A Polygonal Approach for Automation in Extraction of Serial Modular Roofs
Yair Avrahami, Yuri Raizman, Yerach Doytsher · Photogrammetric Engineering & Remote Sensing · 2008
This paper presents a novel approach for automation in roof extraction from two solved aerial images. The approach assumes that roofs are composed of several spatial polygons, and that they can be obtained by extracting all or even only some of them if the model is known. In view of this assumption, innovative algorithms for semi-automatic spatial polygon extraction were developed. These algorithms are based on a 2D approach to solving the 3D reality. Based on these algorithms, an interactive and semi-automatic modelbased approach for automation in roof extraction was developed. The approach is composed of two phases: manual (interactive) and automatic. In the manual (interactive) phase, the operator needs to choose an Expanded Parameterized Model (EPM) from a knowledge base and select one pre-prepared Interactive Option for Extraction (IOE) of the roof. Then, the operator needs to point according to the guidelines of the chosen option in the left image space. In the automatic phase, the selected spatial polygons are extracted, the parameters of the selected model are calculated and the roof is reconstructed. The approach was examined and the results we obtained had standard accuracy. It appears that the approach can be implemented on many types of roofs and under diverse photographic conditions. In this paper, the algorithms, the experiments and the results are detailed.