Image analysis for the detection and quantification of concrete bugholes in a tunnel lining

Abstract

A measurement and quantification system for concrete bugholes (surface air voids) on sidewalls was developed to quantify the surface quality of tunnel-lining concrete. The developed system uses and evaluates red/green/blue values of color images taken by a commercial digital still camera. A comparative test shows that the developed system has higher accuracy than image analyses using thresholding and can estimate bugholes with accuracy almost equal to that of a detailed visual inspection. The results confirm that even small bugholes (<1 mm) can be detected in color image analysis, whereas such bugholes are hardly detected in the detailed visual survey. In addition, color image analysis improves the calculations of the area of multiple bugholes distributed randomly over a concrete surface. Fundamental tests employing image analysis demonstrate that the prevalence of bugholes increases with an increase in the negative angle of the concrete form and a decrease in concrete workability. The system is applicable to the quantitative evaluation of a concrete surface having visible and invisible bugholes. Results indicate that the developed color image analysis can contribute to the reasonable and appropriate evaluation of bugholes and replace a detailed survey that requires much human resource and has a long inspection time.