Structure extraction of straight wing aircraft using consistent line clustering

Abstract

In this paper, we propose a new method to automatically extract the main structure (fuselage and wings) of straight wing aircraft in photographic images using line clustering. We exploit the location, length and orientation features (parallel constraint) of line segments to group them into line clusters. The different line clusters correspond to the different structure of straight wing aircraft, and we can extract fuselage and wing of straight wing aircraft accordingly. We adopt line segment detector to extract line features, mean-shift and k-means++ approaches are applied to group line segments. The experimental results demonstrate that our method is effective for the structure extraction of straight wing aircraft.