Stereoscopic Image Quality Assessment Based on Wavelet Transform and Structure Characteristics

Chunping Hou; Honghu Lin

doi:10.3788/LOP55.061005

Journals >Laser & Optoelectronics Progress >Volume 55 >Issue 6 >Page 061005 > Article

Laser & Optoelectronics Progress
Vol. 55, Issue 6, 061005 (2018)

Stereoscopic Image Quality Assessment Based on Wavelet Transform and Structure Characteristics

Chunping Hou and Honghu Lin^*;

Author Affiliations

School of Electrical and Information Engineering, Tianjin University, Tianjin 300072, China

show less

DOI: 10.3788/LOP55.061005 Cite this Article Set citation alerts

Chunping Hou, Honghu Lin. Stereoscopic Image Quality Assessment Based on Wavelet Transform and Structure Characteristics[J]. Laser & Optoelectronics Progress, 2018, 55(6): 061005 Copy Citation Text

show less

Fig. 1. Flowchart of the proposed algorithm

Download full size

Fig. 2. Wavelet decomposition

Download full size

Fig. 3. Dual-tree complex wavelet transform

Download full size

Fig. 4. Different types of distortion images

Download full size

Fig. 5. Feature distributions of 5 distortion types on an image. (a) Phase amplitude distribution; (b) gradient distribution

Download full size

Fig. 6. Prediction model of image quality

Download full size

Algorithm	LIVE3 DIQD Phase 1			LIVE3 DIQD Phase 2
Algorithm	SROCC	PLCC	RMSE	SROCC	PLCC	RMSE
FSIM	0.9135	0.9270	6.0104	0.7874	0.7977	6.4984
MS-SSIM	0.9233	0.9252	6.2245	0.7707	0.7758	7.1221
VIF	0.9087	0.9117	6.7368	0.7167	0.7865	6.9708
Ref. [24]	0.8559	0.8645	8.2424	0.6375	0.6584	8.4956
Ref. [21]	0.9251	0.9350	5.8155	0.8494	0.8628	5.7058
Ref. [11]	0.9347	0.9388	5.6470	0.8935	0.9113	4.6480
Ref. [22]	0.8920	0.8870	6.9950	0.8250	0.8180	6.5000
Ref. [23]	0.3830	0.6260	14.8270	0.5430	0.5680	9.2940
Proposed	0.9336	0.9440	5.3718	0.9074	0.9215	4.3821

Table 1. Performance comparison of algorithms

Distortion	Criteria	FSIM	MS-SSIM	VIF	Ref. [24]	Ref. [21]	Ref. [11]	Ref. [22]	Ref. [23]	Proposed
JP2K	PLCC	0.9188	0.9188	0.9373	0.8381	0.9213	0.9520	0.8480	0.9050	0.9716
	SROCC	0.9038	0.8917	0.9015	0.8388	0.8945	0.9127	0.8370	0.8660	0.9279
	RMSE	4.5051	6.8141	6.0168	7.0658	-	3.9631	-	-	3.0330
JPEG	PLCC	0.6234	0.6859	6.0168	0.2866	0.5200	0.7546	0.6260	0.7290	0.8712
	SROCC	0.5772	0.6123	0.5807	0.0931	0.4951	0.7164	0.6380	0.6750	0.8117
	RMSE	4.8394	6.3445	6.4097	6.2650	-	4.2912	-	-	3.0953
WN	PLCC	0.9283	6.3445	0.9203	0.9280	0.9448	0.9266	0.9250	0.9040	0.9617
	SROCC	0.9345	0.9320	0.9221	0.9284	0.9405	0.9289	0.9310	0.9140	0.9314
	RMSE	5.6401	8.0413	8.6781	6.1964	-	6.2569	-	-	4.3513
G blur	PLCC	0.9370	0.9435	0.9568	0.9475	0.9592	0.9583	0.8990	0.6170	0.9573
	SROCC	0.9223	0.9261	0.9341	0.9345	0.9403	0.9332	0.8330	0.5550	0.8667
	RMSE	4.8019	6.3973	5.4161	4.6291	-	4.1371	-	-	4.0158
FF	PLCC	0.7839	0.8018	0.8600	0.7086	0.8594	0.8620	0.7070	0.5030	0.8431
	SROCC	0.7298	0.7231	0.8042	0.4709	0.7963	0.8286	0.6490	0.6400	0.7265
	RMSE	7.1759	9.8997	8.4537	8.7671	-	6.2993	-	-	6.3050

Table 2. Single distortion performance comparison on LIVE3 DIQD Phase 1

Algorithm	FSIM^[9]	MS-SSIM^[19]	VIF^[20]	Ref. [24]	Proposed
Time cost /s	1.7771	0.9426	1.6726	0.7503	19.1035

Table 3. Comparison of computation time

Chunping Hou, Honghu Lin. Stereoscopic Image Quality Assessment Based on Wavelet Transform and Structure Characteristics[J]. Laser & Optoelectronics Progress, 2018, 55(6): 061005

Download Citation

Set citation alerts for the article

Tools

Set citation alerts for the article

Save the article for my favorites

Paper Information