Image Quality Measures

Image quality measures are calculated for a distorted image with reference to an original image

Description

Image / Picture Quality Measures

In this application, different image quality measures are calculated for a distorted image with reference to an original image. To test the application, a set of 20 distorted images is included in this package. The list of Image Quality measures implemented in this package include,

1. Structural Content (SC)
2. Mean Square Error (MSE)
3. Peak Signal to Noise Ratio (PSNR in dB)
4. Normalized Cross-Correlation (NCC)
5. Average Difference (AD)
6. Maximum Difference (MD)
7. Normalized Absolute Error (NAE)

Refer "Reference.png" for the mathematical expressions implemented in this package.

Original images are kept in the folder "OriginalImages".
Distorted images are kept in the folder "DistortedImages".

• Required Products: Image Processing Toolbox
• MATLAB release: MATLAB 6.5 (R13)

[Christopher]

Works well

[Shrinivas]

Good application for quality comparison of two images.
Sir, I would like to know
* To what extent distImg is close to OrigImg ( how much % )

[Deni]

Sir, Could you please explain what Does 3 values given by PSNR and MSE for .jpeg extension files means?? (is it because .jpeg consist 3 samples, so it gives 3 values for PSNR and MSE??)

[RAJKUMAR Palaniappan]

hello Sir
can i have your Email id so that i can clarify many more doubts

Thanking you

[uitm put]

dear sir,

i have tried your algorithm in Image and Picture Quality Measures.
But i could not measure the PSNR using lena (from folder "OriginalImages") and lena (from folder "DistortedImages").
In Matlab occured the message " ??? Error using ==> minus, Matrix dimensions must agree".
Could you please give me any ideas how to solve this problem.

Regards,
Puteri Zirwatul Nadila
Postgraduate Student UITM Malaysia
([e-mail])

[Tian]

 

[Sean de]

Don't overwrite the (VERY) important built-in function: error!

Maximum Difference is a pretty much useless measure. You want the maximum absolute difference since you're looking for magnitude.
max(abs(image1(:)-image2(:)))

Assumptions are made in the PSNR calculation that the images have a max of 255 (that for uint8) - this is never alluded to in the documentation.

No error checking.

This suite has potential but the documentation needs to be augmented.

Three stars for now.

[Sean de]

 

[Adi Navve]

Adding to Sean de, I don't know what does "NormalizedCrossCorrelation" compute but it is not normalized cross correlation, not even close. This is wrong and misleading. One should have more responsibility over what he claims in his published code.

[Emad Taha]

very good work

[Akshay Bhosale]