Fast Haar Transform

Powerful Essays
FAST HAAR TRANSFORM BASED FEATURE EXTRACTION FOR MULTIMODAL BIOMETRIC SYSTEM ABSTRACT
In many real-world applications, uni-modal biometric systems often face significant limitations due to sensitivity to noise, interclass variability, data quality, non universality, and other factors. Attempting to improve the performance of individual matchers in such situations may not prove to be highly effective. Multi-biometric systems seek to alleviate some of these problems by providing multiple pieces of evidence of the same identity. These systems help achieve an increase in performance that may not be possible using a single-biometric indicator.
In this project we use multimodal biometric fast recognition method. Subspace learning is the process of finding a proper feature subspace and then projecting high-dimensional data onto the learned low-dimensional subspace. The projection operation requires many floating-point multiplications and additions, which makes the projection process computationally expensive. To tackle this problem, this project proposes two simple-but-effective fast subspace learning and image projection methods, fast Haar transform (FHT) based principal component analysis. The advantages of this methods result from employing both the FHT for subspace learning and the integral vector for feature extraction. Experimental results on face,iris and fingerprint databases demonstrated their effectiveness and efficiency.

LIST OF ABBREVIATIONS
FHT Fast Haar Transform
PCA Principal Component Analysis
FLD Fisher’s Linear Discriminant
DSP Digital Signal Processing
RGB Red Green Blue
FAR False Accept Rate
FRR False Reject Rate
FTE Failure To Enroll rate
GAR Genuine Accept Rate
EER Equal Error Rate



Bibliography: [1] P.N. Belhumeur, J.P. Hespanha, and D.J. Kriegman, “Eigenfaces vs. Fisherfaces: Recognition using class specification linear projection,” IEEE Trans. Pattern anal. Mach. Intell., vol.19,no. 7, pp. 711-720, Jul. 1997. [2] D. Cai, X. He, and J. Han, “SRDA: An efficient algorithm for largescale discriminant analysis,” IEEE Trans. Knowl. Data Eng., vol. 20, no. 1, pp. 1-12, Jan. 2008. [3] P. Getreur, Writing Fast Matlab Code 2008 [Online]. Available:WWW. Eng.auburn.edu/~dmbevly/matlab/ [4] X. He, D. Cai, and J. Han, “Learning a maximum margin subspace for image retrieval,” IEEE Trans. Knowl. Data Eng., vol. 20, no.2, pp. 189-201, Feb.2008. [5] X. He,S. Yan, Y. Hu,P.Niyogi,and H. Zhang, “Face recognition using Laplacian faces, “ IEEE Trans. Pattern Anal. Mach.Intell., vol. 27, no.3, pp. 328-340, Mar. 2005.

You May Also Find These Documents Helpful

  • Powerful Essays

    Fast Fourier Transform

    • 2163 Words
    • 9 Pages

    persons who discussed the FFT (Fast Fourier Transform) for the first time in history. In past years, researchers believed that a discrete Fourier transform can also be calculated and classified as FFT by using the Danielson-Lanczos lemma theorem. By using this theorem, this process is slower than other, as it is slightly tainted in speed due to the power of N (exponent of N) are not 2. Therefore, if the number of points i.e. N is not a power of two, then the transform will only gives you the sets of…

    • 2163 Words
    • 9 Pages
    Powerful Essays
  • Good Essays

    Lecture 11 Fast Fourier Transform (FFT) Weinan E1,2 and Tiejun Li2 1 Department of Mathematics, Princeton University, weinan@princeton.edu 2 School of Mathematical Sciences, Peking University, tieli@pku.edu.cn No.1 Science Building, 1575 Examples Fast Fourier Transform Outline Examples Fast Fourier Transform Applications Applications Examples Fast Fourier Transform Applications Signal processing Filtering: a polluted signal 1.5 1 0.5 0 −0.5 −1 −1.5 0 200 400 600 800…

    • 1866 Words
    • 34 Pages
    Good Essays
  • Satisfactory Essays

    exploit the wavelets which represent the visual feature. We use Haar and D4 wavelet to decompose color images into multilevel scale and wavelet coefficients, with which we perform image feature extraction and similarity match by means of F-norm theory. Furthermore, we also provide a progressive image retrieval strategy to achieve flexible CBIR. We tested five categories of color images in the experiments. The retrieval performance of D4 and Haar wavelet is compared with wavelet histograms in terms of recall…

    • 3812 Words
    • 16 Pages
    Satisfactory Essays
  • Satisfactory Essays

    hilbert transform

    • 326 Words
    • 4 Pages

    2. Analysis of Signals Figure 2.45.: Approximate FTs of two bandlimited signals ¾º½ º Ì ÀÐ ÖØ ÌÖ Ò× ÓÖÑ The Hilbert transform of a function is by definition, H {x(t)} = xh (t) = ∞ x(τ ) dτ t −τ −∞ 1 π (2.171) which is the convolution of x(t) with 1/π t, H {x(t)} = xh (t) = x(t) ∗ 1 πt (2.172) if we take the FT of this convolution, Xh (ω ) = X (ω ) × F 1 πt (2.173) From Example 2.24, F {sgn(t)} = 2 jω (2.174) and using duality from…

    • 326 Words
    • 4 Pages
    Satisfactory Essays
  • Satisfactory Essays

    Advanced Transforms

    • 814 Words
    • 4 Pages

    and STFT | ELEM018: Advanced Transform Methods | | Arsal Javid | 12/1/2011 | [Type the abstract of the document here. The abstract is typically a short summary of the contents of the document. Type the abstract of the document here. The abstract is typically a short summary of the contents of the document.] | Section 2: The following code was used to calculate perform the DFT Function in Matlab: function sw = dft(st) % DFT - Discrete Fourier Transform M = length(st); N = M;…

    • 814 Words
    • 4 Pages
    Satisfactory Essays
  • Powerful Essays

    Laplace Transforms

    • 965 Words
    • 4 Pages

    Laplace Transforms Gilles Cazelais May 2006 Contents 1 Problems 1.1 Laplace Transforms . . . . . . 1.2 Inverse Laplace Transforms . 1.3 Initial Value Problems . . . . 1.4 Step Functions and Impulses 1.5 Convolution . . . . . . . . . . 2 Solutions 2.1 Laplace Transforms . . . . . . 2.2 Inverse Laplace Transforms . 2.3 Initial Value Problems . . . . 2.4 Step Functions and Impulses 2.5 Convolution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .…

    • 965 Words
    • 4 Pages
    Powerful Essays
  • Better Essays

    Discrete Cosine Transform

    • 6299 Words
    • 26 Pages

    Cosine Transform∗ Gilbert Strang† Abstract. Each discrete cosine transform (DCT) uses N real basis vectors whose components are π cosines. In the DCT-4, for example, the jth component of vk is cos(j + 1 )(k + 1 ) N . These 2 2 basis vectors are orthogonal and the transform is extremely useful in image processing. If the vector x gives the intensities along a row of pixels, its cosine series ck vk has the coefficients ck = (x, vk )/N . They are quickly computed from a Fast Fourier Transform. But a…

    • 6299 Words
    • 26 Pages
    Better Essays
  • Good Essays

    Discrete wavelet transform 2 Others Other forms of discrete wavelet transform include the non- or undecimated wavelet transform (where downsampling is omitted), the Newland transform (where an orthonormal basis of wavelets is formed from appropriately constructed top-hat filters in frequency space). Wavelet packet transforms are also related to the discrete wavelet transform. Complex wavelet transform is another form. Properties The Haar DWT illustrates the desirable properties of wavelets…

    • 1342 Words
    • 6 Pages
    Good Essays
  • Better Essays

    The Discrete Cosine Transform (DCT): Theory and Application 1 Syed Ali Khayam Department of Electrical & Computer Engineering Michigan State University March 10th 2003 1 This document is intended to be tutorial in nature. No prior knowledge of image processing concepts is assumed. Interested readers should follow the references for advanced material on DCT. ECE 802 – 602: Information Theory and Coding Seminar 1 – The Discrete Cosine Transform: Theory and Application 1. Introduction…

    • 5443 Words
    • 22 Pages
    Better Essays
  • Satisfactory Essays

    Jiawei Huang 37154135 Fourier Transform Assignment 1. Fourier transform of sine wave (code): import numpy as np import matplotlib.pyplot as plt from scipy.fftpack import fft,fftfreq dt = 0.01 time = np.arange(0,5.,dt) f_1 = 3. a_1 = 2.3 y = a_1*np.sin(2.*np.pi*time*f_1) plt.plot(time,y) plt.xlabel("Time t [s]") plt.ylabel("Wave") plt.title("Wave Signal") plt.show() n = time.shape[-1] transform = (fft(y)[:n/2]) * 2./n frequency = fftfreq(n,time[1]-time[0])[:n/2]…

    • 736 Words
    • 13 Pages
    Satisfactory Essays