PROJECT TITLE :

Road Extraction From Very High Resolution Remote Sensing Optical Images Based on Texture Analysis and Beamlet Transform

ABSTRACT:

Road extraction from terribly high resolution sensors may be a terribly fashionable topic in panchromatic and multispectral remote sensing image analysis. Despite the vast variety of methods proposed in the literature to deal with this problem, in observe, most are quite limited and don't account for geometric and radiometric variability. Our aim is to propose a novel road extraction approach in a position to efficiently extract roads and scale back computation time using texture analysis and multiscale reasoning primarily based on the beamlet rework. The proposed methodology consists of two stages: 1) road edge candidate selection and a pair of) multiscale reasoning with the beamlet transform. In the primary step, mathematical morphology is applied to tell apart rectilinear structures, and road edge candidates are identified using the Canny edge detector. In the second part, multiscale reasoning using the beamlet transform allows native and global information to be combined. World data is introduced to differentiate main road axes at coarser scales, and local segments in finer scales, which are aggregated to reconstruct the road network. Rules primarily based on the spatial relationships between segments belonging to totally different levels of resolution are also introduced at this stage. The experiments are performed based on the photographs acquired from the town of Port-au-Prince in Haiti during the earthquake of January 2010. The results demonstrate the accuracy and potency of our algorithm.