Blade track tracking and state detection of wind turbine blades based on YOLOv8

With the wind turbine blade as the key-parts, Its operation state should be monitored at all times, In order to accelerate the innovation and upgrading of offshore wind farm inspection technology, The machine vision method is used to extract the operation data of the wind turbine blades, To realize the dynamic detection of the blade running status, The YOLOv8 algorithm is used to identify the coordinates of wind turbine blade tip, Tracking the position of the key points of the blade tip, To obtain the blade-tip trajectory data, Analysis of the blade status, Due to the lack of data set on part of wind turbine blades, Established with different power generation sizes, A dataset of ${2, 1 8 0}$ images including the three-dimensional blade model, Validation method on a low-power experimental fan, It can effectively detect the blade offset deformation angle parameters.