Appearance Quality Inspection And Assessment Of Concrete Structures Based On Image Analysis

Currently, manual inspection performed by a qualified inspector is the primary inspection method, the detection results are extremely dependent on individual experiences, lack of objectivity, and difficult to quantify deterioration process accurately. However, concrete engineering structures are widely used in our country, manual inspection of concrete surface is laborious and time-consuming. In addition, the influence of environmental is great in the process of detection and the experienced inspectors are quite deficient, and corresponding quantitative evaluation system has not been established yet. In view of the deficiencies of manual inspection, it is necessary to introduce an appearance quality evaluation system which is more objective and mainly based on quantitative analysis.Based on the application of image analysis technology, this paper presents a novel concrete surface defects detection and assessment approach that can detect concrete surface cracks, discoloration, air-pockets and holes. The program is written in Matlab and performed experimental study to verify the validity of the proposed method. Finally, concrete surface defects detection and assessment system has been established based on three-stage fuzzy comprehensive assessment method. The main investigations are summarized as follows:(1)For the detection of crack width, a new non-contact method was proposed in this paper,which using the calibration block on the concrete surface and revising the original deformed images,then enhancing the contrast of the revised image and smoothing them, etc. Finally, the width of crack in the image can be calculated. The method was verified by the laboratory experiment of a cracked concrete girder using vertical and oblique photographic techniques to detect the cracks and the accuracy can reach 93.4% and 90.9%, repectively, it shows that this new method can effectively detect the width of crack.(2)For discoloration, this paper presents the standard deviation of gray-scale image as evaluation index, and analyzes the influencing relationship of sunshine and shooting distance, and establishes assessment approach of discoloration based on trapezoidal fuzzy distribution. Experimental results show that the recognition results and manual visual assessment results agree well.(3)As to air-pockets and holes on concrete surface, this paper presents a novel method of combination of edge detection and morphological operation. The detection rate and accuracy rate are used to evaluate effectiveness of this method. 10 images are tested, and experimental results show that the average of detection rate is 92.1% with standard deviation of 5.1%, so this method can detect the number of air-pockets and holes completely and have much better stability. The average of accuracy rate is 96.4% with standard deviation of 3.4% only, it shows that the identified air-pockets or holes are almost always true, so this method has strong robustness.(4)Based on the image analysis results of quantitative detection of defects, using three fuzzy synthetic evaluation method, this paper established a quality evaluation system for concrete bridge appearance.