A comprehensive framework for characterizing and quantifying soil desiccation crack patterns: Quantitative parameters and development of CIAS

• Published in: Engineering geology 2024-12, Vol.343, p.107820, Article 107820
• Main Authors: Tang, Chao-Sheng, Tian, Ben-Gang, Cheng, Qing, Leng, Ting, Shi, Bin
• Format: Article
• Language: English
• Subjects: Crack Image Analysis System-CIAS; Geometrical and morphological parameters; Image processing; Quantitative framework; Soil desiccation crack patterns; Statistical metrics
• ISSN: 0013-7952
• DOI: 10.1016/j.enggeo.2024.107820

Desiccation cracking has a significant impact on the hydro-mechanical properties of soils, yet quantifying crack patterns remains challenging. This study presents a quantitative framework with a total of 26 parameters for characterizing the geometric and morphological characteristics of soil desiccation crack patterns, including soil clod parameters (soil clod area, soil clod perimeter, number of clods, and the probability density distribution of clod parameters, etc.) and crack network parameters (crack length, crack width, crack intersection angle, number of crack segments, surface crack ratio, crack density, connectivity index, etc.). To implement this quantitative framework, the Crack Image Analysis System (CIAS) was developed to automatically identify and analyse complex crack patterns through image preprocessing, clod identification, crack network identification and batch processing. CIAS was then applied to quantify the crack images of soil with different thicknesses, validating its efficacy. To comprehensively describe the geometric and morphological characteristics of crack networks, it is recommended to use the number of soil clods per unit area, surface crack ratio, crack density, and connectivity index as key parameters. These metrics effectively capture information on crack spacing, area, length, width, and connectivity. This comprehensive framework for characterizing and quantifying crack images is of great significant for geological engineering. Moreover, it holds great potential for application in other different disciplines such as geotechnical, hydraulic, mineral engineering and material even planetary science. •A comprehensive framework for quantifying crack images is established.•26 parameters are defined for quantifying geometric and morphological features.•A powerful Crack Image Analysis System (CIAS) is developed and verified.•Four essential parameters for quantifying desiccation crack patterns are proposed.