Method to Improve Accuracy of PSF parameters of Motion Blur Images using Window Functions

Sanyuan Ju; Shuhui Gao

doi:10.3788/LOP223009

Laser & Optoelectronics Progress, Volume. 61, Issue 4, 0437007(2024)

Method to Improve Accuracy of PSF parameters of Motion Blur Images using Window Functions

Sanyuan Ju and Shuhui Gao^*

Author Affiliations

School of Investigation, People's Public Security University of China, Beijing 100038, China

show less

Abstract Get PDF(in Chinese)

Figures & Tables(15)

Fig. 1. Fingerprint degradation image model

Download full size

Fig. 2. Less information after blocking (left) and multi-angle application (right)

Download full size

Fig. 3. Lena image (left), motion blurred image (middle), and spectrum of the blurred image (right)

Download full size

Fig. 4. Windowed, median-filtered image (left), and binarized image (right)

Download full size

Fig. 5. Open operation processing image (left) and Canny edge detection image (right)

Download full size

Fig. 6. Radon transform estimating the blur angle. (a) Radon transform of the traditional algorithm; (b) Radon transform of the windowing algorithm; (c) Radon transform of the proposed algorithm

Download full size

Fig. 7. Schematic of the Radon transform in the fuzzy angle direction

Download full size

Fig. 8. Flowchart of the proposed algorithm

Download full size

Fig. 9. Restoration of motion-blurred images

Download full size

Table 1. Results of the fuzzy parameters of the Lena image estimated by the algorithm in Ref.[10] and the proposed algorithm

View table

Table 1. Results of the fuzzy parameters of the Lena image estimated by the algorithm in Ref.[10] and the proposed algorithm

True blurred angle /（°）	Angle estimated by the algorithm in Ref.［10］/（°）	Angle estimated by the proposed algorithm /（°）	True blurred length /pixel	Length estimated by the algorithm in Ref.［10］/pixel	Length estimated by the proposed algorithm /pixel
20	20	20.0	15	15	15.14
40	40	40.0	20	19	19.74
60	60	60.0	25	25	25.00
80	81	80.0	30	30	29.70
100	98	100.0	35	34	35.00
120	120	120.0	40	39	40.00

Table 2. Results of the fuzzy parameters of the Cameraman image estimated by the algorithm in Ref.[10] and the proposed algorithm

View table

Table 2. Results of the fuzzy parameters of the Cameraman image estimated by the algorithm in Ref.[10] and the proposed algorithm

True blurred angle /（°）	Angle estimated by the algorithm in Ref.［10］/（°）	Angle estimated by the proposed algorithm /（°）	True blurred length /pixel	Length estimated by the algorithm in Ref.［10］/pixel	Length estimated by the proposed algorithm /pixel
15	15	90.0	5	6	2.01
30	30	30.5	10	10	10.24
45	44	44.0	15	16	14.86
60	61	60.0	20	21	19.87
75	76	75.0	25	26	25.00
90	92	90.0	30	29	30.00

Table 3. Results of the fuzzy parameters of the Lena image estimated by the algorithm in Ref.[11] and the proposed algorithm

View table

Table 3. Results of the fuzzy parameters of the Lena image estimated by the algorithm in Ref.[11] and the proposed algorithm

True blurred angle /（°）	Angle estimated by the algorithm in Ref.［11］/（°）	Angle estimated by the proposed algorithm /（°）	True blurred length /pixel	Length estimated by the algorithm in Ref.［11］/pixel	Length estimated by the proposed algorithm /pixel
30	30	30.0	25	25	25.21
50	50	50.0	30	31	30.00
70	70	70.0	35	35	35.29
90	90	90.0	40	40	40.00
110	109	109.5	45	44	45.67
130	131	130.0	50	50	50.00

Table 4. Results of the fuzzy parameters of the Cameraman image estimated by the algorithm in Ref.[11] and the proposed algorithm

View table

Table 4. Results of the fuzzy parameters of the Cameraman image estimated by the algorithm in Ref.[11] and the proposed algorithm

True blurred angle /（°）	Angle estimated by the algorithm in Ref.［11］/（°）	Angle estimated by the proposed algorithm /（°）	True blurred length /pixel	Length estimated by the algorithm in Ref.［11］/pixel	Length estimated by the proposed algorithm /pixel
25	25.0	25.0	15	15	14.76
40	40.0	40.0	20	20	20.00
55	55.0	55.5	25	24	25.21
70	70.5	70.0	30	30	30.00
85	84.0	85.5	35	35	35.18
100	100.0	100.0	40	41	40.00

Table 5. Restoration results of the proposed algorithm for different images
View table
Table 5. Restoration results of the proposed algorithm for different images
Image kind PSNR /dB SSIM
Car 34.15 0.84
Flower 33.76 0.82
Tree 32.05 0.85

Table 6. Restoration results of different algorithms on the GOPRO dataset
View table
Table 6. Restoration results of different algorithms on the GOPRO dataset
Parameter Algorithm in Ref.［9］ Algorithm in Ref.［11］ Proposed algorithm
PSNR /dB 24.08 27.18 33.32
MSE 1.74 1.51 0.83
SSIM 0.64 0.82 0.84

Tools

Get Citation

Copy Citation Text

Sanyuan Ju, Shuhui Gao. Method to Improve Accuracy of PSF parameters of Motion Blur Images using Window Functions[J]. Laser & Optoelectronics Progress, 2024, 61(4): 0437007

Download Citation

EndNote(RIS)BibTex Plain Text

Set citation alerts for article

Save article for my favorites

Paper Information

Category: Digital Image Processing

Received: Nov. 9, 2022

Accepted: Jan. 17, 2023

Published Online: Feb. 27, 2024

The Author Email: Gao Shuhui (gaoshuhui@ppsuc.edu.cn)

DOI:10.3788/LOP223009

Topics

laser devices and laser physics

Lasers and Laser Optics

Laser physics

laser manufacturing

Instrumentation, Measurement and Metrology

Table 1. Results of the fuzzy parameters of the Lena image estimated by the algorithm in Ref.[10] and the proposed algorithm

Table 1. Results of the fuzzy parameters of the Lena image estimated by the algorithm in Ref.[10] and the proposed algorithm

Table 2. Results of the fuzzy parameters of the Cameraman image estimated by the algorithm in Ref.[10] and the proposed algorithm

Table 2. Results of the fuzzy parameters of the Cameraman image estimated by the algorithm in Ref.[10] and the proposed algorithm

Table 3. Results of the fuzzy parameters of the Lena image estimated by the algorithm in Ref.[11] and the proposed algorithm

Table 3. Results of the fuzzy parameters of the Lena image estimated by the algorithm in Ref.[11] and the proposed algorithm

Table 4. Results of the fuzzy parameters of the Cameraman image estimated by the algorithm in Ref.[11] and the proposed algorithm

Table 4. Results of the fuzzy parameters of the Cameraman image estimated by the algorithm in Ref.[11] and the proposed algorithm

Table 5. Restoration results of the proposed algorithm for different images

Table 5. Restoration results of the proposed algorithm for different images

Table 6. Restoration results of different algorithms on the GOPRO dataset

Table 6. Restoration results of different algorithms on the GOPRO dataset