Premium
Multiband Versus Multispectral Supervised Classification of Architectural Images
Author(s) -
Lerma José L.
Publication year - 2001
Publication title -
the photogrammetric record
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.638
H-Index - 51
eISSN - 1477-9730
pISSN - 0031-868X
DOI - 10.1111/0031-868x.00169
Subject(s) - multispectral image , pattern recognition (psychology) , artificial intelligence , euclidean distance , pairwise comparison , computer science , a priori and a posteriori , principal component analysis , identification (biology) , cover (algebra) , remote sensing , mathematics , geography , engineering , epistemology , mechanical engineering , philosophy , botany , biology
For the conservation of historic monuments, there may be considerable value in automating the methods of detection and analysis of surface condition and deterioration. This paper describes tests using a range of multiband and multispectral images for the assessment of architectural façade cover by means of supervised image classifications. From the spectral training sets, both pairwise distances (the Euclidean distance and the Jeffries‐Matusita (J‐M) distance) are calculated and are used to predict the a posteriori accuracy of image classification. Furthermore, the effects of increasing the number of spectral bands (blue, green, red and near‐infrared) in the supervised maximum‐likelihood classification procedures are also analysed, as are the benefits of applying principal components. The resultant multiband datasets increased both the J‐M distance and the classification accuracy of the architectural façade, and thus enabled better identification and recognition of the different kinds of façade‐cover features.