Table 2. The results of the proposed method and other existing methods based on PSNR (first rows) and SSIM (second rows) criteria on Gaussian simulated datasets.

Sequence	Criteria	Two-stage	NDL	SGL	Centroid	RPCA	JSUB	Proposed
Car-front	PSNR	15.3815	19.9009	16.7093	19.5172	24.0959	20.9223	25.1344
	SSIM	0.5448	0.8136	0.6801	0.8163	0.9137	0.8375	0.9460
Road	PSNR	26.58	27.4061	23.9782	30.03	33.8682	32.1232	32.0174
	SSIM	0.7822	0.8036	0.7638	0.8608	0.9063	0.9005	0.9226
Mean	PSNR	20.9807	23.6535	20.3437	24.7736	28.9820	26.5227	28.5759
	SSIM	0.6635	0.8086	0.7219	0.8385	0.91	0.869	0.9343

Note:

The best result in each row is shown in bold.