LCE/7.5.

1/RC 01

TEACHING NOTES
Department: ELECTRONICS & COMMUNICATION ENGINEERING Unit: II Date: Topic name: Image Transforms - 1 No. of marks allotted by JNTUK: Books referred: 01. Digital Image Processing by R C Gonzalez and R E Woods 02. www.wikipedia.org 03. www.google.com Image Transforms: The Fourier Transform is an important image processing tool which is used to decompose an image into its sine and cosine components. The output of the transformation represents the image in the Fourier or frequency domain, while the input image is the spatial domain equivalent. In the Fourier domain image, each point represents a particular frequency contained in the spatial domain image. The Fourier Transform is used in a wide range of applications, such as image analysis, image filtering, image reconstruction and image compression. The DFT (Discrete Fourier Transform) is the sampled Fourier Transform and therefore does not contain all frequencies forming an image, but only a set of samples which is large enough to fully describe the spatial domain image. The number of frequencies corresponds to the number of pixels in the spatial domain image, i.e. the image in the spatial and Fourier domain is of the same size. For a square image of size NN, the two-dimensional DFT is given by:

Where f(a,b) is the image in the spatial domain and the exponential term is the basis function corresponding to each point F(k,l) in the Fourier space. The equation can be interpreted as: the value of each point F(k,l) is obtained by multiplying the spatial image with the corresponding base function and summing the result. The basic functions are sine and cosine waves with increasing frequencies, i.e. F(0,0) represents the DC-component of the image which corresponds to the average brightness and F(N-1,N-1) represents the highest frequency. In a similar way, the Fourier image can be re-transformed to the spatial domain. The inverse Fourier transform is given by:

To obtain the result for the above equations, a double sum has to be calculated for each image point. However, because the Fourier Transform is separable, it can be written as,

Page 1 of 7

Faculty/Date:

HOD/Date: LCE/7.5.1/RC 01

TEACHING NOTES
Department: ELECTRONICS & COMMUNICATION ENGINEERING Unit: II Date: Topic name: Image Transforms - 2 No. of marks allotted by JNTUK: Books referred: 01. Digital Image Processing by R C Gonzalez and R E Woods 02. www.wikipedia.org 03. www.google.com Where

Using these two formulas, the spatial domain image is first transformed into an intermediate image using N one-dimensional Fourier Transforms. This intermediate image is then transformed into the final image, again using N onedimensional Fourier Transforms. Expressing the two-dimensional Fourier Transform in terms of a series of 2N one-dimensional transforms decreases the number of required computations. Even with these computational savings, the ordinary one-dimensional DFT 2 has N complexity. This can be reduced to Nlog2N if we employ the Fast Fourier Transform (FFT) to compute the one-dimensional DFTs. This is a significant improvement, in particular for large images. There are various forms of the FFT and most of them restrict the size of the input image that may be transformed, often to N = 2n where n is an integer. The mathematical details are well described in the literature. The Fourier Transform produces a complex number valued output image which can be displayed with two images, either with the real and imaginary part or with magnitude and phase. In image processing, often only the magnitude of the Fourier Transform is displayed, as it contains most of the information of the geometric structure of the spatial domain image. However, if we want to retransform the Fourier image into the correct spatial domain after some processing in the frequency domain, we must make sure to preserve both magnitude and phase of the Fourier image. The Fourier domain image has a much greater range than the image in the spatial domain. Hence, to be sufficiently accurate, its values are usually calculated and stored in float values. Walsh Transform: The Hadamard transform (also known as the Walsh-Hadamard transform, Hadamard-Rademacher-Walsh transform, Walsh transform, or Walsh-Fourier transform) is an example of a generalized class of Fourier transforms. It is named Page 2 of 7

for the French mathematician Jacques Solomon Hadamard, the GermanAmerican mathematician Hans Adolph Rademacher, and the American mathematician Joseph Leonard Walsh. It performs an orthogonal, symmetric, involutional, linear operation on 2m real numbers (or complex numbers, although the Hadamard matrices themselves are purely real). The Hadamard transform can be regarded as being built out of size-2 discrete Fourier transforms (DFTs), and is in fact equivalent to a multidimensional DFT of size . It decomposes an arbitrary input vector into a superposition of Walsh functions. Faculty/Date: HOD/Date: LCE/7.5.1/RC 01

TEACHING NOTES
Department: ELECTRONICS & COMMUNICATION ENGINEERING Unit: II Date: Topic name: Image Transforms - 3 No. of marks allotted by JNTUK: Books referred: 01. Digital Image Processing by R C Gonzalez and R E Woods 02. www.wikipedia.org 03. www.google.com The Hadamard transform Hm is a 2m 2m matrix, the Hadamard matrix (scaled by a normalization factor), that transforms 2m real numbers xn into 2m real numbers Xk. The Hadamard transform can be defined in two ways: recursively, or by using the binary (base-2) representation of the indices n and k. Recursively, we define the 1 1 Hadamard transform H0 by the identity H0 = 1, and then define Hm for m > 0 by:

Where the 1/2 is a normalization i.e., sometimes omitted. Thus, other than this normalization factor, the Hadamard matrices are made up entirely of 1 and 1. Equivalently, we can define the Hadamard matrix by its (k, n)-th entry by writing, And Where the kj and nj are the binary digits (0 or 1) of n and k, respectively. In this case, we have:

This is exactly the multidimensional DFT, normalized to be unitary, if the inputs and outputs are regarded as multidimensional arrays indexed by the nj and kj, respectively. Page 3 of 7

Some examples of the Hadamard matrices follow:

(This H1 is precisely the size-2 DFT. It can also be regarded as the Fourier transform on the two-element additive group of Z/ (2))

Faculty/Date:

HOD/Date: LCE/7.5.1/RC 01

TEACHING NOTES
Department: ELECTRONICS & COMMUNICATION ENGINEERING Unit: II Date: Topic name: Discrete Cosine Transform No. of marks allotted by JNTUK: Books referred: 01. Digital Image Processing by R C Gonzalez and R E Woods 02. www.wikipedia.org 03. www.google.com

Where is bitwise dot product of the binary representations of the numbers i and j. For example, Agreeing with the above (ignoring the overall constant). Note that the first row, first column of the matrix is denoted by H00 The rows of the Hadamard matrices are the Walsh functions. Discrete Cosine Transform: A discrete cosine transform (DCT) expresses a sequence of finitely many data points in terms of a sum of cosine functions oscillating at different frequencies. DCTs are important to numerous applications in science and engineering, from lossy compression of audio and images (where small highfrequency components can be discarded), to spectral methods for the numerical solution of partial differential equations. The use of cosine rather than sine functions is critical in these applications: for compression, it turns out that cosine Page 4 of 7

functions are much more efficient (as explained below, fewer are needed to approximate a typical signal), whereas for differential equations the cosines express a particular choice of boundary conditions. In particular, a DCT is a Fourier-related transform similar to the discrete Fourier transform (DFT), but using only real numbers. DCTs are equivalent to DFTs of roughly twice the length, operating on real data with even symmetry (since the Fourier transform of a real and even function is real and even), where in some variants the input and/or output data are shifted by half a sample. There are eight standard DCT variants, of which four are common. The most common variant of discrete cosine transform is the type-II DCT, which is often called simply "the DCT"; its inverse, the type-III DCT, is correspondingly often called simply "the inverse DCT" or "the IDCT". Two related transforms are the discrete sine transforms (DST), which is equivalent to a DFT of real and odd functions, and the modified discrete cosine transforms (MDCT), which is based on a DCT of overlapping data. Faculty/Date: HOD/Date: LCE/7.5.1/RC 01

TEACHING NOTES
Department: ELECTRONICS & COMMUNICATION ENGINEERING Unit: II Date: Topic name: Haar Transform & Hotelling Transform - 1 No. of marks allotted by JNTUK: Books referred: 01. Digital Image Processing by R C Gonzalez and R E Woods 02. www.wikipedia.org 03. www.google.com Haar Transform: The Haar transform is the simplest of the wavelet transforms. This transform cross-multiplies a function against the Haar wavelet with various shifts and stretches, like the Fourier transform cross-multiplies a function against a sine wave with two phases and many stretches. The Haar transform is derived from the Haar matrix. An example of a 4x4 Haar matrix is shown below:

The Haar transform can be thought of as a sampling process in which rows of the transform matrix act as samples of finer and finer resolution. Compare with the Walsh transform, which is also 1/1, but is non-localized. Hotelling Transform: In statistics, principal components analysis (PCA) is a technique to simplify a dataset; more formally it is a transform used for reducing dimensionality in a dataset while retaining those characteristics of the dataset that contribute most to its variance. These characteristics may be the 'most important', but this is not necessarily the case, depending on the application. Page 5 of 7

PCA is also called the Karhunen-Love transform or the Hotelling transform. PCA has the specialty of being the optimal linear transform for keeping the subspace that has largest variance. However this comes at the price of greater computational requirement, e.g. if compared to the discrete cosine transform. Unlike other linear transforms, the PCA does not have a fixed set of basis vectors. Its basis vectors depend on the data set. The principal component w1 of a dataset x can be defined as (assuming zero mean, i.e. E(x)=0)

(See arg max for the notation) With the first k-1 component, the k-th component can be found by subtracting the first k-1 principal components from x:

And by substituting this as the new dataset to find a principal component in:

Faculty/Date:

HOD/Date: LCE/7.5.1/RC 01

TEACHING NOTES
Department: ELECTRONICS & COMMUNICATION ENGINEERING Unit: II Date: Topic name: Hotelling Transform - 2 No. of marks allotted by JNTUK: Books referred: 01. Digital Image Processing by R C Gonzalez and R E Woods 02. www.wikipedia.org 03. www.google.com A simpler way to calculate the components wi uses the covariance matrix of x, the measurement vector. By finding the eigenvalues and eigenvectors of the covariance matrix, we find that the eigenvectors with the largest eigenvalues correspond to the dimensions that have the strongest correlation in the dataset. The original measurements are finally projected onto the reduced vector space. Related (or even more similar than related?) is the calculus of empirical orthogonal functions (EOF). Another method of dimension reduction is a self-organizing map. If PCA is used in pattern recognition an often useful alternative is the linear discriminant analysis that takes into account the class separability, which is not the case for PCA.

Page 6 of 7

Faculty/Date:

HOD/Date:

Page 7 of 7