Automatic Rooftop Extraction in Nadir Aerial Imagery of Suburban Regions Using Corners and Variational Level Set Evolution

Orientation-selective building detection in aerial images

“…Instead of applying a shape template or only providing a pixel-level location, the present approach estimates the real building outline, by applying the non-parametric ChanVese active contour algorithm from Chan and Vese (2001) similarly to Cote and Saeedi (2013). As the application of this iterative technique in not a novel approach for detecting building contours, we will not go into details.…”

“…However, in many cases, not all building objects could be represented by only two templates, moreover, the given features were not always enough to distinguish the buildings from the background. Cui et al (2008) and Cote and Saeedi (2013) introduced a method using Harris corner points; Sirmacek and Unsalan (2011) tested directional Gabor filter based feature points, Harris corner points and the FAST points of Rosten et al (2010) for extracting different local feature vectors (LFV), estimating a joint probability density function for urban area detection, assuming that around such points there is a high probability for urban characteristics. This technique motivated our previous work in Kovacs and Sziranyi (2013) for introducing the Modified Harris for Edges and Corners (MHEC) feature point set for efficient urban area detection.…”

Automatic Rooftop Extraction in Stereo Imagery Using Distance and Building Shape Regularized Level Set Evolution

“…And there are a number of approaches available regarding this topic (Weidner and Förstner, 1995;Sohn and Dowman, 2007;Cote and Saeedi, 2013;Wang et al, 2015;Liasis and Stavrous, 2016).…”

“…Level set has been also involved in building extraction using remote sensing data (Ahmadi et al, 2010;Kim and Shan, 2011;Li et al, 2014). Cote and Saeedi (2013) used corner points detected from multispectral aerial images as building corner candidates and further refined through level set curve evolution. Liasis and Stavrous (2016) proposed to use the building masks from morphological filtering as the initial building boundaries for level set evolution.…”

Building Detection From Monocular VHR Images by Integrated Urban Area Knowledge

“…Lately, various building detection approaches based on local features have been introduced: four different local feature descriptors were tested and two fusion methods were introduced in [1] to improve the results of detection; a graph-based strategy was developed in [2], utilizing corners and edges of buildings; a marked point process (MPP) framework was built in [3] and [4] using lower and higher level features to find the locations of buildings; an automatic approach was introduced in [5] combining the strength of energy-based approaches with the distinctiveness of corner features. In [6], it was shown that shadow information can be efficiently used for building localization, introducing a new fuzzy landscape strategy to uncover the relation between buildings and shadows and using iterative graph cuts for automated building detection.…”