Author(s)

L.-M. Chang, P.-H. Chen and Y.AbdelRazig

Abstract

Intelligent diagnosis on bridge painting defects using image processing techniques L.-M. Chang', P.-H. Chen* and Y.AbdelRazig^ ^School of Civil Engineering, Purdue University, USA ^Department of Civil Engineering, Florida State University, USA Abstract Up to now, most bridge painting inspection still relies on human visual diagnosis, which is subjective, less efficient and inaccurate. In order to shorten the evaluation time and increase the accuracy, an intelligent diagnosis system was developed with the hope to standardize and automate the evaluation process. The system hybridizes image processing techniques and neural networks, which provide expert knowledge through training, to automatically diagnose the defects on an image. With the development of this system, vast number of images can be diagnosed instantly and intelligently. 1 Introduction One of the major causes for the deficiencies of U.S. bridges is the deterioration of anti-corrosion coatings (Hunt [6]). In spite o

Keywords