imagery

TL;DR

This paper introduces a lightweight SegFormer-based deep learning model for real-time segmentation of hotspots and snail trails in thermal images of solar panels, improving accuracy and efficiency for drone-based inspections.

Contribution

The work develops a novel lightweight segmentation model tailored for thermal PV imagery, outperforming existing models in accuracy and enabling real-time drone deployment.

Findings

SegFormer model achieves higher accuracy than U-Net, DeepLabV3, PSPNet, and Mask2Former.

The model effectively segments small and irregular defects in thermal images.

Real-time deployment on edge devices is feasible with the proposed lightweight architecture.

Abstract

Accurate detection of defects such as hotspots and snail trails in photovoltaic modules is essential for maintaining energy efficiency and system reliablility. This work presents a supervised deep learning framework for segmenting thermal infrared images of PV panels, using a dataset of 277 aerial thermographic images captured by zenmuse XT infrared camera mounted on a DJI Matrice 100 drone. The preprocessing pipeline includes image resizing, CLAHE based contrast enhancement, denoising, and normalisation. A lightweight semantic segmentation model based on SegFormer is developed, featuring a customised Transformwer encoder and streamlined decoder, and fine-tuned on annotated images with manually labeled defect regions. To evaluate performance, we benchmark our model against U-Net, DeepLabV3, PSPNet, and Mask2Former using consistent preprocessing and augmentation. Evaluation metrices includes per-class Dice score, F1-score, Cohen's kappa, mean IoU, and pixel accuracy. The SegFormer-based model outperforms baselines in accuracy and efficiency, particularly for segmenting small and irregular defects. Its lightweight design real-time deployment on edge devices and seamless integration with drone-based systems for automated inspection of large-scale solar farms.