The problem may also be considered as a multi-sources information fusion process. For example the fusion of the numeric source (the image and derived information such as a classification for example) and a symbolic source (the expert knowledge, out-image data and derived information).

 Nowadays out-image data are often available in Geographic Information Systems (GIS) and some of these systems manage with imprecise and uncertain data.

When the out-image data is not available in a GIS, the corresponding information may be gained in the digital image by ad hoc techniques (for example detection of urban areas and thoroughfares …, detection of water areas and linear elements.

In case of very imprecise geographic databases, specific algorithms may be developed to improve the GIS accuracy especially for example in the case of updating vectored databases by means of satellite images.

 All methods combining classical pattern recognition techniques (image based) and artificial intelligence techniques (based on expert knowledge modeling and out-image data) have induced spin off concerning data management.

Such spin off concerns for example forestry forecasting by spatio-temporal techniques and problem-solving capabilities within geographical information systems.