This paper reviews the state-of-the-art of three key problems related to the residential area extraction:the quantitative description of residential areas in remotely sensed image,extraction models and accuracy assessment methods.

An approach for extracting outline of residential area through perceptual organization and multi-scale boundaries of textured regions is presented.

Choosing TM images in 1986 and 2002,through multiband color composition,the spectrum models for extracted information of residential areas are established,and the distributing maps of residential areas in 1986 and 2002 are made.

For automatic recognition of residential areas filled with hatched lines in color scanned topographical maps,Gabor filters are firstly used to detect hatched lines with their texture features and find the candidate residential areas.

A new method is proposed for automated recognzing the change types of residential areas using topological relations between new and old boundaries of residential areas.

Spatial pattern in residential area and influencing factors in oasis landscape;

Firstly,the characteristics of residential areas on the image are studied.

A residential area texture description based on the 3×3 region grey deviations is designed, and the Gauss blur is applied to make the residential area in the texture feature image possessing accordant grey value and limited contrast relative to the background area so as to obtain self-adaptive threshold for image segmentation.

Spatial characteristics of farmland and residential land in dry valley of Minjiang River upper reaches;