Name
Title | ||
|---|---|---|
Mapping Vegetation-Covered Urban Surfaces Using Seeded Region Growing in Visible-NIR Air Photos |
Abstract | ||
|---|---|---|
Unreliability involved in the extraction of shaded vegetation-covered surfaces (VS) is a common problem in urban vegetation mapping. Serving as a solution to it, a novel method named Nonlinear Fitting-based Seeded Region Growing (NFSRG) is explored. With NFSRG, a series of classified results are organized by a seeded-region-growing process. In order to adapt to the variable separability between VS and background, the growing is limited in several weighted buffers defined by some nonlinear fitting relationships. When searching new VS members (member means both pixel and patch) within such a buffer, a gradually reduced weight makes the buffer width continually narrowed as the separability worsens. To avoid unexpected entrances of water and smooth shaded background members, a during-growing constraint, named expansion rate, is proposed. Accuracy assessments reveal that more than 96% of VS members can be accurately extracted by the proposed method. |
Year | DOI | Venue |
|---|---|---|
2015 | 10.1109/JSTARS.2014.2362308 | Selected Topics in Applied Earth Observations and Remote Sensing, IEEE Journal of |
Keywords | Field | DocType |
classification,seeded region growing,shadow,urban,vegetation,china,accuracy,remote sensing,earth,indexes | Computer vision,Vegetation,Nonlinear system,Remote sensing,Region growing,Artificial intelligence,Pixel,Mathematics,Seeding | Journal |
Volume | Issue | ISSN |
PP | 99 | 1939-1404 |
Citations | PageRank | References |
0 | 0.34 | 4 |
Authors | ||
3 | ||