Automated Visual Inspection of Road Surface Cracks

Abstract

Pavement maintenance requires knowing the state of the road surface. Human inspection is the most common method for evaluating this state. Recently, the automated visual inspection has been addressed, but some important questions remain open concerning the variable ambient lighting, shadows, device synchronisation and the large amount of data. In the present paper, an automated visual inspection system is presented. Images are obtained using laser lighting and linear cameras onboard a vehicle. Longitudinal and transversal cracks are detected and classified using a novel approach based on combining traditional features and Gabor filters. A Differential Global Positioning System (DGPS), a web camera and an Inertial Profiler to measure the International Roughness Index (IRI) are also considered in order to obtain comprehensive information about the road state. Implementation details are given concerning image acquisition and processing, system architecture and data synchronisation. Field results are presented which prove the suitability of the approach.

Pavement crack detection using the Gabor filter

“…The automatic filter selection method or automatic crack detection algorithm is not discussed for both kinds of cracks. It seems that the filters are selected manually for each type of crack after visual inspection [4].…”

“…In contrast to the previous studies, the algorithm presented in this study can detect every type of crack such as longitudinal, traverse, box or alligator crack despite of its orientation. The case study shown in the research is a part of alligator crack with crack spread in different directions and it is detected successfully [4].…”

Ongoing tests and improvements of the MPS algorithm for the automatic crack detection within grey level pavement images

“…The pavement images can provide information on the presence of cracks through the pixel intensities and the shape of the darker image features. Many image processing techniques exist for the detection of cracking on grey level images, e.g., [4][5][6][7][8][9][10][11][12][13][14].…”

Enhanced automatic detection of road surface cracks by combining 2D/3D image processing techniques

“…Image processing results obtained in previous works, [3,4], have been improved by incorporating geometric information obtained from the 3D profile of the road.…”

“…Genetic algorithms have been used for determining the threshold values in crack segmentation [7]; empirical mode decomposition (EMD) has been used to divide the complete spatial and frequency characteristic of the image features into different components [8]; filtered images have been used for crack segmentation [4,9]; and anisotropic methods for crack segmentation in different textured roads, such as Free-Form Anisotropy (FFA) [10], have also been developed, as well as filters based on the entropy of the patterns of intensities in image regions [11].…”