Automatic building segmentation from remote sensing images using multi-layer level set framework

Guo Jing; Jiang Jie; Cao Shixiang

Towards high resolution remote sensing images, combining with features of buildings, a novel method to extract buildings based on multi-layer level set framework was proposed. Firstly, as far as the impact of shadow and vegetation was concerned, it should be removed on the basis of the separation of gray value thresh and the joint distribution of hue and saturation. Then, an improved C-V level set segmentation algorithm combining with building features of roof's gray and obvious boundaries was applied to extract building regions of similar gray-scales on each gray layer, and thus all building regions of different gray-scales could be extracted layer by layer, followed by layers of segmented regions integration. Finally, the non-building regions were excluded by using normal areas of buildings and related position between buildings and shadows. The experiment results demonstrate that, compared with the traditional level set methods, this one can detect each single building of gray heterogeneity and buildings of multiple shapes and different gray-scales. Meanwhile, compared to the traditional C-V method, it largely reduces the leakage segmentation ratio by 25% and over-segmentation by 22%.

1. Key Laboratory of Precision Opto-mechatronics Technology,Ministry of Education,Beihang University,Beijing 100191,China

Received Date: 2013-08-05

Rev Recd Date: 2013-09-03

Publish Date: 2014-04-25

Key words:

Abstract: Towards high resolution remote sensing images, combining with features of buildings, a novel method to extract buildings based on multi-layer level set framework was proposed. Firstly, as far as the impact of shadow and vegetation was concerned, it should be removed on the basis of the separation of gray value thresh and the joint distribution of hue and saturation. Then, an improved C-V level set segmentation algorithm combining with building features of roof's gray and obvious boundaries was applied to extract building regions of similar gray-scales on each gray layer, and thus all building regions of different gray-scales could be extracted layer by layer, followed by layers of segmented regions integration. Finally, the non-building regions were excluded by using normal areas of buildings and related position between buildings and shadows. The experiment results demonstrate that, compared with the traditional level set methods, this one can detect each single building of gray heterogeneity and buildings of multiple shapes and different gray-scales. Meanwhile, compared to the traditional C-V method, it largely reduces the leakage segmentation ratio by 25% and over-segmentation by 22%.

HTML

Reference (23)

[1]	Liow Y, Pavlidis T. Use of shadows for extracting buildings in aerial images computer vision [J]. Graphics and Image Processing, 1990, 49(2): 242-277.
[2]
[3]	Wegner J D, Rosenhahn B. Segment-based building detection with conditional random fields [C]//Joint Urban Remote Sensing Event, 2011: 205-208.
[4]
[5]	Liu Yang, Tian Xiaojian, Wang Qing, et al. Application of efficient image segmentation method based on local fractal in the infrared cloud image processing[J]. Optics and Precision Engineering, 2011, 19(6): 1367-1374. (in Chinese) 刘洋, 田小建, 王晴, 等. 采用局部分形的高效图像分割方法在红外云图处理中的应用[J]. 光学精密工程, 2011, 19(6): 1367-1374.
[6]
[7]
[8]	Liu Xinyue, Fang Xiaoxuan, Huang Lianqing. Adaptive segmentation of digital mammograms through reinforcement learning [J]. Optics and Precision Engineering, 2005, 13(5): 575-583. (in Chinese)
[9]	Wang Zhi, Jin Guang, Sun Xiaowei. An algorithm of image segmentation for overlapping grain image [J]. Optics and Precision Engineering, 2005, 13(5): 592-598. (in Chinese)
[10]
[11]	Sirmacek B, Unsalan C. A probabilistic framework to detect buildings in aerial and satellite images[J]. IEEE Transactions on Geoscience and Remote Sensing, 2011, 49(1): 211-221.
[12]
[13]
[14]	Chan T F, Vese L A. Active contours without edges [J]. IEEE Transaction on Image Processing, 2001, 10(2): 266-277.
[15]
[16]	Karantzalos K, Paragios N. Recognition-driven two-dimensional competing toward automatic and accurate building detection[J]. IEEE Transactions on Geoscience and Remote Sensing, 2010, 47(1): 133-144.
[17]
[18]	Cao Guo, Yang Xin, Mao Zhihong. A two-stage level set evolution scheme for man-made objects detection in aerial images [C]// IEEE Computer Society Conference on Computer Vision and Pattern Recognition, 2005: 1031-1038.
[19]	Karantzalos K, Argialas D. A region-based level set segmentation for automatic detection of man-made objects from aerial and satellite images [J]. Photogrammetric Engineering Remote Sensing, 2010, 75(6): 667-677.
[20]
[21]	Li Feng, Peng Jiaxiong, Zhang Xiang. Remote sensing image segmentation for nonstationary random field models [J]. Infrared And Laser Engineering, 2003, 32(4): 387-396. (in Chinese)
[22]
[23]	Tian Yan, Su Jianzhong, Tian Jinwen,et al. Numerical algorithm for image segmentation based on variational method [J]. Infrared and Laser Engineering, 2002, 31 (4): 310-313. (in Chinese)

Automatic building segmentation from remote sensing images using multi-layer level set framework

Abstract

References

Proportional views

通讯作者: 陈斌, bchen63@163.com

Article Metrics

Related

Proportional views