Abstract: As a special granular material, the mechanical properties and characteristics of sand are significantly affected by particle shape such as packing density, shearing characterization, compaction, and crushability. However, grain shape quantification techniques to distinguish one sandy material from another have not reached a stage of development for inclusion in the interpretation of the behavior of sand. In this paper, the light microscope was employed to capture the grain shape digital images, which were subsequently converted into binary picture to produce contour boundary for particle shape with the help of ImageJ. The Java program language was used to compile a new plugin to determine several grain shape parameters of all sand samples in our test. Finally, relative density and direct shear tests were conducted to determine the limit void ratios and shearing strength index. It is concluded that the overall outline coefficient, sphericity, angularity could be taken as the key shape parameters to distinguish grains; and these key parameters are related closely with the angle of dilation and critical state angle as well. The proposed method can be used as an effective means to accomplish the sand particle shape quantification; and the key shape parameters can also be applied to the mechanical properties analysis and the related numerical simulation work

Key words: sand, particle shape, digital image processing, shape parameters, dilation angle