Open Access Open Access  Restricted Access Subscription Access

A Review: No Reference Image Quality Assessment Based On Multi-Expert Convolutional Neural Networks

Shilpa Ajith

Abstract


Using No Reference (NR) Image Quality Assessment (IQA) algorithm can measure the quality of distorted images without referencing the original image. This property is importance in image processing field. Diversity of the distortion types and image contents, leads difficulty in the existing NR IQA algorithms to maintain the best performance. As a solution to this problem, we develop a novel NR IQA algorithm based on multi-expert convolutional neural networks (CNNs), which consists of distortion type classification, IQA algorithms based on CNN and fusion algorithm. First, a distortion type classifier identifies various distortion type of the input image. Then, propose a multi-expert CNN based IQA algorithms for each of this distortion types. Finally, a fusion algorithm is adopted which aggregate the classification result of both the distortion types and multi-expert CNN based image quality predictions.

 

Keywords: IQA, no reference image quality assessment, distortion type classification, CNN, multi-expert CNN


Full Text:

PDF

References


Sang Q., Qi H., Wu X., et al. No-reference image blur index based on singular value curve. J. Vis. Commun. Image Represent. Vol. 25, no. 7, 1625-1630p, 2014.

Zhang J., Ong S.H., Le T.M. Kurtosis-based no-reference quality assessment of JPEG2000 images. Signal Process, Image Commun. vol. 26, no. 1, 1323p, 2011.

Yang G., Liao Y., Zhang Q., et al. No-Reference quality assessment of noise-distorted images based on frequency mapping. IEEE Access, vol. 5, 23146-23156p, Oct. 2017.

Saad M.A., Bovik A.C., Charrier C. Blind image quality assessment:A natural scene statistics approach in the DCT domain. IEEE Trans Image Process. vol. 21, no. 8, 3339-3352p, Aug. 2012.


Refbacks

  • There are currently no refbacks.