Soil crack morphology analysis using image processing techniques

Abstract

The present paper demonstrated an image processing technique of surface soil crack analysis. The geometric features of cracks, such as width, length, and surface area are estimated. These parameters are important, because they influence both the soil hydraulics and mechanics. The crack intensity factor was introduced as a descriptor of the extent of surficial cracking. The correlation analysis indicates that area-weighted mean ratio of soil-crack area to perimeter and index r has a much closed positive relationship with cracks intensity and the area weighted mean of crack fractal dimension declines continuously as the degree of development of soil cracks increases, indicating that the degree of complexity of the soil cracks also gradually decreases. However, traditional visual assessment, which is the primary method in use, is slow and expensive. The present works involve image processing technique for the automatic detection and analysis of cracks in the digital image of a concrete surface.