This image database contains 72 images. There are 6 versions of each still image content. The original added to 5 power of blur. The 12 original images come from the Berkeley Segmentation Dataset as specified in our paper. The database was created to evaluate a no-reference metric that quantifies perceived image quality induced by blur.
Database at FTP, no password. Link: ftp://ftp.ivc.polytech.univ-nantes.fr/Intentionally_Blurred_Background_Images/
References and Citation
Please, cite the following paper in your reference if you use this database for your work [LWR11].
LWR11: H. Liu, J. Wang, J. Redi, P. Le Callet, H. Ingrid. An efficient no-reference metric for perceived blur, 2011 3rd European Workshop on Visual Information Processing (EUVIP), Jun 2011, Paris, France.