针对风力发电机叶片人工检测低效，缺陷诊断难的问题，提出一种基于无人机与图像处理的风力发电机叶片缺陷识别方法。通过Halcon 12与Visual Studio 2015的联合开发，实现图像处理流程、检测结果输出以及缺陷回放等功能，包括相机标定、通过快速自适应加权中值滤波处理图像、动态阈值分割叶片图像缺陷特征，利用区域处理识别裂纹和砂眼等缺陷，并对缺陷进行分类与测量以及输出对叶片质量的分析报告等，实现风力发电机叶片表面缺陷的自动检测功能。通过实例验证了该方法在风力发电机叶片表面缺陷检测中的较高精确性与算法稳定性。
School of Mechanical, Electronic and Control Engineering, Beijing Jiaotong University, Haidian District, Beijing 100044, China
Abstract:
Aiming at the problems of inefficient manual detection of wind turbine (WT) blades and difficult diagnosis of defects, a method of defects detection for WT blades based on unmanned aerial vehicle (UAV) and image processing is proposed. Through the joint development of Halcon 12 and Visual Studio 2015, the functions of image processing flow, test result output and defects playback are realized, including camera calibration, image processing through fast adaptive weighted median filtering and dynamic threshold segmentation. Through the regional processing to identify defects such as cracks and trachoma, the defects are classified and measured, and the analysis report of WT blades quality can be output to finally realize the automatic detection function of the surface defects of the WT blades. The high accuracy and stability of the algorithm of this method in detecting WT blades surface defects are demonstrated experimentally.
