American Society of Civil Engineers


Digital Image Processing Methods for Assessing Bridge Painting Rust Defects and Their Limitations


by Sangwook Lee, (Ph.D. Candidate, School of Civil Engineering, Purdue University, West Lafayette, Indiana 47907. E-mail: lee15@purdue.edu) and Luh-Maan Chang, (Associate Professor, School of Civil Engineering, Purdue University, West Lafayette, Indiana 47907. E-mail: changlm@purdue.edu)
Section: AI/Machine Learning, pp. 1-12, (doi:  http://dx.doi.org/10.1061/40794(179)80)

     Access full text
     Purchase Subscription
     Permissions for Reuse  

Document type: Conference Proceeding Paper
Part of: Computing in Civil Engineering (2005)
Abstract: Accurate and objective rust defect assessment is required to maintain a good-quality steel bridge painting surfaces and make a decision whether a bridge shall completely or partially be repainted. For more objective rust defect recognition, digital image recognition methods have been developed for the past few years and they are expected to replace or complement conventional painting inspection methods. Efficient image processing methods are also essential for the successful implementation of steel bridge coating warranty contracting where the owner, usually a state agency, and the contractor inspect steel bridge coating conditions regularly and decide whether additional maintenance actions are needed based on the processed data. Previously developed image recognition methods for painting rust defect assessment can be summarized as two: the NFRA (Neuro-Fuzzy Recognition Approach) method and the SKMA (Simplified K-Means Algorithm) method. The NFRA method uses artificial intelligence techniques to separate rust pixels from background pixels. The SKMA method segments object pixels and background pixels in a digitized image using a statistical method, called the K-means algorithm. Even if both methods pass through different processing procedures, one common thing is that they first convert original color images to grayscale images and further process the grayscale images. This article presents the application of previously developed image processing methods for painting rust defect evaluations and discusses their limitations under several specific environmental conditions which are often encountered while acquiring digital images.


ASCE Subject Headings:
Coating
Digital techniques
Imaging techniques
Steel bridges