Fast Blind Restoration of QR Code Images Based on Blurred Imaging Mechanism

Rongjun CHEN; Zhijun ZHENG; Huimin ZHAO; Jinchang REN; Hongzhou TAN

doi:10.3788/gzxb20215007.0710003

Journals >Acta Photonica Sinica >Volume 50 >Issue 7 >Page 91 > Article

Acta Photonica Sinica
Vol. 50, Issue 7, 91 (2021)

Fast Blind Restoration of QR Code Images Based on Blurred Imaging Mechanism

Rongjun CHEN^1、2, Zhijun ZHENG¹, Huimin ZHAO¹, Jinchang REN^1、3、*, and Hongzhou TAN²

Author Affiliations

¹School of Computer Science， Guangdong Polytechnic Normal University， Guangzhou50665， China

²School of Electronics and Information Technology， Sun Yat-sen University， Guangzhou510006， China

³National Subsea Centre， Robert Gordon University， Aberdeen， AB10 7RQ， U.K.

show less

DOI: 10.3788/gzxb20215007.0710003 Cite this Article

Rongjun CHEN, Zhijun ZHENG, Huimin ZHAO, Jinchang REN, Hongzhou TAN. Fast Blind Restoration of QR Code Images Based on Blurred Imaging Mechanism[J]. Acta Photonica Sinica, 2021, 50(7): 91 Copy Citation Text

show less

Fig. 1. Schematic diagram of the lens imaging

Download full size | View in the Article

Fig. 2. QR code image design diagram

Download full size | View in the Article

Fig. 3. Method architecture diagram of the paper

Download full size | View in the Article

Fig. 4. Comparison diagram of QR code symbols in different sizes and positions

Download full size | View in the Article

Fig. 5. Comparison of four restoration algorithms

Download full size | View in the Article

Fig. 6. Comparison of four restoration algorithms.

Download full size | View in the Article

QR code symbols	$E d g e B$	$d$	$O_{x}$	$O_{y}$	$S$	$D$	$R$
Original size	81	31	146	147	87	77	10
9/10 times	86	28	146	145	92	82	10
8/10 times	92	23	142	148	98	88	10
7/10 times	98	21	145	141	105	95	10
Upper left	44	23	94	119	50	40	10
Upper right	154	23	204	120	160	150	10
Lower left	49	23	99	183	55	45	10
Lower right	159	23	209	182	165	155	10

Table 1. Comparison of restoration parameters of QR code symbols of different sizes and positions

View in the Article

Figures	Image quality	Blurred	YAN Y et al ^［23］	BAI Y C et al ^［13］	WEN F et al ^［15］	Ours
Fig. 5（1）	PSNR	60.956 7	61.493 8	61.684 8	61.073 8	61.663 9
Fig. 5（1）	SSIM	0.995 3	0.996 8	0.997 3	0.996 6	0.998 3
Fig. 5（2）	PSNR	60.895 1	66.765 0	64.787 6	62.461 1	64.892 0
Fig. 5（2）	SSIM	0.994 6	0.999 5	0.999 3	0.998 3	0.999 3
Fig. 5（3）	PSNR	60.892 9	61.404 2	61.215 3	61.740 6	62.647 7
Fig. 5（3）	SSIM	0.995 3	0.996 5	0.996 6	0.996 8	0.998 6

Table 2. Analysis of restoration effect

View in the Article

Algorithms	Recognition rate			Average operation time T/s
Algorithms	Zxing （v3.4.0）	Wochacha （v10.0.1）	Wechat （v8.0.5）	Average operation time T/s
YAN Y et. al ^［23］	24.62%	72.31%	79%	14.879 6
BAI Y C et. al ^［13］	20.00%	89.23%	90.77%	9.035 5
WEN F et. al ^［15］	6.92%	44.62%	51.51%	9.633 1
Ours	70.77%	86.92%	93.08%	0.329 2

Table 3. Analysis of operation time and recognizability

Rongjun CHEN, Zhijun ZHENG, Huimin ZHAO, Jinchang REN, Hongzhou TAN. Fast Blind Restoration of QR Code Images Based on Blurred Imaging Mechanism[J]. Acta Photonica Sinica, 2021, 50(7): 91

Download Citation

Tools

Save the article for my favorites

Paper Information