Use of multivariate analysis to minimize collecting of infrared images and classify detected objects

Thomas Svensson; Dietmar Letalick

doi:10.1117/12.2241821

21 October 2016 Use of multivariate analysis to minimize collecting of infrared images and classify detected objects

Thomas Svensson, Dietmar Letalick

Proceedings Volume 9988, Electro-Optical Remote Sensing X; 99880M (2016) https://doi.org/10.1117/12.2241821
Event: SPIE Security + Defence, 2016, Edinburgh, United Kingdom

Abstract

An infrared image contains spatial and radiative information about objects in a scene. Two challenges are to classify pixels in a cluttered environment and to detect partly obscured or buried objects like mines and IEDs. Infrared image sequences provide additional temporal information, which can be utilized for a more robust object detection and an improved classification of object pixels. A manual evaluation of multi-dimensional data is generally time consuming and inefficient and therefore various algorithms are used. By a principal component analysis (PCA) most of the information is retained in a new, reduced system with fewer dimensions. The principal component coefficients (loadings) are here used both for classifying detected object pixels and for reducing the number of images in the analysis by computing of score vectors. For the datasets studied, the number of required images can be reduced significantly without loss of detection and classification ability. This allows for a more sparse sampling and scanning of larger areas when using a UAV, for example.

Citation Download Citation

Thomas Svensson and Dietmar Letalick "Use of multivariate analysis to minimize collecting of infrared images and classify detected objects", Proc. SPIE 9988, Electro-Optical Remote Sensing X, 99880M (21 October 2016); https://doi.org/10.1117/12.2241821

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