Search

As for the concrete cracks with complex noise such as honeycomb pockmarks on the surface, a high anti-noise extraction method of concrete cracks based on a dynamic threshold was proposed. Firstly, the noise was weakened and removed by gray-scale transformation, Gaussian filtering, ad information enhancement. The brightness of the crack in the image was enhanced, and the contrast between the crack and the background was improved. Then, the OTSU threshold segmentation algorithm was used to remove the pseudo cracks. Based on the difference between the area of cracks and that of interference noise, the dynamic threshold was introduced to realize the extraction of all kinds of cracks. Finally, based on edge detection and crack skeleton refinement, three characteristic paameters including the crack length, the width and the area were calculated. The results of four groups of crack images taken from the laboratory experiments and the actual bridge show that the average widths were 2.01, 1.07, 1.34 and 0.97 mm, respectively, proving that the proposed method is suitable for the crack extraction of various types under different backgrounds and has strong robustness. [ABSTRACT FROM AUTHOR]