Timely and accurate FOD (Foreign Object Debris) detection is essential to keep airport runway operating safely. However, manually inspection by workers is time consuming. An algorithm based on image processing theory is proposed to automatically detect FOD. According to the features of complex background and FOD, the images are cut into small pieces and several features such as Harris corner, GLCM, Gray-scale range, are computed and assembled according to the nature of airport pavement Threshold method and SVM are performed to check out images with FOD. The experiment results show that the proposed method can detect FOD from complex background with detection rate of 98%.%机场道面异物是威胁跑道运行安全的常见病害,及时、准确的检测异物具有现实意义.针对现有的人工目视检测方法,本文基于图像处理理论,提出了一种机场道面异物的自动检测算法.根据机场道面的复杂背景和常见异物的特点,本文采取了分块的方法,选择Harris角点、灰度共生矩阵、灰度级分布范围等特征,分别用阈值法和SVM法对实际机场道面异物图像进行检测.初步实验证明,该方法可以有效检测出机场道面复杂背景下的异物,实验结果显示,检测正确率达到了98％.
展开▼
