Trilateral filter smoothing and denoising algorithm for gated light ranging and imaging

Chenhao XIA; Xinwei WANG; Liang SUN; Bo SONG; Songtao FAN; Yan ZHOU

doi:10.3788/IRLA20240204

Infrared and Laser Engineering, Volume. 53, Issue 8, 20240204(2024)

Trilateral filter smoothing and denoising algorithm for gated light ranging and imaging

Chenhao XIA^1,2, Xinwei WANG^1,2,3, Liang SUN¹, Bo SONG^1,2, Songtao FAN¹, and Yan ZHOU^1,2,3

Author Affiliations

¹Optoelectronic System Laboratory, Institute of Semiconductors, Chinese Academy of Sciences, Beijing 100083, China

²Center of Materials Science and Optoelectronics Engineering, University of Chinese Academy of Sciences, Beijing 100049, China

³School of Electronic, Electrical and Communication Engineering, University of Chinese Academy of Sciences, Beijing 100049, China

show less

Abstract Get PDF(in Chinese)

Figures & Tables(14)

Fig. 1. Trilateral filter smoothing and denoising algorithm for gated LiRAI. (a) Principle of gated LiRAI based on laser range-gated 3D imaging; (b) Pipeline of trilateral filter smoothing and denoising algorithm

Download full size

View in Article

Fig. 2. Illustration of trilateral filtering

Download full size

View in Article

Fig. 3. Simulated pyramid point clouds

Download full size

View in Article

Fig. 4. Select edge regions

Download full size

View in Article

$Comparison of simulated pyramid point cloud smoothing effects between trilateral filtering and bilateral filtering with \begin{document}${\sigma _n}$\end{document}=0.2 m$

Fig. 5. Comparison of simulated pyramid point cloud smoothing effects between trilateral filtering and bilateral filtering with \begin{document}${\sigma _n}$\end{document}=0.2 m

Download full size

View in Article

Fig. 6. Comparison of simulated pyramid point cloud smoothing effects between trilateral filtering and bilateral filtering with σ_n=0.3 m

Download full size

View in Article

Fig. 7. Gated LiRAI system and targets

Download full size

View in Article

Fig. 8. Select regions for quantitative analysis

Download full size

View in Article

Fig. 9. Select regions to analyze PSNR of 2 plants

Download full size

View in Article

Fig. 10. Filtered 3D image of real scene. (a) Gated image; (b) Depth image; (c) Point cloud

Download full size

View in Article

Table 1. Comparison of simulated pyramid PSNR between trilateral filter and bilateral filter when σ_n= 0.2 m

View table

View in Article

Table 1. Comparison of simulated pyramid PSNR between trilateral filter and bilateral filter when σ_n= 0.2 m

Step/m	σ_d	Original		Bilateral filter		Trilateral filter
Step/m	σ_d	PSNR/dB	PSNR_e/dB	PSNR/dB	PSNR_e/dB	PSNR/dB	PSNR_e/dB
0.2	0.3	35.06	35.06	43.84	41.98	44.47	43.59
0.2	0.4	35.06	35.06	46.51	42.43	47.75	45.71
0.3	0.3	35.06	35.06	43.14	40.41	44.11	42.84
0.3	0.4	35.06	35.06	45.32	39.74	47.14	44.43

Table 2. Comparison of simulated pyramid PSNR between trilateral filter and bilateral filter when σ_n=0.3 m

View table

View in Article

Table 2. Comparison of simulated pyramid PSNR between trilateral filter and bilateral filter when σ_n=0.3 m

Step/m	σ_d	Original		Bilateral filter		Trilateral filter
Step/m	σ_d	PSNR/dB	PSNR_e/dB	PSNR/dB	PSNR_e/dB	PSNR/dB	PSNR_e/dB
0.2	0.3	31.54	31.54	36.79	36.38	36.93	36.81
0.2	0.4	31.54	31.54	39.57	38.85	39.83	39.51
0.3	0.3	31.54	31.54	36.53	35.88	36.81	36.57
0.3	0.4	31.54	31.54	39.13	37.76	39.62	38.99

Table 3. Comparison of trilateral filtering and bilateral filtering on real-scene data

View table

View in Article

Table 3. Comparison of trilateral filtering and bilateral filtering on real-scene data

Targets spacing/m	Noiseσ_n/m	Filterσ_d/m	Original		Bilateral filter		Trilateral filter
Targets spacing/m	Noiseσ_n/m	Filterσ_d/m	PSNR/dB	PSNR_e/dB	PSNR/dB	PSNR_e/dB	PSNR/dB	PSNR_e/dB
0.1	0.13	0.3	33.30	33.49	45.30	42.61	45.43	45.85
	0.13	0.4	33.30	33.49	45.96	42.66	46.24	46.73
	0.24	0.3	27.89	27.85	34.47	34.62	34.61	35.11
	0.24	0.4	27.89	27.85	37.56	37.62	37.67	38.40
0.2	0.12	0.3	33.81	33.09	42.01	37.46	42.66	40.34
	0.12	0.4	33.81	33.09	42.37	37.40	43.28	40.81
	0.27	0.3	27.68	26.92	33.70	32.10	34.03	32.71
	0.27	0.4	27.68	26.92	36.38	34.21	36.92	35.40
0.3	0.13	0.3	33.08	31.79	38.90	33.98	39.32	35.50
	0.13	0.4	33.08	31.79	39.02	33.76	39.80	35.90
	0.27	0.3	27.30	27.00	32.42	30.68	32.72	31.59
	0.27	0.4	27.30	27.00	34.55	31.91	35.02	33.21

Table 4. Comparison of trilateral and bilateral filter for two different plants
View table
View in Article
Table 4. Comparison of trilateral and bilateral filter for two different plants
Target Original Bilateral filter Trilateral filter
PSNR/dB PSNR/dB PSNR/dB
Plant 1 44.09 36.84 59.32
Plant 2 33.55 32.69 38.65

Tools

Get Citation

Copy Citation Text

Chenhao XIA, Xinwei WANG, Liang SUN, Bo SONG, Songtao FAN, Yan ZHOU. Trilateral filter smoothing and denoising algorithm for gated light ranging and imaging[J]. Infrared and Laser Engineering, 2024, 53(8): 20240204

Download Citation

EndNote(RIS)BibTex Plain Text

Set citation alerts for article

Save article for my favorites

Paper Information

Category:

Received: May. 28, 2024

Accepted: --

Published Online: Oct. 29, 2024

The Author Email:

DOI:10.3788/IRLA20240204

Topics

laser devices and laser physics

Lasers and Laser Optics

Laser physics

laser manufacturing

Instrumentation, Measurement and Metrology

Table 1. Comparison of simulated pyramid PSNR between trilateral filter and bilateral filter when σn= 0.2 m

Table 1. Comparison of simulated pyramid PSNR between trilateral filter and bilateral filter when σn= 0.2 m

Table 2. Comparison of simulated pyramid PSNR between trilateral filter and bilateral filter when σn=0.3 m

Table 2. Comparison of simulated pyramid PSNR between trilateral filter and bilateral filter when σn=0.3 m

Table 3. Comparison of trilateral filtering and bilateral filtering on real-scene data

Table 3. Comparison of trilateral filtering and bilateral filtering on real-scene data

Table 4. Comparison of trilateral and bilateral filter for two different plants

Table 4. Comparison of trilateral and bilateral filter for two different plants

Table 1. Comparison of simulated pyramid PSNR between trilateral filter and bilateral filter when σ_n= 0.2 m

Table 1. Comparison of simulated pyramid PSNR between trilateral filter and bilateral filter when σ_n= 0.2 m

Table 2. Comparison of simulated pyramid PSNR between trilateral filter and bilateral filter when σ_n=0.3 m

Table 2. Comparison of simulated pyramid PSNR between trilateral filter and bilateral filter when σ_n=0.3 m