Time-dependent crack analysis of lime and nano-treated expansive soil using image processing technique

Crack formation in soil samples due to environmental factors such as chemical exposure has critical implications for geotechnical engineering and environmental applications. Especially in weak soil, when stabilised with chemicals, crack formation is evident, as the failure is brittle. This paper presents a comprehensive investigation into the quantification and analysis of cracks formed in chemically modified soil samples subjected to six distinct combinations. The chemicals used in the study are lime at its initial lime consumption level and nano-alumina under two different water contents (liquid limit and optimum moisture content). The cracks are captured at an interval of 0, 3, 7, and 21 days. The primary objective is to understand and evaluate the effectiveness of crack formation and to explore various techniques for detecting cracks. Digital image processing techniques, such as edge detection and boundary analysis for characterising crack morphology, density, and length of the crack, are proposed. Experimental findings highlight significant variations in crack patterns across the tested environments. The integration of pre-processing techniques with Canny edge detection and contour-based analysis proved effective in automating image-based crack characterisation and enabling precise quantitative assessment. The study proposes an optimised parameter tuning framework for enhancing detection accuracy, thereby contributing to a deeper understanding of crack formation mechanisms under certain conditions.