Image Filter (Part of Unit 2)

Q.1 What is Image Noise?

Image noise is a random variation in the colour or brightness of an image, which can
make it look grainy or unclean. The principal source of noise in digital images arises
during image acquisition and transmission. It can also be caused by a number of
factors, including:
1.Lighting: Poor lighting conditions can cause noise
2.Transmission: Channel noise can occur when an image is transmitted over a
communication channel
3.Exposure: Long exposure times can cause noise
4.Camera equipment: The image sensor or circuitry of a camera can cause noise
5.Digitization: Bit errors can occur after an image is digitized

Examples of Noise: Salt and Pepper Noise, Gaussian Noise

Q. What is Image Filtering ?

-Image filtering is a technique that is utilized in image processing to enhance or
revise the visual appearance of the image. Image filtering involves using a filter/kernel
for every pixel in an image so that a new pixel value can be acquired based on the
values of the existing pixels. The filter helps in defining the weights that have been
applied to the neighbouring pixels at the time of the filtering method.

- Image filtering involves highlighting certain features or removing certain features,

image smoothing, sharpening and edge enhancement.
-Convolution means applying filter to an image. It may be applied under spatial domain
or frequency domain.

Q.Give applications of Image Filtering

1. Noise Reduction — Unwanted noise, such as random changes in pixel values,
frequently accompanies images taken by cameras or produced by digital processes.
Noise may be successfully reduced or eliminated using picture filters, creating images
that are clearer and more appealing to the eye.
2. Smoothing And Blurring — Using image filters to smooth or blur a picture
diminishes abrupt transitions and minute details. this benefits applications like picture
denoising, spot removal,and aesthetics
3. Edge Detection — Edge detection image filters highlight the borders separating
various areas or objects in a picture. In processes like object identification, picture
segmentation,and feature extraction, edges must be detected.
4. Image Enhancement — Filters may bring out particular details or qualities in an
image. Ex. sharpening filters can enhance the sharpness and definition of edges,
giving the appearance of greater detail in the image.

Q. Explain Spatial Domain Filters/ Smoothing Filters

Spatial filtering is a process in image processing where the filtering operation is applied
directly to the pixels of an image. This is done by moving a filter mask from point to
point across the image. It modifies each pixel's value based on the values of
neighbouring pixels.

Types of spatial filters:

1.Lowpass filtering : Lowpass filters attenuate or eliminate high-frequency
components, while leaving low frequencies untouched (that is, the filter "passes" low
frequencies). Since high frequency components characterize edges and other sharp
details in image, the effect of lowpass filtering is blurring.
2. Highpass filtering High pass filters attenuate or eliminate low-frequency
components, while leaving high frequencies untouched (that is, the filter "passes" high
frequencies). Since low frequency components characterize overall contrast and
average intensity, the effect of high pass filtering is sharpening of edges and other
sharp details.
3.Smoothing Spatial Filters: 1. Smoothing Spatial Filters are used for blurring and
noise reduction within an image. Noise reduction can be accomplished by blurring.
2.They are used to soften sharp edges.
-Blurring is used in preprocessing steps, such as removal of small details from
an image prior to object extraction and bridging of small gaps in lines or curves.
-Image smoothing improves the image quality and enhances an image by reducing
noise. Hence image smoothing is an important module in various image processing
softwares.

Types of Smoothing Spatial Filters:

1. Smoothing Linear Filter (Mean Filter)
-Smoothing Linear Filters are also known as Mean or averaging filters.
-A smoothing linear filter calculates the average of the pixel values within the
neighbourhood defined by the filter mask.
-The fundamental concept of mean filtering is that each pixel in an image is
substituted with the average value derived from neighbouring pixels confined within a
specified window or kernel.
-The filter replaces each pixel's value with the average gray level of its neighbouring
pixels, leading to a smoother appearance.
-It reduces sharp transitions in gray levels, thereby smoothing the image and

reducing noise or small details.

2.Types of Smoothing Linear Filters (Mean Filters) (Q. Describe Averaging Filter)

A.Averaging Filter: Averaging Filter is a specific type of smoothing filter in image

processing.
Applications: 1. Averaging filters are used to reduce irrelevant details in an image by
eliminating high-frequency components (like noise) and retaining low frequency
components. 2.Used for removing Gaussian noise from images.
A spatial averaging filter in which all the co-efficients are equal is called a Box
Filter.
Averaging Filter is also known as a Low pass filter because it retains low-
frequency components (smooth areas) while reducing high-frequency components
(sharp edges and noise).
Normalization: A m x n mask has a normalizing constant of 1/mn.
Coefficients: All coefficients in the mask are positive.

B. Weighted Averaging Filter (Q. Describe Weighted Averaging Filter)

In Weighted Averaging Filter, pixels are multiplied by different coefficients. The
Center pixel is multiplied by a higher value than any other thus giving this pixel more
importance in the calculation of average.
Since more weightage is given to the centre value, the contribution of centre value
becomes more than the rest of the values. The other pixels are inversely weighted as a
function of their distance from the centre of the mask. This helps to control the blurring
of image.
This is also known as the “Gaussian Filter”

3.Sharpening Spatial Filters: Sharpening Spatial Filter is applied to a digital image

to give it a sharper appearance. The principal objective of a Sharpening Spatial Filter is
to highlight finer detail in an image or to enhance detail that has been blurred, either in
error or as a natural effect of a particular method of image acquisition.
Application: 1. They are used for image smoothing and to give an image a sharper
appearance. 2. Sharpening enhances the edges and fine details in an image, making
features more distinct.
-Sharpening Spatial Filter is also called High Pass Filter (Laplacian Filter). It is a
second-order derivative operator/filter/mask. It detects the image along with horizontal
and vertical directions collectively. The sum of the values of this filter is 0.
A.Highpass filtering High pass filters attenuate or eliminate low-frequency
components, while leaving high frequencies untouched (that is, the filter "passes" high
frequencies). Since low frequency components characterize overall contrast and
average intensity, the effect of high pass filtering is sharpening of edges and other
sharp details.
Mask coefficients must have positive values at centre and negative values elsewhere.
Sum of coefficients must be zero.

B. Sharpening filter: High Boost Filter: High boost filtering is an image sharpening
technique in digital image processing.
Application:It is used to enhance the edges and details of an image without affecting
the smooth areas.This technique is applied to an image after using the Laplacian filter
to the image because the Laplacian filter highlights the edges and details of the image.
The high-boost filter can be used to enhance high frequency component while still
keeping the low frequency components.

High Boost Filter Diagram

C. Laplacian Filter: Laplace smoothing technique is mainly use to detect image

edges. It highlights gray level discontinuities. It is based on second spatial derivation of
an image.
To define Laplacian operator, below equation has been used.

Laplace edge detector uses only one kernel. To detect the edges of an image, this
kernel detects 2nd order derivatives of image’s intensity levels by using only single
pass. We can use “kernel 2" to detect edges with diagonals. It will give better
approximation. Also, Laplace method gives faster calculations than others.

4.Order Statistics Filter: It is based on the ordering of the pixels contained in the
image area encompassed by the filter. It replaces the value of the centre pixel with the
value determined by the ranking result. Edges are better preserved in this filtering.
Types of Order statistics filter:
(i) Minimum filter: 0th percentile filter is the minimum filter. The value of the
center is replaced by the smallest value in the window.

(ii) Maximum filter: 100th percentile filter is the maximum filter. The value of
the center is replaced by the largest value in the window.

(iii) Median filter: Each pixel in the image is considered. First neighbouring
pixels are sorted and original values of the pixel is replaced by the median of the
list. The principal function of median filtering is to force pixels with very
distinct intensities to be more like their neighbours, thus eliminating intensity
spikes.

iv) Median Filter: Median filters are used for noise reduction with less blurring than
linear smoothing filters of same size. The principal function of median filtering is to
force pixels with very distinct intensities to be more like their neighbours, thus
eliminating intensity spikes. Note on Median filter
Application:1. Median filters are used for noise reduction with less blurring than linear
smoothing filters of same size.
2.Median filters are particularly effective in the presence of salt and pepper noise (
impulse noise) because of its appearance as black and white dots superimposed on
an image.
Median filters work by ranking pixels contained within the mask. Each pixel in the
image is considered. First neighbouring pixels are sorted and original values of the
pixel are replaced by the median of the list.The centre value is replaced with the
median value determined by the ranking result.