General Mean Pooling Strategy for Color Image Quality Assessment

Yuemei Ma; Haiying Chen; Guojun Liu

doi:10.3788/LOP55.021007

Journals >Laser & Optoelectronics Progress >Volume 55 >Issue 2 >Page 021007 > Article

Laser & Optoelectronics Progress
Vol. 55, Issue 2, 021007 (2018)

General Mean Pooling Strategy for Color Image Quality Assessment

Yuemei Ma¹, Haiying Chen^1、2, and Guojun Liu^、*

Author Affiliations

¹ School of Mathematics and Statistics, Ningxia University, Yinchuan, Ningxia 750021, China

¹ School of Preparatory Education for Nationalities, Ningxia University, Yinchuan, Ningxia 750021, China

² School of Mathematics and Statistics, Wuhan University, Wuhan, Hubei 430072, China

show less

DOI: 10.3788/LOP55.021007 Cite this Article Set citation alerts

Yuemei Ma, Haiying Chen, Guojun Liu. General Mean Pooling Strategy for Color Image Quality Assessment[J]. Laser & Optoelectronics Progress, 2018, 55(2): 021007 Copy Citation Text

show less

Reference image and distorted images of varying degrees in TID2013 database. (a) Reference image; (b) distorted image (EMSE=76.54, RPSNR=29.29); (c) distorted image (EMSE=0.88, RPSNR=48.68)

Fig. 1. Reference image and distorted images of varying degrees in TID2013 database. (a) Reference image; (b) distorted image (EMSE=76.54, RPSNR=29.29); (c) distorted image (EMSE=0.88, RPSNR=48.68)

Download full size

Fig. 2. SROCC curves varying with r. (a) General mean pooling strategy 1; (b) general mean pooling strategy 2

Download full size

Fig. 3. SROCC mesh graphs varying with ω. (a) Variation in SROCC of GM-C-SSIM2 with ω2 and ω3; (b) variation in SROCC of GM-C-GSSIM2 with ω2 and ω3; (c) variation in SROCC of GM-C-FSIM2 with ω1 and ω2

Download full size

Fig. 4. Scatter plots between objective scores and MOS of each evaluation algorithm in TID2013 database. (a) C-SSIM; (b) C-GSSIM; (c) C-FSIM; (d) GM-C-SSIM1; (e) GM-C-GSSIM1; (f) GM-C-FSIM1; (g) GM-C-SSIM2; (h) GM-C-GSSIM2; (i) GM-C-FSIM2

Download full size

Algorithm	SROCC	KROCC	PLCC	RMSE
SSIM^[3]	0.7417	0.5588	0.7895	0.7608
C-SSIM	0.8115	0.6203	0.8221	0.7058
GM-SSIM1^[14]	0.7665	0.5869	0.8093	0.7283
GM-SSIM2^[14]	0.7825	0.5885	0.8021	0.7404
GM-C-SSIM1	0.8509	0.6596	0.8527	0.6475
GM-C-SSIM2	0.8561	0.6689	0.8526	0.6478
GSSIM^[4]	0.7586	0.5768	0.8179	0.7132
C-GSSIM	0.8508	0.6625	0.8629	0.6266
GM-GSSIM1^[14]	0.7919	0.6115	0.8515	0.6500
GM-GSSIM2^[14]	0.7466	0.5573	0.7968	0.7491
GM-C-GSSIM1	0.8708	0.6845	0.8800	0.5889
GM-C-GSSIM2	0.8748	0.6937	0.8911	0.5626
FSIM^[5]	0.8015	0.6289	0.8589	0.6349
C-FSIM^[5]	0.8510	0.6665	0.8769	0.5959
GM-FSIM1^[14]	0.8157	0.6439	0.8690	0.6133
GM-FSIM2^[14]	0.8137	0.6399	0.8649	0.6223
GM-C-FSIM1	0.8613	0.6783	0.8834	0.5810
GM-C-FSIM2	0.8841	0.6975	0.8933	0.5571

Table 1. Performance comparison of IQA in TID2013 database

Yuemei Ma, Haiying Chen, Guojun Liu. General Mean Pooling Strategy for Color Image Quality Assessment[J]. Laser & Optoelectronics Progress, 2018, 55(2): 021007

Download Citation

Set citation alerts for the article

Tools

Set citation alerts for the article

Save the article for my favorites

Paper Information