Abstract: The digital image information of soilrock aggregate mixtures of a target slope was collected by using digital image processing technique. Factors that might adversely affect the quality of digital image were filtered through the preprocessing of the digital photos. The dimensions, grain size distribution and other characteristics of the soilrock aggregate were obtained through gray scale, binary processing, and domain threshold adjustment.  The image pixel coordinates of the soilrock aggregates were converted the numerical identification that could be recognized by Particle Flow Code 2D (PFC2D) software and the numerical identification information in the software can be restored into digital image. The digital direct shear test results were also compared with the large scale direct shear test results on soilrock aggregate samples.

Key words: digital image processing, particle flow code 2d, rocksoil aggregate mixture, large direct shear test