TA NE 
A La WY 
SV 
Lerma, José Luis 
G, R), visible and near-infrared spectral classification with four spectral bands (B, G, R, near-IR) and finally, spectral 
(with the last four bands) and textural classification of multispectral close-range images; the three textural bands were 
computed from the grey level co-ocurrence matrix (GLCM) (Haralick et 31.,1992): homogeneity (3x3 window), mean 
(50x50 window) and homogeneity (50x50 window). 
2 STUDY SITE AND MATERIALS 
The study site is the Santos Juanes’ Church located on the old part of the city of Valencia. The materials consisted of 
limestone and brick rendered with masonry cement (inside the rosette). After a careful direct observation of the facade 
and depending on characteristics such as material, color, texture and relevance, several masonry wall materials and 
damages (features) were identified: (1) cleaned cement mortar used to render the rosette, (2) dirty cement mortar used 
to render the rosette, (3) flaking (thus, appearing at least the bricks), (4) polished limestone, (6) washed limestone, (7) 
dark limestone. 
3 IMAGE ACQUISITION AND PREPROCESSING 
The facade site was photographed with two cameras, a metric one (9 x 12 cm) used for the visible radiation and an 
amateur one (35 mm.) used for the reflected near-infrared radiation. For the latter purpose the following material was 
required: a black-and-white infrared-sensitive film, a Kodak Wratten gelatin filter No. 87 (Kodak, 1990).Each analogue 
photograph was digitized in order to get digital images; the quantization level of each digital image was of 8 bits/pixel. 
tified because they were taken with different cameras and they were also taken 
The spectral images were digitally rec 
from different locations. The four spectral images were merged altogether and resampled with a bilinear interpolation. 
A view of the extracted spectral set can be seen in Fig 1. 
  
(d) 
Figure 1. Spectral images: (a) blue, (b) green, (c) red and (d) near-infrared. 
International Archives of Photogrammetry and Remote Sensing. Vol. XXXIII, Part B5. Amsterdam 2000. 481