The variability in strata of foundation soil in marine environments makes it tedious to design foundations for offshore structures. Hence, it is essential to investigate and evaluate the strength properties of this type of soil. This study investigates the variability of the soil strata (which is quantified by the index of the mixing degree between coarse and fine particles) and its influence on the stability of the soil by mixing coarse and fine particles at varying proportions. A series of discrete element method triaxial shear tests were conducted on binary geotechnical mixtures with a varying proportion of coarse content (25%, 50% and 75%) and different mixing degrees (ranging from 0.0 to 1.0). The macroscopic results show that the peak shear strength increases with an increase in mixing degree, and the increase is more obvious with increasing coarse content, while the critical shear strength is independent of the mixing degree. The main evaluation of the number, mean normal force and distribution of the coarse–fine (cf) contact helps to clarify the meso-mechanisms that result in the variations in peak shear strength and critical shear strength with mixing degree. The increase in the peak strength may primarily be due to the increased number and globalized distribution of coarse–fine contact. However, the decreased contact force of coarse–fine contact counterbalances the strength gain due to the increased number and globalized distribution, which maintain the stability of the critical strength.

Granular materials are widespread in nature and human production, and their macro-mechanical behavior is significantly affected by granule movement. The development of computer vision has brought some new ideas for measuring the numerical information (including the amount of translation, the rotation angle, velocity, acceleration, etc.) of dynamic granular materials. In this paper, we propose a numerical measurement method for dynamic granular materials based on computer vision. Firstly, an improved video instance segmentation (VIS) network is introduced to perform end-to-end multi-task learning, and its temporal feature fusion module and tracking head with long-sequence external memory can improve the problems of poor video data quality and high similarity in appearance of granular materials, respectively. Secondly, the numerical information can be extracted through a series of post-processing steps. Finally, the effectiveness of the measurement method is verified by comparing the numerical measurement results with the real values. The experimental results indicate that our improved VIS obtains an average precision (AP) of 76.6, the relative errors and standard deviations are maintained at a low level, and this method can effectively be used to measure the numerical information of dynamic granular materials. This study provides an intelligent proposal for the task of measuring numerical information of dynamic granular materials, which is of great significance for studying the spatial distribution, motion mode and macro-mechanical behavior of granular materials.