Knowledge-based image analysis for 3D edge extraction and road reconstruction

Abstract

Road network extraction from aerial images has received attention in photogrammetry and computer vision for decades.We present a concept for road network reconstruction from aerial images using knowledge-based image analysis. Incontrast to other approaches, the proposed approach uses multiple cues about the object existence, employs existingknowledge, rules and models, and treats each road subclass differently to increase success rate and reliability of theresults. Finding 3D edges on the road and especially the road borders is a crucial component of our procedure and is thefocus of this paper. We developed an algorithm for automatically matching line segments across images. The algorithmexploits line geometrical and photometrical attributes and line geometrical structures. A framework to integrate theseinformation sources using probability relaxation is developed and implemented to deliver locally consistent matches.Results of straight line matching are presented and future work is discussed.