BladeYOLO: Wind Turbine Blade Defect Detection with Limited Annotations and Weak-Saliency Awareness

2026-07-30

AI Summary

This study addresses the challenges of wind turbine blade defect detection, including data scarcity, subtle defect appearance, low contrast, and complex background interference, which undermine the robustness of existing methods. To overcome these limitations, this work proposes BladeYOLO, a novel framework that integrates a DINOv3-pretrained Vision Transformer backbone with a YOLOv12-L detection head. The approach further introduces two key innovations: a Mamba-guided weak defect enhancement module and a lightweight Fourier decomposition–based style injection mechanism, collectively enhancing defect perception under low-label and cross-environment conditions. Experimental results on the WTBlade-Defect and Wind Surface Defect datasets demonstrate a 3.5% improvement in mAP50 and a 2.5% gain in mAP50–95, confirming the method’s superior performance and generalization capability in annotation-scarce scenarios.

Abstract

Wind turbine blade defect detection remains highly challenging in real-world inspection scenarios due to limited on-site data and the subtle visual characteristics of defects. In practice, blade defects are often small-scale, low-contrast, and difficult to distinguish from complex backgrounds, which significantly limits the robustness of existing detectors. To address these challenges, we propose BladeYOLO, a defect detection framework for wind turbine blades. Specifically, we integrate a Vision Transformer (ViT) backbone initialized with DINOv3 self-supervised pre-trained weights into YOLOv12-L, enabling the transfer of large-scale generic visual priors to blade defect detection and improving feature representation under limited training annotations. To enhance the perception of subtle defects, we further develop a Mamba-guided Weak-Defect Enhancement module, which consists of a Detail-Enhanced Multi-scale Branch for preserving high-frequency structural cues and a Cross-Mamba module for progressively propagating high-level semantic guidance to shallow features. In addition, we introduce a lightweight Style-Injector module that captures environment-related style information via Fourier decomposition and injects it into selected ViT self-attention layers, thereby improving robustness against environment-induced appearance variations. Extensive experiments demonstrate that BladeYOLO achieves superior performance on the WTBlade-Defect dataset, with additional annotation-budget experiments showing its favorable performance under reduced training annotations. Evaluation on the public Wind Surface Defect dataset further provides supportive evidence for the cross-dataset robustness of BladeYOLO. In particular, on this public dataset, BladeYOLO outperforms the best competing method by 3.5\% in mAP$_{50}$ and 2.5\% in mAP$_{50-95}$.

Problem

• wind turbine blade defect detection
• limited annotations
• weak-saliency defects
• low-contrast defects
• small-scale defects

Innovation

• Vision Transformer
• Mamba-guided Enhancement
• Weak-Saliency Awareness
• Self-Supervised Pretraining
• Style Injection