"Multi-Ratio Sand-Cement Microscopic Image Dataset (MSCMID): A Benchmark for Automated Mix-Ratio Identification"

" Background: Accurate determination of cement-to-sand proportions is critical for ensuring the structural integrity and quality of construction mortars. Traditional methods for mix-ratio verification are often labor-intensive or require destructive testing. This dataset introduces a novel, micro-scale optical approach to material characterization, facilitating the development of automated, non-destructive quality control tools through computer vision.Methodology: The dataset consists of high-resolution microscopic images of dry sand and cement mixtures. To ensure a uniform, monolayer distribution of particles for clear imaging, a standardized \"Salo-tape\" adhesion method was employed. Various proportions (e.g., 1:1, 1:2, 1:3 etc.) were prepared by weight, spread over the adhesive substrate, and captured using an optical microscope. This preparation technique minimizes particle clustering, providing a distinct view of the morphological differences between the finer cement particles and the coarser sand grains."