Fast segmentation of high-resolution satellite images using watershed transform combined with an efficient region merging approach

Authors:
Qiuxiao Chen;Chenghu Zhou;Jiancheng Luo;Dongping Ming
Affiliations:
The State Key Lab of Resources & Environmental Information System, Chinese Academy of Sciences, Beijing, China;The State Key Lab of Resources & Environmental Information System, Chinese Academy of Sciences, Beijing, China;The State Key Lab of Resources & Environmental Information System, Chinese Academy of Sciences, Beijing, China;The State Key Lab of Resources & Environmental Information System, Chinese Academy of Sciences, Beijing, China
Venue:
IWCIA'04 Proceedings of the 10th international conference on Combinatorial Image Analysis
Year:
2004

Citing 4
Cited 2

Watersheds in Digital Spaces: An Efficient Algorithm Based on Immersion Simulations

IEEE Transactions on Pattern Analysis and Machine Intelligence
Computer Vision

Computer Vision
Adaptive Split-and-Merge Segmentation Based on Piecewise Least-Square Approximation

IEEE Transactions on Pattern Analysis and Machine Intelligence
Hybrid image segmentation using watersheds and fast region merging

IEEE Transactions on Image Processing

A basin morphology approach to colour image segmentation by region merging

ACCV'07 Proceedings of the 8th Asian conference on Computer vision - Volume Part I
Automatic recognition of landforms on mars using terrain segmentation and classification

DS'06 Proceedings of the 9th international conference on Discovery Science

Quantified Score

Hi-index	0.00

Visualization

Abstract

High-resolution satellite images like Quickbird images have been applied into many fields. However, researches on segmenting such kind of images are rather insufficient partly due to the complexity and large size of such images. In this study, a fast and accurate segmentation approach was proposed. First, a homogeneity gradient image was produced. Then, an efficient watershed transform was employed to gain the initial segments. Finally, an improved region merging approach was proposed to merge the initial segments by taking a strategy to minimize the overall heterogeneity increased within segments at each merging step, and the final segments were obtained. Compared with the segmentation approach of a commercial software eCognition, the proposed one was a bit faster and a bit more accurate when applied to the Quickbird images.