Blind document image quality prediction based on modification of quality aware clustering method integrating a patch selection strategy

Highlights：

• A blind image quality assessment measure based on foreground information is proposed.

• The proposed measure is based on a Bag-of-Visual-Words and patch selection strategy.

• The proposed measure is particularly adapted for document images.

• The proposed measure is competitive with the state-of-the-art methods.

• The proposed method is suitable for expert systems embedded in mobile applications.