A forest map updating expert system based on the integration of low level image analysis and photointerpretation techniques

In remote sensing a large amount of data is now available, such as very high spatial resolution images and imaging spectrometers, artificial intelligence can be applied to the management of this new information. An expert system (ES) can easily integrate and exploit various type of information, such as soil type map, DEM (digital elevation model) or images. We describe a new method combining an ES and a low level tool. The reasoning of the ES is applied to pixels poorly classified or unclassified by the conventional methods.