Share Email Print
cover

Proceedings Paper

Matching clouds and shadows based on high-resolution satellite image
Format Member Price Non-Member Price
PDF $14.40 $18.00

Paper Abstract

Clouds are the obstruction of visual and infrared remote sensing and their shadows may also lead up to an intolerable bias of the true reflectance of the underlying terrain elements. Thus a reliable cloud and shadow mask is essential before the further processing. Clouds cast shadows on the earth's surface. On the high resolution remote sensing images, clouds' profiles and their shadows' are resemblant. Based on this truth, we employed a robust image matching algorithm called Modified Partial Hausforff Distance(MPHD) to find the match with every cloud and its shadow and finally calculated the pixel distance between them. Before the match task we took into account topologic relationships such as coverage and fragmentation to improve the match result. Not only were the match pairs detected but also the pixel distances from each cloud to its shadow were obtained. Then we can use a pixel distance to predict a shadow of a cloud by translating the cloud. Given sunbeam's direction and viewing angles we may get cloud height with simple geometry calculation.

Paper Details

Date Published: 8 December 2006
PDF: 7 pages
Proc. SPIE 6408, Remote Sensing of the Atmosphere and Clouds, 640819 (8 December 2006); doi: 10.1117/12.693703
Show Author Affiliations
Xiaodong Chen, State Key Lab. of Satellite Ocean Environment Dynamics (China)
Jianyu Chen, State Key Lab. of Satellite Ocean Environment Dynamics (China)
Delu Pan, State Key Lab. of Satellite Ocean Environment Dynamics (China)
Zhihua Mao, State Key Lab. of Satellite Ocean Environment Dynamics (China)
Haiqing Huang, State Key Lab. of Satellite Ocean Environment Dynamics (China)


Published in SPIE Proceedings Vol. 6408:
Remote Sensing of the Atmosphere and Clouds
Si-Chee Tsay; Teruyuki Nakajima; Ramesh P. Singh; R. Sridharan, Editor(s)

© SPIE. Terms of Use
Back to Top