How to detect corrupted image or noisy image using MATLAB ?

if we pass some image to imread('some.jpg') and if that image is corrupted or noisy then is it possible to detect it .

If it is detected then code should give some message to user that image is noisy or corrupted .

[Walter Roberson]

Is this about corruption and/or added noise after the image was originally produced? (Or perhaps even errors in writing the saving the image)? Would these have been deliberately introduced or a problem with the storage or a problem with transmission (e.g., a corrupted packet when downloading over WiFi) ?

Or are you talking about images that were "always" corrupted or noisy; for example turbulence while taking an astronomical photograph.

Answer 1

[Image Analyst]

If you have a reference (perfect, uncorrupted) image, then you can just compare it using standard methods like SAD (Sum of absolute differences), MSE, PSNR, or SSIM.

If you don't have a reference image, there are methods that can judge perceptual image quality. For example right now, today, there is an SPIE conference going on here in San Francisco on 'Methodologies for Quantifying Perceptual Quality" with papers like "Noisy Images - JPEG compressed: subjective and objective image quality evaluations" (to be given tomorrow) and many similar papers.

For more info, see http://en.wikipedia.org/wiki/Image_quality and http://en.wikipedia.org/wiki/Video_quality

[Image Analyst]

If you have video snapshots, see measurements such as Adjectival Categorical Judgment (ACJ) and Double Stimulus Continuous Quality Scale (DSCQS) here: http://en.wikipedia.org/wiki/Comparison_of_video_codecs

No, I don't have MATLAB programs for those.