Where mixtures are prepared, placed or compacted, the ingredients likely segregate. Part of the reasons is ingredient differences at least in density, size, shape and surface roughness and, when placing the mixture, leads to flow-induced segregation. Considering the large quantity and small scale of ingredients, it is challenging to examine the segregation through hands-on tests. Alternatively, digital image processing (DIP), an image-based computer simulation approach, provides the toolbox to tackle this task. DIP was applied to observe segregation occurred to rubber sand mixtures when they form heaps. Through segmenting the digital images into a binary picture, the DIP method enables material ingredients identification and three-dimensional mapping of mixture segregation. This helps reach a better understanding of mixture heterogeneity when incorporating artificial material into conventional geotechnical materials. To gain an insight into the mixture heterogeneity, the DIP results were used to validate a discrete element model and the model was then used to examine the influence of particle properties on the segregation. Some interesting validating tests and example problems are provided to further understand the mixture segregation problems.