Abstract:
The development of technology in High spatial resolution has provided opportunities for the application of remote sensing, and the limitation of traditional multi-scale segmentation which is based on region is becoming more obviously. The paper provides a new method of multi-scale segmentation which is combined by edge detection and Regional feature detection. We used nonlinear filtering to segment, and then extracted the information of damaged house caused by earthquake successfully by uniting the edge detection Canny operator embed confidence and the advanced Regional feature detection of multi-scale segmentation using Mean Shift operator. The new method restrains effectively the phenomenon of over segmentation and under segmentation, maintains the space structure of object. The total confident of Kappa reached 0.9623 in the experiment and the time of multi-scale segmentation has shorten less than 20 seconds while processing two images, and the results finally provides the information after disaster.
Keywords:
Edge Detection; Regional Feature Detection; Nonlinear Filtering; Multi-Scale Segmentation; Canny Operator; Mean Shift Operator
References:
[1] G.F. Zhang, “Mu1ti-sea1e remote sensing image segmentation based on granularity theory,” Wuhan University (2010).
[2] J.J. Zhu, X.P. Du,X.T. Fan et al., “Multi-scale Edge Detection and Multi-scale Segmentation of Imagery,” Geography and Geo-Information Science.vol.02, (2013), pp.45-48+127.
[3] W. Tao,H. Jin, L. Liu, “A new image thresholding method based on graph cuts,” ICASSP,IEEE International Conference on Acoustics, speech and signal processing (2007)
[4] L. Wang, “Analysing Classification and Segmentation parameters selection in High resolution remote sensing image using based on Object,” Central South University (2014)
[5] B. Zhang, “He Binbin. Multi-scale Segmentation of High-resolution Remote Sensing Image Based on Improved Watershed Transformation,” Journal of Geo-Information Science, no.1, (2007), pp.142-150.
[6] S.Wang, “A panoramic image multi-scale segmentation method,” Kunming University Of Science And Technology (2014)
[7] J.X. Zhou, “Study on Mean Shift Segmentation and application of Remotely Sensed Imagery,” Central South University (2012)
[8] H.J. You, “SAR Change Detection by Multi-scale Segmentation and Optimization,” Geomatics and Information Science of Wuhan University,vol.5, (2011), pp.531-534.
[9] R.Q. Ye, R.Q. Niu, L.P. Zhang. “Mineral Features Extraction and Analysis Based on Multiresolution Segmentation of Petrographic Images,” Journal of Jilin University (Earth Science Edition),vol.04, (2011), pp.1253-1261.
[10] J.C. Guo,P.J. Li,X.B. Xiao, “A Hierarchical Segmentation Method for Multispectral Imagery,” Acta Scientiarum Naturalium Universitatis Pekinensis, vol.2,(2009),pp.306-310.
[11] H.Z. Ji, High Resolution Remote Sensing Image Multi-scale Segmentation Support by Spectral Graph Theory,” Wuhan University (2011)
[12] L. Fan,Y.Z. Cheng,L.G. Wang et al.,“Estimation of Winter Planting Area Using Object-oriented Method Based Onmuliti-scale Segmentation,” Chinese Journal of Agricultural Resources and Regional Planning, vol.6, (2010), pp.44-51.
Citations:
APA:
Aquesta, C. (2021). Damaged House Information Extracting of After Disaster Using Multi-Scale Grid Computing Segmentation. International Journal of Grid and Distributed Computing (IJGDC), ISSN: 2005-4262 (Print); 2207-6379 (Online), NADIA, 14(2), 47-54. doi: 10.33832/ijgdc.2021.14.2.05.
MLA:
Aquesta, Carmela, et al. “Damaged House Information Extracting of After Disaster Using Multi-Scale Grid Computing Segmentation.” International Journal of Grid and Distributed Computing (IJGDC), ISSN: 2005-4262 (Print); 2207-6379 (Online), NADIA, vol. 14, no. 2, 2021, pp. 47-54. IJGDC, http://article.nadiapub.com/IJGDC/vol14_no2/5.html.
IEEE:
[1] C. Aquesta, "Damaged House Information Extracting of After Disaster Using Multi-Scale Grid Computing Segmentation." International Journal of Grid and Distributed Computing (IJGDC), ISSN: 2005-4262 (Print); 2207-6379 (Online), NADIA, vol. 14, no. 2, pp. 47-54, July 2021.