Paper
18 April 2003 Texture-based cloud detection in MODIS images
Ramon E. Vasquez, Vidya B. Manian
Author Affiliations +
Proceedings Volume 4882, Remote Sensing of Clouds and the Atmosphere VII; (2003) https://doi.org/10.1117/12.463042
Event: International Symposium on Remote Sensing, 2002, Crete, Greece
Abstract
Statistical and spatial features are computed to detect clouds from MODIS Terra images. The presence of cirrus clouds is specifically studied. 1.38 μm band of MODIS is used. Cirrrus clouds that do not appear in visible imagery can be detected in this band. Although clouds appear to be uniform, they exhibit a texture pattern. This is evident from the texture measures computed for cloud pixels. Therefore, a supervised classification procedure that makes use of texture features is adopted. Second order gray level features, spectral features, wavelet features are computed. The non-cloud pixels and cloud pixels are identified, followed by classification of the cloud pixels. A comparison of the ability of the features is presented. To show the advantage of texture classification, the classified results are compared with results of cloud pixel identification from thresholding of pixel values. Images obtained from the MODIS sensor over the Caribbean region is classified. Wavelet features have been found to better characterize the cloud texture in the MODIS bands 17 and higher.
© (2003) COPYRIGHT Society of Photo-Optical Instrumentation Engineers (SPIE). Downloading of the abstract is permitted for personal use only.
Ramon E. Vasquez and Vidya B. Manian "Texture-based cloud detection in MODIS images", Proc. SPIE 4882, Remote Sensing of Clouds and the Atmosphere VII, (18 April 2003); https://doi.org/10.1117/12.463042
Lens.org Logo
CITATIONS
Cited by 3 scholarly publications.
Advertisement
Advertisement
RIGHTS & PERMISSIONS
Get copyright permission  Get copyright permission on Copyright Marketplace
KEYWORDS
Clouds

Wavelets

MODIS

Image classification

Image segmentation

Remote sensing

Feature extraction

Back to Top