Influence of the distortion type on the image quality assessment when reducing its sizes

In this paper, the influence of various types of distortion of an image on its quality while reducing its sizes, is investigated. To assess the image quality, it is proposed to use the method of comparison with the standard using a previously developed measure based on the proximity of the values of the parameters of the Weibull distribution, which describes the gradient field of the image. The well-known TID2013 image database was used as the material, which includes 3000 images distorted by 24 types of distorting algorithms with five levels. Each image of the base was reduced by 2, 4 and 8 times by the two most common methods and compared with the original image-original. The calculations were performed for five types of distortions implemented in the database. To make a decision on the acceptability of the applied quality measure, the calculated measure values were compared with the subjective quality ratings provided along with the documentation on the TID2013 database. The comparison was carried out using Spearman’s correlation coefficient. It is shown that the average values of correlations for all images at three types of distortions are very high, while for the other two they are unacceptably low. An attempt has been made to explain this situation by the properties of distorting algorithms that change the structural properties of the image to varying degrees.

The possibility of comparing images of the same scene, but with different resolutions, is demonstrated.

1. N. Ponomarenko, L. Jin, O. Ieremeiev, V. Lukin, K. Egiazarian, J. Astola, B. Vozel, K. Chehdi, M. Carli, F. Battisti, C.-C. Jay Kuo // Signal Processing: Image Communication. – 2015. – V. 30. – P. 57–77.

2. Starovoitov F. V., Starovoitov V. V. Comparative analysis of no-reference quality measures for digital images // System analysis and applied information science. – 2017. – № 1. – С. 24–32.

3. Y. I. Golub, F. V. Starovoitov, V. V. Starovoitov. Impact of image size reducing for image quality assessment // System analysis and applied information science. – 2020. – № 2. – С. 35–45.

4. Asatryan D., Egiazarian K. Quality Assessment Measure Based on Image Structural Properties // Proc. of the International Workshop on Local and Non-Local Approximation in Image Processing, Finland, Helsinki, 2009. – P. 70–73.

5. Asatryan D. G. Image blur estimation using gradient field analysis [In Russian]. Computer Optics. – 2017. – Vol. 41. – № 6. – P. 957–962.

6. Аsatryan D. A Novel Technique for No-Reference Image Quality Assessment // Proceedings of International Conference on Computer Science and Information Technologies, 2019. – P. 201–203.

7. Asatryan D. Gradient-based technique for image structural analysis and applications // Computer Optics. – 2019. – Vol. 43. – № 2. – P. 245–250.