EVALUATION OF THE IMAGE TEXTURE PATTERNS AND VEGETATION INDEX
CLASSES INDICATORS FOR SOCIOECONOMIC POPULATION STANDARDS
Vänia Maria Vianna Ceccato
Celina Foresti
Maria de Lourdes Neves de Oliveira Kurkdjian
Instituto Nacional de Pesquisas Espaciais - INPE
Divisäo de Sensoriamento Remoto e Meteorologia - DSM
Caixa Postal 515
12201-970 - Säo José dos Campos, SP, Brazil
ISPRS Commission VII / Working Group 9
ABSTRACT
The use of aerial photography as a potential tool for socioeconomic characterization of urban population
has been successfully implemented since the fifties. Similarly, the use of aerial photographic and orbital
data processed with suitable image analysis algorithms (vegetation index) have produced satisfactory
information about urban green areas. Recent studies have indicated that vegetation indices can be used as
a qualitative measure of quality of life in urban areas. The objective of this work is to assess the
relationship between image texture patterns, vegetation index classes and income of resident population.
Five texture patterns were extracted from panchromatic SPOT paper prints at 1:25,000 scale based on a
method developed at INPE and described in this paper. Four classes of vegetation index were defined
using TM/Landsat images processed with the vegetation index algorithm, and aerial photographs. The
data about population income were obtained from sample observation. The test area was the city of Rio
Claro. Sáo Paulo, Brazil. An one-way analysis of variance indicated the existence of relationship between
image texture patterns, vegetation index classes and the income level of resident population. These
results support the decision of considering the information extracted from the images as indicators of the
quality of life in urban areas.
128