Performance of Quickbird image and LiDAR data fusion for 2d/3d city mapping

Abstract

In this paper we present a practical and convenient 2D city mapping and 3D digital surface model construction technique based on data merge of spatial, spectral, and textural information of QuickBird High Resolution Satellite Imagery with precise Digital Surface Model (DSM) information of LiDAR data for an urban area within the city of Cairo. Due to recent increased demand for city mapping from various authorities at scale of 1:5000 for the purpose of up-grading and improving city conditions, with reasonable cost and limited time frame, the merge between QuickBird and LiDAR is investigated and tested. The result of this merge is very encouraging and proved its suitability for 2D and 3D city mapping and visualization in a cost-effective way. The integration of spectrally and spatially complementary remote multi sensor data can facilitate and improve visual and image interpretation. Compared with an existing 1:5000 planimetric reference map, a mean value exceeding 93% was achieved for manual building and road network extraction from QuickBird and LiDAR fused image. Also, building height comparison between LiDAR enhanced data and ground truth verification resulted in a mean absolute height difference of 47cm with a standard deviation of 58cm. Besides, the advantages of using QuickBird/LiDAR merged data was evident in the areas of cloud coverage, improved built-up area classification, better interpretability, as well as 3D/Digital City Model (DCM) visualization and mapping.