Scribd
Upload
8 views

IPT Module 2

Uploaded by

b20cs36
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
8 views

IPT Module 2

Uploaded by

b20cs36
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 11

Module 2

Basic concept of spatial domain and frequency domain: In image processing, the spatial domain is the
domain of the original image, where the pixel values are represented as a function of their spatial
coordinates. The frequency domain is the domain of the Fourier transform of the image, where the
frequency components of the image are represented as a function of their spatial frequencies.

The spatial domain is a good representation for images when we want to perform operations that
involve the spatial relationships between pixels, such as edge detection, smoothing, and filtering. The
frequency domain is a good representation for images when we want to perform operations that
involve the frequency content of the image, such as image compression, denoising, and sharpening.

Here are some examples of how the spatial domain and the frequency domain are used in image
processing:

TRACE KTU
Edge detection: Edge detection algorithms work by identifying the locations of sharp changes
in intensity in an image. These changes in intensity correspond to high-frequency components
in the frequency domain.
Smoothing: Smoothing algorithms work by averaging the values of neighboring pixels in an
image. This reduces the amount of high-frequency noise in the image.
Filtering: Filters can be used to enhance or remove certain features from an image. For
example, a low-pass filter can be used to smooth an image, while a high-pass filter can be used
to enhance edges.
Image compression: Image compression algorithms work by reducing the amount of data
required to represent an image. This is done by removing high-frequency components from the
image, which are not as noticeable to the human eye.
Image denoising: Image denoising algorithms work by removing noise from an image. This is
done by identifying the noise components in the frequency domain and then removing them.
Image sharpening: Image sharpening algorithms work by enhancing the edges in an image.
This is done by increasing the contrast between the edges and the surrounding pixels.
The spatial domain and the frequency domain are two powerful tools that can be used to
manipulate and analyze images. By understanding the key differences between these two
domains, you can use them to improve the quality and functionality of your images.
Need of Image Transform, Basic properties of unitary transform:
Image transforms are used in image processing for a variety of tasks, including:
Compression: Image transforms can be used to compress images by representing them in a
more compact form. This is done by decomposing the image into a set of basis functions, and
then representing each basis function with a small number of coefficients.
Feature extraction: Image transforms can be used to extract features from images. This is done
by identifying the basis functions that are most important for representing the image content.
These features can then be used for tasks such as classification, object detection, and
segmentation.
Image restoration: Image transforms can be used to restore images that have been corrupted by
noise or other distortions. This is done by decomposing the image into a set of basis functions,
and then using the coefficients to reconstruct the image.
Unitary transforms are a special type of image transform that have the following properties:
Energy preservation: The sum of the squares of the coefficients in a unitary transform is the
same as the sum of the squares of the original image pixels. This property ensures that the
image information is not lost during the transform.
Orthogonality: The basis functions in a unitary transform are orthogonal to each other. This
means that they are uncorrelated, which makes them well-suited for compression and feature
extraction.

TRACE KTU
Invertibility: A unitary transform is invertible, which means that the original image can be
reconstructed from the coefficients. This property is important for image restoration
applications.
Some common unitary transforms used in image processing include:
Discrete Fourier transform (DFT): The DFT is a unitary transform that decomposes an image
into a set of basis functions that are sinusoids of varying frequencies.
Discrete cosine transform (DCT): The DCT is a unitary transform that decomposes an image
into a set of basis functions that are cosine functions of varying frequencies.
Wavelet transform: The wavelet transform is a family of unitary transforms that decomposes
an image into a set of basis functions that are localized in both space and frequency.
Image transforms are a powerful tool that can be used for a variety of tasks in image processing.
By understanding the basic properties of unitary transforms, you can use them to improve the
quality, efficiency, and flexibility of your image processing applications.
Discrete Fourier Transform:
Proof DFT is Unitary:

TRACE KTU
4 order DFT Transform coefficients (Derivation):

TRACE KTU
TRACE KTU
TRACE KTU
TRACE KTU
TRACE KTU
TRACE KTU
Discrete Cosine Transform:
Hadamard Transform:

TRACE KTU
TRACE KTU

You might also like