Analysis of Images from Experiments Investigating Fragmentation of Materials

Abstract

Image processing techniques have been used extensively to identify objects of interest in image data and extract representative characteristics for these objects. However, this can be a challenge due to the presence of noise in the images and the variation across images in a dataset. When the number of images to be analyzed is large, the algorithms used must also be relatively insensitive to the choice of parameters and lend themselves to partial or full automation. This not only avoids manual analysis which can be time consuming and error-prone, but also makes the analysis reproducible, thus enabling comparisons between images which have been processed in an identical manner. In this paper, we describe our approach to extracting features for objects of interest in experimental images. Focusing on the specific problem of fragmentation of materials, we show how we can extract statistics for the fragments and the gaps between them.