Evaluation of image analysis methods used for quantification of particle angularity

Abstract

AbstractAngularity is an important parameter in the characterization of particle morphology that is used to interpret the transport history of particles in sedimentary deposits. In the past, visual classification using silhouette charts was widely used to determine particle angularity, but this approach is subjective and time‐consuming. With advances in modern image analysis techniques and low‐cost software packages, it is possible to rapidly quantify particle angularity more objectively than using visual classification methods. This study re‐examines the performance of three existing image analysis methods and one new image analysis procedure, applied to six rock and sediment samples that were visually different in angularity. To facilitate comparison between the angularity results, measurements were reduced to rankings for each aggregate sample. These results show that the four image analysis methods rank the angularity of the samples differently, and that none rank the mean angularity index in the same order as the angularity ranking using visual classification. Therefore, further research is needed to develop an image analysis method that can quantify the angularity of sedimentary particles more precisely.