Ambiguous versus non ambiguous characterization of components from pictures of cataclasites

Abstract

Abstract:The quantitative analysis of shape parameters of components in cataclastic rocks is often complicated by the fact that the individual components are difficult to identify. Frequently, the components have the same mineralogical composition as the matrix and the size distribution of components and the resolution of the images make it difficult to discriminate between the individual fragments. This work investigates the differences of component shape analysis from processing pictures of cataclasites digitized by different persons. The approach undertaken herein assumes that the ambiguity of the forms is expressed in a diversity of component identification, respectively in different selection of the components. The extent of this diversity is quantified and analyzed as to how far statistical parameters reflect this ambiguity. Components in the samples may be more or less unambiguous with respect to the identification of particle shape parameters. The present study demonstrates that two parameters, the Aspect Ratio and the angularity (Elliptical Parisfactor) might be used in order to reflect this ambiguity in the component identification.

Effects of the introduction of pre-treated wastewater in a shallow lake reed stand

“…This is a widely known multivariate classification method in which each case starts in a separate cluster and joins up to the other clusters as the linkage distance grows, until only one cluster remains [41]. This method has been successfully applied in hydrology [42][43][44][45], hydrogeology [46], geology [47][48][49][50][51], chemistry [52] and anthropology [53] to find similar and homogeneous groups of observations. The validity of the groupings was verified using linear discriminant analysis (LDA), which separates the observations with linear planes resulting in a percentage of correctly classified cases [54,55].…”