摘要
在水泥刻槽路面损坏检测中,由于刻槽的存在给基于图像的裂缝识别造成许多噪声干扰,长期以来无法对水泥刻槽路面应用快速检测设备进行自动裂缝识别。为了解决这一问题,对水泥刻槽路面裂缝识别方法进行了研究,提出根据几何特征识别刻槽、去除刻槽的方法。首先,用形态学的方法去除路面噪声,然后通过试验得到适用于水泥路面裂缝的最佳边缘检测方法及最佳阈值,检测出图像中的裂缝和刻槽。根据刻槽的直线及角度特征,利用Hough变换对边缘检测结果进行刻槽提取,而后将其从边缘检测结果中剔除,得到裂缝。结果表明,用该方法可以有效地去除刻槽误识别,准确识别裂缝,与传统的边缘检测方法相比,识别精度有了明显提高,解决了长期以来水泥刻槽路面无法进行自动识别的问题,提高了水泥刻槽路面破损检测效率。
During the crack detection of grooved cement concrete pavement,because of the existence of groove,lots of noises was introduced to image-based crack detection.Therefore,it is very difficult to apply the rapid road detection equipment to detect cracks in grooved cement pavement automatically.To solve this issue,the method for detecting cracks in grooved cement concrete pavement was analyzed,then the way of detecting and removing groove was proposed according to its geometrical characteristics.First,morphology was applied to remove the noises on the pavement image.Second,the most appropriate edge detecting method and threshold suitable for cracks in cement concrete pavement were obtained through an experiment to detect the crack and groove on images.Based on its linear and angular features,groove will be extracted from the edge detection result by Hough transformation,then be removed from the edge detection result,finally cracks could be obtained.The results illustrate that comparing with traditional edge detecting algorithm,this algorithm can avoid groove mis-detection effectively,and accuracy of crack detection could be improved significantly.Consequently,the problem of suffering from crack detection on grooved cement concrete pavements for a long time was solved,and efficiency in crack detecting on grooved cement concrete pavements was increased.
出处
《公路交通科技》
CAS
CSCD
北大核心
2012年第3期45-50,共6页
Journal of Highway and Transportation Research and Development
基金
国家高技术研究发展计划(八六三计划)项目(2006AA11Z101)
关键词
道路工程
水泥裂缝识别
图像识别
水泥刻槽
road engineering
cement crack detection
image identification
cement groove
作者简介
荣婧（1980-），女，山西蒲县人，博士研究生．（j．rong@fioh．cn）
