Blind quality assessment of image and video based on fragile watermarking and robust features
Multidimensional Systems and Signal Processing
Blind quality assessment is an important feature required by many modern multimedia communication systems. Reduced reference approaches employing fragile watermarking techniques have been proposed earlier for this purpose. This paper proposes an approach which makes use of robust image features and fragile watermarking techniques resulting in a reduced reference strategy. Other than the input image, the approach requires no external information. Robust image features are used to construct the watermark from the image to be transmitted, which is then embedded as a fragile watermark in the image itself. The same features are once again used to construct the reference watermark from the received image, with which the extracted fragile watermark is compared to determine the amount of degradation suffered by it. Construction of the watermark from the image itself, dispenses with the need for embedding an extraneous watermark which must be made known to the user separately. Singular Value Decomposition is used for the extraction of the fragile watermark from the received image while a different approach is used for reduced reference quality assessment of video sequences, making use of consecutive video frames. These form the significant contributions of the present work. The performances of the proposed algorithms have been demonstrated through extensive simulations using different kinds of gray-scale and color images and a number of popularly used test video sequences.
Bhattacharya, Ankan and Palit, Sarbani, "Blind quality assessment of image and video based on fragile watermarking and robust features" (2018). Journal Articles. 1228.