Detection and Measurement of Surface Cracks in Concrete Structures using Image Processing

 Richelle Gallardo-Zafra¹ and Yozo Fujino²

 

ABSTRACT

Cracks can be manifestations of structure distress hence it is essential to detect cracks early so that measures to prevent further deterioration can be taken. Considerable interest exists in developing an automated crack inspection system that will improve the efficiency of the data collection, consistency and uniformity of data and data quality over manual visual inspection. This paper presents a procedure developed for detecting and measuring surface cracks on concrete structures using image processing. It consists of image acquisition, edge detection using the Difference Vector edge operator, thresholding, filling, non-crack elimination, skeletonization, and crack measurement. Through tests performed, it was verified that cracks of different thickness can be detected and widths measured. Results showed that the accuracy of the measurement using this procedure is affected by the configuration and the length of crack section considered as well as the camera lens distance from concrete surface. The study shows the capability of the proposed procedure for crack detection and measurement.

 Keywords: Concrete structures, surface crack detection, crack measurement, image processing

 ¹Assistant Professor, Department of Civil Engineering, UPLB, College, Laguna, Philippines; ²Professor, Department of Civil Engineering, The University of Tokyo, Tokyo, Japan.