A GIS vectorization model for quad-tree satellite images

Quad tree is a common raster data structure that uses recursive decomposition to divide a raster into a hierarchy of quadrants. This papers aims to develop a vectorization model for the conversion of quad tree satellite images to vector format that can be used and processed in geographic information systems (GIS) projects. The proposed model consists of four main phases: applying noise removal filter, segmenting images using edge detection methods, recognition of segments boundaries using a contour detection algorithm, and approximation of the sets of connected points by generalization and simplification procedures to create Shapefiles vector map data. The proposed vectorization model is implemented using Pascal. Results showed that the model is acceptable for mapping purposes. Future work may include developing a vector linker component, capable of linking different vectors at different blocks at the same layer.