Abstract: The visual inspection based on an endoscope has played an important role in modern fault diagnosis of aero-engines. An automatic measurement approach to cracks of the aero-engine inner parts is studied in this paper. In this method, the area of the crack needs to be selected, it is subjected to image re-processing, and the pseudo cracks are removed, the crack is measured and calculated. Automatic measurement system has developed based on this method. The automatic measurement system includes crack segmentation, crack identification, and measurement, and it can solve the disadvantages of traditional inspection method, such as, excessive reliance on the inspector’s experience, low inspection efficiency, and poor accuracy of measurement. Several experiments are used to verify the effectiveness and practicality of the developed system.

Key words: aero-engine, image processing, crack measurement

摘要： 基于内窥镜的视觉检测技术在现代航空发动机故障诊断中发挥着重要的作用。针对航空发动机内部部件损伤产生的裂纹，提出了一种裂纹自动测量方法。该方法需要选取裂纹所在区域，对其进行图像预处理，去除图像中的伪裂纹，进行裂纹测量计算。基于该方法，开发了集裂纹分割、裂纹识别、裂纹测量为一体的航空发动机内部损伤裂纹的自动测量系统，解决了孔探裂纹检查中严重依赖工作人员经验、测量结果准确性较差、效率较低等问题。通过多组发动机孔探图像的裂纹测量实验证明了所开发系统的正确有效性和普遍实用性。

关键词: 航空发动机, 图像处理, 裂纹测量