Pattern recognition using maximum likelihood estimation and orthogonal subspace projection

M. M. Islam; M. S. Alam

doi:10.1117/12.679642

30 August 2006 Pattern recognition using maximum likelihood estimation and orthogonal subspace projection

M. M. Islam, M. S. Alam

Proceedings Volume 6311, Optical Information Systems IV; 63110Y (2006) https://doi.org/10.1117/12.679642
Event: SPIE Optics + Photonics, 2006, San Diego, California, United States

Abstract

Hyperspectral sensor imagery (HSI) is a relatively new area of research, however, it is extensively being used in geology, agriculture, defense, intelligence and law enforcement applications. Much of the current research focuses on the object detection with low false alarm rate. Over the past several years, many object detection algorithms have been developed which include linear detector, quadratic detector, adaptive matched filter etc. In those methods the available data cube was directly used to determine the background mean and the covariance matrix, assuming that the number of object pixels is low compared to that of the data pixels. In this paper, we have used the orthogonal subspace projection (OSP) technique to find the background matrix from the given image data. Our algorithm consists of three parts. In the first part, we have calculated the background matrix using the OSP technique. In the second part, we have determined the maximum likelihood estimates of the parameters. Finally, we have considered the likelihood ratio, commonly known as the Neyman Pearson quadratic detector, to recognize the objects. The proposed technique has been investigated via computer simulation where excellent performance has been observed.

Citation Download Citation

M. M. Islam and M. S. Alam "Pattern recognition using maximum likelihood estimation and orthogonal subspace projection", Proc. SPIE 6311, Optical Information Systems IV, 63110Y (30 August 2006); https://doi.org/10.1117/12.679642

ACCESS THE FULL ARTICLE

INSTITUTIONAL
Select your institution to access the SPIE Digital Library.

SELECT YOUR INSTITUTION

PERSONAL
Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.

PERSONAL SIGN IN

No SPIE Account? Create one

PURCHASE THIS CONTENT

SUBSCRIBE TO DIGITAL LIBRARY

50 downloads per 1-year subscription

Members: $195

Non-members: $335 ADD TO CART

25 downloads per 1 - year subscription

Members: $145

Non-members: $250 ADD TO CART

PURCHASE SINGLE ARTICLE

Includes PDF, HTML & Video, when available