Abstract:

We proposed the idea of transforming the registration of vector spatial data and raster images to vector spatial data matching and carried out related research. The main idea of this paper: first,feature extraction based on shape feature of remote sensing images and vectorization, and then to match vectorization data and existing vector spatial data using vector spatial data matching method; then to choose some ground control points in the matching entities, and then get the corresponding pixel coordinates of the ground control points by the ground control points of the vectorization data is fed back to the original remote sensing image; final, to geometric correction of the original image using the ground control points coordinates of the existing vector spatial data and the corresponding pixel coordinates of the original image, in order to achieve the registration purpose of the vector spatial data and raster data.

Key words: shape feature, vector data, image data, feature matching, feature extraction, Beamlet transform, registration, Fourier shape descriptors