Affordable Access

Publisher Website

A multi-attribute method for comparing geological lineament interpretations

Remote Sensing of Environment
Publication Date
DOI: 10.1016/0034-4257(78)90044-5
  • Earth Science


Abstract Growing interest in lineament tectonics and regional fracture patterns, particularly since the advent of LANDSAT imagery, has led to the publication of a multiplicity of lineament interpretations. Such interpretations tend to be subjective and hence can be highly variable. Variations in lineament patterns can occur between interpretations of the same scene by one observer on multiple occasions as well as between interpretations of the same scene by several observers who may have used different methods. There is also the problem of assessing the similarity or differences between interpretations from two different scenes having similar geological patterns. In this paper we present a multi-attribute method for comparing two lineament patterns. With this technique the lineament patterns can be described in terms of location, direction, length and curvature classes. However, in the present paper only the first three attributes have been used and the information has been put into seven categories of vectors whose components consist of the length of lineaments in each class. These vectors range from “very fine” to “coarse”. In the fine vectors the classes have a hierarchical class structure consisting of other types of classes. In the coarsest vectors only one type of class is considered. By calculating the dot product of each vector for one interpretation with its equivalent vector for another interpretation, a range of similarity coefficients is obtained. These provide a local (sub-scene) comparison as well as an overall (scene) comparison of the two interpretations. The results of a series of experiments, covering a variety of applications, are given to illustrate the range of coefficients of similarity or association that can be expected.

There are no comments yet on this publication. Be the first to share your thoughts.


Seen <100 times