一种新颖的路面裂纹修补识别方法研究

Research on A Novel Pavement Crack Repair Identification Method

针对目前公路路面裂缝的图像处理中难以区分未修补裂缝与已修补裂缝的难点问题,提出一种新颖的路面裂纹修补识别方法,通过Hessian矩阵特征向量与特征值中提取的各向异性对裂缝检测的敏感性,能较好地提取路面裂缝或修补裂纹细节信息,确定裂缝是否修补。通过实际拍摄的未修补裂缝、已修补裂纹的数据进行解析处理,结合边缘检测的能量强度与检测结果中的非平滑特征可以区分出裂缝是否被修补。

Focusing on the difficulty of distinguishing unrepaired cracks and repaired cracks in the current image processing of highway pavement cracks,this paper proposes a novel pavement crack repair identification method.It takes advantage of the sensitivity of anisotropy extracted from the eigenvector and eigenvalue of the Hessian matrix to crack detection,extracting the details of the pavement cracks or repaired cracks to determine whether the crack is repaired.According to the analytic processing of actual data of unrepaired and repaired cracks and the energy intensity of edge detection and non-smooth feature of the detection result,whether cracks are repaired or not can be distinguished.