Nowadays, the significative improvements of remote sensing images have offered a better representation of intra-urban morphological patterns. However, high spatial resolution images have a great number of spectral noise, difficulting the discrimination of image objects. The analysis of image textural information is an alternative to minimize this problem. In this context, this paper verifies the capabilities of CBERS-2 imagens on the differentiation of morphological patterns between areas with different inhabitability conditions. Initially, the spectral information of a CBERS-2 image are enhanced and classified into two groups: occupied and non-occupied spaces. Then, the textural information of the binary classified image are calculated through a sliding box algorithm. Finally, the lacunarity values obtained are compared with values of a Inhabitability Index (IH) created from Census 2000 variables. The results show that, in general, non-occupied spaces are bettter distributed in urban areas with higher inhabitability conditions. 
For more information see the paper: A morfologia da habitabilidade intra-urbana: o uso de imagem CBERS-2 na análise de padrões morfológicos no Recife which was presented at XIII Brasilian Remote Sensing Symposium SBSR 2007 , in Florianópolis, Brazil.
