Abstract: A system and method of mapping a persistent feature change of image data at a geographic location includes selecting a plurality of satellite images from a geographic location on different dates, producing a plurality of two-date satellite change images from pairs of the satellite images, comparing the plurality of satellite change images, and detecting a persistent feature change of image data in the compared satellite change images. The system may be implemented by selecting a plurality of satellite images from a database from different dates. A change detection module measures a change between pairs of the satellite images to produce a plurality of two-date satellite change images. A change confirmation module compares the two-date satellite change images and to confirm a persistent feature change.

Abstract: A system and method of mapping a persistent feature change of image data at a geographic location includes selecting a plurality of satellite images from a geographic location on different dates, producing a plurality of two-date satellite change images from pairs of the satellite images, comparing the plurality of satellite change images, and detecting a persistent feature change of image data in the compared satellite change images. The system may be implemented by selecting a plurality of satellite images from a database from different dates. A change detection module measures a change between pairs of the satellite images to produce a plurality of two-date satellite change images. A change confirmation module compares the two-date satellite change images and to confirm a persistent feature change.

Abstract: The present invention involves a system and method for processing information from high resolution data from maps, aerial photographs and other spatial data with low resolution data from limitation in order to accurately identify changes in the studied areas. The system and method relies upon a vectorized database which has been classified according to topographical features. Change is defined by a Z-statistic. The Z-statistic is calculated in a two-pass cross-correlation technique when pixel brightnesses are compared to a mean brightness for the typical topographical class and to a standard deviation for that class. The normalized difference values are accumulated across different bands, are scaled and then compared to a threshold which defines high, medium and low change values. A map is automatically produced illustrating areas of significant change.