Classification of pixels in a noisy grayscale image of polar ice

IEEE Transactions on Geoscience and Remote Sensing - Tập 40 Số 8 - Trang 1879-1884 - 2002
S. Das Peddada1, J.T.G. Hwang2
1Department of Statistics, University of Virginia, Charlottesville, VA, USA
2Department of Mathematics, Cornell University, Ithaca, NY, USA

Tóm tắt

Often, in synthetic aperture radar (SAR) images of polar ice, one encounters shadow-like features across the images. Such features make it difficult to classify pixels into ice and water. Accordingly, it becomes a challenge to determine the true size and boundaries of ice floes in an SAR image of polar ice. We develop a simple statistical procedure which classifies pixels of an image by eliminating the effects of shadow-like features. Methodology developed in this paper is illustrated using some noisy SAR images of ice floes in the Arctic sea.

Từ khóa

#Pixel #Gray-scale #Ice #Synthetic aperture radar #Arctic #Monte Carlo methods #Classification algorithms #Radar tracking #Smoothing methods #Image sampling

Tài liệu tham khảo

10.1080/03610928008827932 10.1093/biomet/65.2.273 10.1029/1999JC000110 10.1109/36.508408 titterington, 1985, Statistical Analysis of Finite Mixture Distributions higdon, 1998, auxiliary variable methods for markov chain monte carlo with applications, J Amer Stat Assoc, 93, 585, 10.1080/01621459.1998.10473712 10.1109/36.655340 10.2307/2529563 bose, 1997, a region based segmentation method for multichannel image data, J Amer Stat Assoc, 92, 92, 10.1080/01621459.1997.10473606 mclachlan, 1988, Mixture Models Inference and Applications to Clustering 10.1080/01621459.1985.10478110