Scribd
Upload
7 views5 pages

Image Enhancement and Restoration Techniques

The document describes 5 image processing projects: 1. Reduce additive periodic noise from an image using Fourier analysis. 2. Enhance details in a corridor image obscured by glare using homographic filtering. 3. Restore a degraded flyer image using least squares filtering in the frequency domain. 4. Similarly restore a degraded truck image. 5. Compress an image using Walsh-Hadamard wavelet transform and calculate compression efficiency and error.

Uploaded by

AliffPadzi
Copyright
© All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
7 views5 pages

Image Enhancement and Restoration Techniques

The document describes 5 image processing projects: 1. Reduce additive periodic noise from an image using Fourier analysis. 2. Enhance details in a corridor image obscured by glare using homographic filtering. 3. Restore a degraded flyer image using least squares filtering in the frequency domain. 4. Similarly restore a degraded truck image. 5. Compress an image using Walsh-Hadamard wavelet transform and calculate compression efficiency and error.

Uploaded by

AliffPadzi
Copyright
© All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd

PROJECT 1

Image File : [Link]

The enclosed image has been corrupted by additive periodical noise


defined as:

𝑖 𝑥, 𝑦 = 𝐴 sin 𝜔0 𝜋 𝑥 + 𝑦

where 𝐴 = 20 and 𝜔0 = 0.1. Perform image enhancement in order to


reduce or eliminate 𝑖 𝑥, 𝑦 in 𝑓 𝑥, 𝑦

Hint:

First show the Fourier transform of 𝑖 𝑥, 𝑦 is given by the pair of conjugate


impulses

𝐴 𝜔0 𝜔0 𝜔0 𝜔0
𝐼 𝑢, 𝑣 = −𝑗 𝛿 𝑢− ,𝑣 − −𝛿 𝑢+ ,𝑣 +
2 2𝜋 2𝜋 2𝜋 2𝜋

Second design suitable filter to remove 𝐼 𝑢, 𝑣


PROJECT 2
Image File : [Link]

The image shown is the front corridor of the EE building captured in mid-
afternoon sunlight. As a result the details inside the corridor are obscured
by the glare from the outside the building.

Perform image enhancement by bringing the details of objects inside the


corridor.

Hint:

Use homographic filtering with Gaussian high pass filter defined as:

𝐷 2 𝑢 ,𝑣
− 2
𝐻 𝑢, 𝑣 = 𝛾𝐻 − 𝛾𝐿 1− 𝑒 𝐷0 + 𝛾𝐿

Choose suitable 𝐷0 , 𝛾𝐻 and 𝛾𝐿 values.


PROJECT 3
Image File : [Link]

The image 𝑓(𝑥, 𝑦) shown a flyer which has been degraded with
degradation function

1 𝑛−1
𝑔 𝑥, 𝑦 = 𝑘=0 𝑓 𝑥 − 𝑘, 𝑦
𝑛

where n is an integer. Derive the degradation function in frequency domain.


Use this function to perform image restoration using Least Square Filtering
technique. Try to read the texts printed on the flyer from the restored
image.

Show results for various n and 𝛾.

Hint:

Refer lecture notes and Gonzalez and Woods, pg: 266-269.


PROJECT 4
Image File : [Link]

The image 𝑓(𝑥, 𝑦) shown a truck which has been degraded with
degradation function

1 𝑛−1
𝑔 𝑥, 𝑦 = 𝑘=0 𝑓 𝑥, 𝑦 − 𝑘
𝑛

where n is an integer. Derive the degradation function in frequency domain.


Use this function to perform image restoration using Least Square Filtering
technique. Try to read the texts printed on the side of a lorry from restored
image.

Show results for various n and 𝛾.

Hint:

Refer lecture notes and Gonzalez and Woods, pg: 266-269.


PROJECT 5

image file : [Link]

Compare and sketch Walsh-Hadamard wavelets needed to perform W-H


transform of a given image. Hence display the output image after W-H
transform. Compress the image using suitable combination of W-H basis
images (use your visual judgment to determine the number of basis
functions needed by repeatedly compressing and decompressing the
image). Also calculate the compression efficiency and root mean square
error of the compressed image.

Hint:

Treat coloured image as the combination of three primary colurs, i.e. Red
(R), Green (G) and Blue (B). Compress the image separately on R, G and
B planes.

You might also like