Share Email Print

Proceedings Paper

Automatic selection of optimal segmentation scales for high-resolution remote sensing images
Format Member Price Non-Member Price
PDF $14.40 $18.00
cover GOOD NEWS! Your organization subscribes to the SPIE Digital Library. You may be able to download this paper for free. Check Access

Paper Abstract

To extract information from high resolution images is a challenge work.Compared tothe traditional pixel-based approach, the advantages of object-oriented classification methods are well documented. However, the appropriate scale parametersofthese methods are difficult to be determined, andthe choices of scale parametersareof high importance, whichwill havea strong effect on the segmentation effectiveness. Whereas the evaluations of the quality of a segmentation method are still mainly based onsubjective judgment, which is a complicated process and lacksstability and reliability. Thus, an objective and unsupervised method needs to beestablished for selecting suitable parameters for a multi-scale segmentation to ensure the bestresults. In this work, a novicemethod is introduced to choose the optimal parameter for themulti-scale segmentation. For large information in band itself and weak relationship among multispectral bands, valuable bands should be selected from original data and weighed by the degreeofcorrelation. Then thresholds of all 3 selected bands ranging from 20 to 200 (intervals of 10)are created in Definiens Professional 8.7. It considers that a segmentation has two desirable properties: each of the resulting segments should be internally homogeneous and should be distinguishable from its neighborhood. Therefore, the global intra-segment and inter-segment heterogeneity indexes are taken into account to identify the optimal segmentation scale. Finally, cubic spline interpolation is applied to select the optimalsegmentation scale. As a result, the measure combining a spatial autocorrelation indicator and a variance indicator shows that the method can improve the precision in global segmentation.

Paper Details

Date Published: 24 September 2013
PDF: 11 pages
Proc. SPIE 8869, Remote Sensing and Modeling of Ecosystems for Sustainability X, 88691A (24 September 2013); doi: 10.1117/12.2021606
Show Author Affiliations
Ruijuan Yin, East China Normal Univ. (China)
Ctr. for Earth Observation and Digital Earth (China)
Runhe Shi, East China Normal Univ. (China)
Ctr. for Earth Observation and Digital Earth (China)
Wei Gao, East China Normal Univ. (China)
Colorado State Univ. (United States)
Ctr. for Earth Observation and Digital Earth (China)

Published in SPIE Proceedings Vol. 8869:
Remote Sensing and Modeling of Ecosystems for Sustainability X
Wei Gao; Thomas J. Jackson; Jinnian Wang; Ni-Bin Chang, Editor(s)

© SPIE. Terms of Use
Back to Top