Advances in computer vision-based structural health monitoring techniques for wind turbine blades

With the rapid growth of wind energy, larger wind turbines with longer blades have been developed to maximize energy production. However, the increased blade size has also led to higher repair and maintenance costs, which demands efficient and reliable structural health monitoring techniques for wind turbine blades. Computer vision techniques have emerged as promising tools to address the key topics in blade structural health monitoring, including damage detection, damage localization, damage classification, and damage quantification. Despite recent advancements in both traditional computer vision methods and advanced approaches like deep learning, a comprehensive evaluation of the most effective computer vision techniques for structural health monitoring of wind turbine blades remains absent. This includes consideration of task-specific conditions such as environmental variability, computational requirements, and characteristics of different blade damages. This paper analyzes the strengths and limitations of state-of-the-art computer vision techniques for structural health monitoring of wind turbine blades, as well as the imaging modalities applicable to capturing different blade damages. This study also identifies the existing research gaps and future research for advancing computer vision-based structural health monitoring of wind turbine blades.