30 December 1994 Manual versus automatic classification of Landsat TM data for lithological mapping: a case study in the Nama Group (South Africa)
Author Affiliations +
Abstract
Preliminary lithological investigation of a poorly known area can draw substantial help from remotely- sensed data, particularly in arid environments. A reconnaissance mapping of the Nama Group sedi- mentary succession in north-western South Africa (Richtersveld area) was performed with two strategies: visual interpretation and automatic classification. Both strategies share the same purpose: the identification and labelling/coding of the various surface types that are present in the image under study. Although often used alternatively, they can be successfully integrated. The first of such strategies is highly dependent on the interpreter's skill and intuition and therefore rather subjective; automatic classification, on the other hand, is rigidly parametric, although influenced by the creation of the spectral signatures for the training sets. The present paper offers some preliminary evaluations of the results deriving from the two approaches and contributes to the knowledge of an area that has not yet been mapped in detail.
© (1994) COPYRIGHT Society of Photo-Optical Instrumentation Engineers (SPIE). Downloading of the abstract is permitted for personal use only.
Vittoria Zichella, Enzo Pranzini, Arthur Robert Newton, "Manual versus automatic classification of Landsat TM data for lithological mapping: a case study in the Nama Group (South Africa)", Proc. SPIE 2320, Geology from Space, (30 December 1994); doi: 10.1117/12.197293; https://doi.org/10.1117/12.197293
PROCEEDINGS
10 PAGES


SHARE
Back to Top