Research On Surface Defect Detecting Techniques And System Integration For Underwater Structures

Aimed at the requirement of underwater structures detection of bridge structures, in this paper, the surface defects detection system for underwater structures is integrated based on the research of binocular stereo vision for the detection of the CCD camera calibration, the three-dimensional coordinates measurement, feature point extraction and matching, and the applicability of the system proposed is validated by experiment. The major contents consist of the following:Firstly, the mathematical model and structural configuration pattern of three- dimensional measurement for the binocular vision sensor are respectively analyzed and discussed. And then on the basis of the model and the structure configuration, the binocular vision sensor structure is designed. The real projection model of the camera is also studied. The camera parameters'calibration method suited for binocular vision inspection is obtained based on the two-dimension calibration. And the calibration process and the results are also provided.Secondly, based on the dimensional point calculation method for three-dimension coordinate, the structural surface defect detection and evaluation algorithm is proposed in paper. The feature selection method based on the automatic matching and the interactive identification is studied respectively. The SIFT- based method for feature point automatic matching is obtained, and the feature selection method for the interactive identification is also accomplished, the interactive identification tests for concrete block and steel block are correspondingly completed.Thirdly, the surface defect detection system for underwater structures is integrated in paper, which consists of two underwater CCD camera, image acquisition card, computer and surface defect detection software system. This system has the function of collecting image, storing image and evaluating damage. And then the performance evaluation experiments for detection systems are conducted on the case of different distances, different angles, and different water turbidities. Accordingly the performance parameters of detection system are obtained.Finally, with the surface defect detection system, the contrastive experiments are developed in clear water, turbid water and air. The flaws such as spalling, crack, and groove could be found, three-dimensional information about those defects is obtained. The experiment results prove the validity of surface defect detection algorithm, show the surface defect detection system could accurately detected those flaws, and validate the practicability of surface defect detection system.