Efficient Big-Size Light Region Splitting Scheme for High-Resolution SAR Imagery
Building detection is a key step to extract information about urban regions from high-resolution synthetic aperture radar (SAR) images. Light features of building often merge into a big-size region in SAR images due to SAR imaging mechanism and geometric layout of building, and make it difficult to further extract building information. A method that is a hybrid of morphological operators for building detection from high-resolution SAR images is proposed. This method first detects buildings by a combination of two intensity thresholds and morphological reconstruction, then to the big-size light region, a hybrid of several morphological operators are used to split it into proper number of small segments. The proposed method performs better than classic K-means and ISODATA in testing high-resolution TerraSAR-X data, in both shape preservation and building number estimation.
KeywordsSAR imagery Building information extraction Hybrid morphological operators
This work was supported by the National Natural Science Foundation of China (No.40901207, No.41161065).
- 1.Soergel U, Thoennessen U, Brenner A, Stilla U (2006) High-resolution SAR data: new opportunities and challenges for the analysis of urban areas. In: Proceedings of radar sonar and navigation, vol 46, pp 294–300Google Scholar
- 2.Wu Q, Chen R, Sun H Cao Y (2011) Urban building density detection using high resolution SAR imagery. Joint urban remote sensing event 2011, vol 74. Munich, Germany, pp 36–37 Google Scholar
- 3.He W, Hellwich O (2009) Bayesian building extraction from high resolution polarimetric sar data, vol 63. IGARSS2009 Cape Town South Africa, pp 47–53Google Scholar
- 4.Sportouche H, Tupin F, Denise L (2011) A symmetric scheme for building reconstruction from a couple of HR optical and SAR data, vol 35. JURSE 2011, Munich, Germany, pp 73–77Google Scholar
- 5.Thiele A, Thoennessen U, Cadario E, Schulz K, Soergel U (2006) Building recognition in urban areas from multi-aspect high-resolution interferometric SAR data, vol 35. EUSAR 2006, Dresden, Germany, pp 452–456Google Scholar
- 6.Ball G, Hall D (1967) A clustering technique for summarizing multivariate data. Behav Sci 12:153–155, 47:245–247Google Scholar
- 9.Beucher S, Meyer F (1993) The morphological approach to segmentation: the watershed transformation. In: Dougherty ER (ed) Mathematical morphology in image processing, vol 474. pp 433–481Google Scholar