Quantifying texture scale

Introduction Quantitative descriptors that capture texture properties effectively have applications in biomedicine, geoscience and materials science. Ground-truth measurements, against which descriptors can be calibrated and their accuracy can be assessed, are required. These can be hard to obtain for properties such as order and scale for which accepted definitions do not exist. Protonotarios et al (2014) described an algorithm to quantify the order of point patterns. The algorithm was shown to be in accordance with an interval scale for human perception of order obtained through a pairwise comparison psychophysical experiment. We report progress on a similar approach for texture scale.