Building information is extremely important for many applications such as urban planning, telecommunication, or environment monitoring etc. Previous attempts at automating the building detection process from images has met with limited success due to (1) spectral similarities between building rooftops and roads, (2) lack of spatial processing parameters for building geometry. A novel approach is presented in this paper based on aerial images and range images. By using the height information provided by range images, buildings could be easily distinguish from other objects (e.g. roads). A new perceptual grouping technique is introduced for the purpose of organizing the low-level features (arcs and line segments) which are extracted from aerial images. The final contours of the buildings are generated with the help of regularization algorithm. After reconstruction, a refinement is applied by an object-based perceptual grouping method. Finally, the approach is applied to two datasets and promising experimental results are shown
[1]
Wu Yi-rong.
Contour extraction of regular targets in images
,
2007
.
[2]
Kourosh Khoshelham,et al.
Integration of multi-source data for automated building extraction
,
2004
.
[3]
Ronald J. Holyer,et al.
Circle detection for extracting eddy size and position from satellite imagery of the ocean
,
1994,
IEEE Trans. Geosci. Remote. Sens..
[4]
Edward M. Riseman,et al.
Token-based extraction of straight lines
,
1989,
IEEE Trans. Syst. Man Cybern..
[5]
John F. Canny,et al.
A Computational Approach to Edge Detection
,
1986,
IEEE Transactions on Pattern Analysis and Machine Intelligence.
[6]
J. Canny.
A Computational Approach toEdgeDetection
,
1986
.