A procedure is developed to simultaneously extract textural features from all bands of ERTS multispectral scanner imagery for automatic analysis. Multi-images lead to excessively large grey tone N-tuple co-occurrence matrices; therefore, neighboring grey N-tuple differences are measured and an ellipsoidally symmetric functional form is assumed for the co-occurrence distribution of multiimage greytone N-tuple differences. On the basis of past data the ellipsoidally symmetric approximation is shown to be reasonable. Initial evaluation of the procedure is encouraging.
Spectral and textural processing of ERTS imagery
1974-01-01
Conference paper
No indication
English
Combined spectral and spatial processing of ERTS imagery data
Tema Archive | 1974
|Pseudocolor transformation of ERTS imagery
NTRS | 1973
|Crop identification using ERTS imagery
NTRS | 1973
|Geographic applications of ERTS-A imagery
NTRS | 1972
|