Notes while reading book - "Deblurring Images.."

This post would be useful only to those users who are reading the book
Deblurring Images: Matrices, Spectra, and Filtering
by Per Christian Hansen, James G. Nagy, and Dianne P. O’Leary

So if you are one of the privileged few who actually read this type of intellectual goolah, go ahead... dive right in.

Deblurring is used in barcode readers [63 siam]

Challenge 1=====
 -

For given image size, largest noise before inverted noise dominates deblurred image:
1E-7

Value of 'noise' does depend on size of image.
Higher sized image gets corrupted at lower value of noise.
m = n = 128, noise = 1E-6
m = n = 256, noise = 1E-7
m = n = 512, noise = 1E-8

Challenge 2=====

No not able to deblur image with naive approach. Thus unable to read text.



Challenge 3=====

maximum value for relative error values
relative error = 5%,   E_norm =   2.3299e-07
relative error = 10%, E_norm =   4.6598e-07
relative error = 20%, E_norm =   9.3196e-07


Challenge 4=====

Value of k at which text could be read:
k = 53
"The question is not what you look at but what you see."

Interestingly, reduced size image has better clarity of letters:

Challenge 5

Task 6
To create blurred color image, image manipulation in DOUBLE format. But before displaying image, DOUBLE image converted to UINT8 using:

I_Color_blurred = uint8(Id_Color_blurred);
figure(6); imshow(I_Color_blurred, []); title(' Blurred color Image');

Task 7
Weighted average image is different from rgb2gray image obtained directly in MATLAB.

Reason is the implementation of rgb2gray:

RGB2GRAY converts RGB images to grayscale by eliminating the
    hue and saturation information while retaining the
    luminance.

... more posts coming up