基于遗传神经网络的机场跑道局部影响的裂缝检测算法

Research on crack detection method of airport runway based on genetic neural network

摘要:

针对机场跑道裂缝的自主识别和提取过程中存在的阴影、光照不均匀以及效率和精度难以兼顾等一系列问题,提出利用遗传算法优化神经网络的机场道面裂缝检测算法。首先,将拍摄的机场道面裂缝图像进行预处理,包括图像灰度化、高斯滤波以及ROI区域确定。设定神经网络拓扑结构,初始化编码长度以权值阈值及等参数,利用选择、交叉和变异等操作反复执行至最佳进化解,进而搭建匹配的神经网络,获得最大分割阈值。结果表明,遗传神经网络算法在综合评价、召回率、和准确率3个评价指标上均具有显著提升,其均值分别为93.22%、96.28%、90.75%,实现了在复杂背景下对裂缝提取的目标,为机场道面的后期维护和保养提供了技术支持。

Abstract:

In order to solve the problems of complex cracks, high rate of missing cracks and low background contrast in the process of automatic crack identification and extraction of airport runway, a crack detection algorithm based on genetic algorithm and neural network was proposed. First of all, the airport pavement crack images were preprocessed, including image graying, Gaussian filtering and ROI region determination. By setting the network parameters of the genetic algorithm, the selection, crossover and mutation operations are repeatedly executed to the optimal progressive solution, and then the matching neural network is built to obtain the maximum segmentation threshold. The results show that the genetic neural network algorithm has a significant improvement in the comprehensive evaluation, recall rate, and accuracy of the three evaluation indexes, the average of which are 93.22%, 96.28%, 90.75%, respectively, to achieve the target of crack extraction under a complex background, and provide technical support for the later maintenance and maintenance of airport pavement.