Cognitive-mapping and contextual pyramid based Digital Elevation Model Registration and its effective storage using fractal based compression

Image Registration implies mapping images having varying orientation, multi-modal or multi-temporal images to map to one coordinate system. Digital Elevation models (DEM) are images having terrain information embedded into them. DEM-to-DEM registration incorporate registration of DEMs having different orientation, may have been mapped at different times, or may have been processed using different resolutions. Though very important only a handful of methods for DEM registration exist, most of which are for DEM-to-topographical map or DEM-toRemote Sensed Image registration. Using cognitive mapping concepts for DEM registration, has evolved from this basic idea of using the mapping between the space to objects and defining their relationships to form the basic landmarks that need to be marked, stored and manipulated in and about the environment or other candidate environments, namely, in our case, the DEMs. The progressive two-level encapsulation of methods of geo-spatial cognition includes landmark knowledge and layout knowledge and can be useful for DEM registration. Space-based approach, that emphasizes on explicit extent of the environment under consideration, and object-based approach, that emphasizes on the relationships between objects in the local environment being the two paradigms of cognitive mapping can be methodically integrated in this three-architecture for DEM registration. Initially, P-model based segmentation is performed followed by landmark formation for contextual mapping that uses contextual pyramid formation. Apart from landmarks being used for registration key-point finding, Euclidean distance based deformation calculation has been used for transformation and change detection. Initially, P-model based segmentation is performed followed by landmark formation for contextual mapping that uses contextual pyramid formation. Landmarks have been categorized to belong to either being flat-plain areas without much variation in the land heights; peaks that can be found when there is gradual increase in height as compared to the flat areas; valleys, marked with gradual decrease in the height seen in DEM; and finally, ripple areas with very shallow crests and nadirs. For the final storage of coregistered DEMs, fractal based compression has been found to give good results in terms of space and computation requirements. In this paper, an attempt has been made to implement DEMDEM registration based on human spatial cognition method of recollection. This method may further be extended for DEM-totopographic map and DEM-to-remote sensed image registration. Experimental results further cement the fact that DEM registration may be effectively done using the proposed method.