Abstract
As the number of steel bridges which are important SOC of nation has been increased, the requirement of the efficient management technique of steel bridges is increased. The management of steel bridge is composed of structural side and anti corrosion side roughly. In this paper, we introduce an Automatic Coating Inspection and Management System (ACIMS) that diagnoses coating condition of steel bridge automatically using image processing and pattern recognition technique. The system assesses coating of steel bridges using color information and properties of deteriorations, and manages the data of inspection for maintenance. The proposed system is consisted of two modules: Server and client side. The client has the inspection module that detects rust area and peeled region in image by non-parametric modeling of the color distribution. The server manages the data of the inspection results and supplies useful information to user. By using the system, we can avoid errors which are produced by the lack of experience or knowledge of inspector, and reduce time and costs. As the ACIMS integrates the enormous information of steel bridges efficiently, it gives advantages in maintaining coating of steel bridges in both economical and technological side. Through the experiment, the ACIMS was tested on pictures which are obtained by local worker and accuracy which was calculated by comparing the inspection results with those which were obtained by experts is 92.55%.