No-Reference Method for Effective Bandwidth Estimation

Barak Fishbaina*, Leonid Yaroslavskya, Ianir Idesesa, Frederique (Crete) Roffetb

 

aDept. of Electrical Engineering – Physical Electronics, the Iby and Aladar Fleischman Faculty of Engineering, Tel-Aviv University, Tel Aviv 69978, Israel

 

bSTMicroelectronics, 12 rue Jules Horowitz, 38019 Grenoble, France.

 

 

The work done so far is described in:

§        F. Crete, T. Dolmiere, P. Ladret and M. Nicolas, “The Blur Effect: Perception and Estimation with a New No-Reference Perceptual Blur Metric”, proc. of Human Vision and Electronic Imaging XII/Electronic Imaging 2007, SPIE Vol. 6492, San-Jose, CA, 2007.

§        Barak Fishbain, Leonid Yaroslavsky, Ianir Ideses, Frederique Roffet, “No-Reference Method for Effective Bandwidth Estimation”, Proc. Image Quality and System Performance V/ Electronic Imaging 2008, SPIE Vol. 6808, San-Jose, CA, 2008.

Matlab Source Code

The image sequences are provided for personal use and in order to facilitate basic research. We ask not to use these data in any commercial way. We appreciate an email message indicating who has copied the data and understand that the authors will be acknowledged in case results obtained with these data will be published. We explicitly object to redistribution of any versions of our image sequences in order to maintain compatibility of sources in case results obtained by different groups have to be compared. Any further distribution of these sequences will proceed under the same conditions.

Image Degradation Measure

IDM = PerceptualBlurMetric (Image, FiltSize)

 

Image Effective Bandwidth

EffectiveBW = get_matching_rnd_bandwidth(IDM)

 

This file needs the following MAT file Q_rnd_vec1000.mat

 

Test Sets

256x256 random image set with 256 different effective bandwidths.

 



* barak@eng.tau.ac.il; Phone 972-3-640-8014; fax 972-3-641-0189