文献类型：期刊文献

中文题名：基于二维复数离散小波包变换的桥面裂缝检测

英文题名：Detection for bridge surface crack based on 2D complex discrete wavelet packet transform

作者：马芸[1];王国军[1]

第一作者：马芸

机构：[1]新乡学院土木工程与建筑学院

第一机构：新乡学院土木工程与建筑学院

年份：2018

卷号：40

期号：6

起止页码：659-663

中文期刊名：沈阳工业大学学报

外文期刊名：Journal of Shenyang University of Technology

收录：CSTPCD;;Scopus;北大核心:【北大核心2017】；

基金：河南省教育厅科学技术研究重点资助项目(14A560011);新乡学院科技创新基金资助项目(15ZB11)

语种：中文

中文关键词：裂缝;特征提取;图像处理;复数离散小波变换;方向分量;分量提取;边缘信息;小波系数

外文关键词：crack;feature extraction;image processing;complex discrete wavelet transform;direction component;component extraction;edge information;wavelet coefficient

摘要：为了解决传统桥梁维护检测依靠检测人员进行现场测量,造成桥面检测效率低下且耗费巨大的问题,提出了一种基于二维复数离散小波包变换的桥面裂缝检测方法.采用无人机对桥面图像进行采集,并通过二维复数离散小波包变换来提取桥面裂痕特征.根据频域的角度范围设计了一维非分离滤波器计算小波变换需要的方向分量,并将该分量应用到通过2D-CWPT所获得的小波系数中,从而提取任意方向上的图像特征.采用提出的桥面裂缝检测方法对桥面裂缝进行检测,实验提取到了桥面裂纹的多方特征,验证了本方案的有效性.
In order to solve the problem that the traditional bridge maintenance inspection depends on the measurement of in-situ measurement by the inspector,which causes the low bridge surface detection efficiency and high cost,a detection method for bridge surface crack based on 2D complex discrete wavelet packet transform was proposed.The image of bridge surface was acquired with the unmanned aerial vehicle,and the characteristics of bridge surface cracks were extracted with the 2D complex discrete wavelet packet transform.According to the angular range of frequency domain,the direction components needed in the calculation of wavelet transform were designed with the one-dimensional non-separating filter,the components were applied to the wavelet coefficients obtained with the 2D-CWPT,and the image characteristics in arbitrarily direction were extracted.In addition,the proposed method was used to inspect the bridge surface cracks.The multiple characteristics of bridge surface cracks are extracted,and the effectiveness of the proposed method is verified.