基于小波变换与形态学的机场跑道异物检测算法

机场跑道异物(foreign object debris,FOD)检测的精准性和效率直接决定了民航运输业的安全。为了提高机场跑道异物的定位精度,研究中提出基于小波变换与数学形态学相结合的机场跑道异物边缘检测质心定位算法。此算法先对采集到的机场跑道图像进行小波分解,高频部分利用小波变换的尺度边缘检测,并进行小波阈值去噪;低频部分利用数学形态学算子进行形态学边缘检测,然后对得到的高频和低频图像进行融合,并对异物边缘轮廓进行增强,最后利用质心定位法求解异物像素坐标。实验结果表明:小波变换与数学形态学相结合检测出的图像边缘具有较好的互补性,结合了小波变换在边缘精确定位和对噪声的抑制方面较好的性能,数学形态学在检测弱边缘和保留图像细节的优点,通过此算法提取的跑道异物边缘信息细腻且定位准确,能有效识别与机场跑道背景相似的异物并准确定位。

Detection Algorithm for Foreign Object Debris Based on Wavelet Transform and Morphology

The accuracy and high efficiency of the foreign object debris (FOD) detection on the airport runway directly determines the safety of the civil aviation industry. In order to improve the positioning accuracy of airport runway foreign objects, this paper proposes a centroid localization method based on the combination of wavelet transform with the mathematical morphology algorithm. The algorithm firstly performed wavelet decomposition on the acquired target image, and the high-frequency part used multi-scale edge detection of wavelet transform and the wavelet threshold denoising, the low-frequency part used the morphological algorithm to perform morphological edge detection. And then the high-frequency and low-frequency foreign object images were fused. Finally, the method of centroid location was applied to determine the pixel coordinates of the foreign object. The experimental results show that the edge of the runway foreign object extracted by this algorithm is delicate and accurate in positioning, and can effectively identify foreign objects with similar background to the airport runway and be accurately located.