Skip to Main content Skip to Navigation
Conference papers

An Overflow Free Fixed-point Eigenvalue Decomposition Algorithm: Case Study of Dimensionality Reduction in Hyperspectral Images

Abstract : We consider the problem of enabling robust range estimation of eigenvalue decomposition (EVD) algorithm for a reliable fixed-point design. The simplicity of fixed-point circuitry has always been so tempting to implement EVD algorithms in fixed-point arithmetic. Working towards an effective fixed-point design, integer bit-width allocation is a significant step which has a crucial impact on accuracy and hardware efficiency. This paper investigates the shortcomings of the existing range estimation methods while deriving bounds for the variables of the EVD algorithm. In light of the circumstances, we introduce a range estimation approach based on vector and matrix norm properties together with a scaling procedure that maintains all the assets of an analytical method. The method could derive robust and tight bounds for the variables of EVD algorithm. The bounds derived using the proposed approach remain same for any input matrix and are also independent of the number of iterations or size of the problem. Some benchmark hyperspectral data sets have been used to evaluate the efficiency of the proposed technique. It was found that by the proposed range estimation approach, all the variables generated during the computation of Jacobi EVD is bounded within ±1.
Document type :
Conference papers
Complete list of metadata

Cited literature [42 references]  Display  Hide  Download

https://hal.inria.fr/hal-01635958
Contributor : Bibek Kabi <>
Submitted on : Thursday, November 30, 2017 - 2:42:57 PM
Last modification on : Tuesday, March 31, 2020 - 9:12:01 PM

File

1711.10600.pdf
Files produced by the author(s)

Identifiers

  • HAL Id : hal-01635958, version 1

Collections

Citation

Bibek Kabi, Anand S Sahadevan, Tapan Pradhan. An Overflow Free Fixed-point Eigenvalue Decomposition Algorithm: Case Study of Dimensionality Reduction in Hyperspectral Images. 2017 Conference On Design And Architectures For Signal And Image Processing (DASIP). , 〈http://dasip2017.esit.rub.de/〉, Sep 2017, Dresden, Germany. ⟨hal-01635958⟩

Share

Metrics

Record views

404

Files downloads

734