InterImage is an open source knowledge based framework for automatic image interpretation.
With InterImage you can:
- embed your knowledge into the system to automatically interpret remote sensing images
- model your knowledge by creating a semantic network. The nodes of the network represent concepts, classes of image objects expected to be found in a scene
- associate specialized image processing operators (external from the system’s core) to each semantic node
- associate a decision rule to each operator. These rules model your knowledge in an explicit way and provide the system with the information necessary to perform an automatic classification.
InterIMAGE is a framework that may at any time be extended by the addition of new operators that perform specific tasks. The suite of operators that comes along with the system provides the user with some basic functionality such as extraction of objects in images and the calculation of spectral, geometric, topological and texture features.
The only disadvantage, similar to eCognition, the software exists only for windows operating systems. There is a limited, basic tutorial available on the website.
For further details check: http://www.lvc.ele.puc-rio.br/projects/interimage/