Self-Adaptive Image Segmentation Algorithm for Polarization Navigation under Complex Scenes

Fig. 2. Diagrams of polarization camera structure and algorithm processing of 90° polarization image. (a) Structure of polarization camera; (b) algorithm processing of 90° polarization image

Download full size

Fig. 3. Rayleigh scattering model

Download full size

Fig. 4. Flow chart of abnormal pixel segmentation for sun relative azimuth calculation

Download full size

Fig. 5. Flow chart of snake optimization (SO) algorithm

Download full size

Fig. 6. Experimental setup

Download full size

Fig. 7. Sky polarization parameter images under different baffle occlusion rates

Download full size

Fig. 8. Heading angle errors under different baffle occlusion rates

Download full size

Fig. 9. Errors of heading angle under different occlusion rates of trees and buildings

Download full size

Fig. 10. Sky polarization parameter images under different occlusion rates of buildings and trees

Download full size

Table 1. Comparison of test results between two algorithms
View table
Table 1. Comparison of test results between two algorithms
Occlusionrate /% Algorithm Optimal threshold Iteration PSNR Time /s
35 Otsu 75 256 7.2061 0.0076
SO-Otsu 75 4 7.2061 0.0045
50 Otsu 43 256 6.2887 0.0076
SO-Otsu 43 4 6.2887 0.0045
83 Otsu 53 256 9.4736 0.0078
SO-Otsu 53 4 9.4736 0.0044

Table 2. Maximum error and root mean square error of heading angle under obstruction of baffle
View table
Table 2. Maximum error and root mean square error of heading angle under obstruction of baffle
Occlusion rate /% Maximum error /（°） Root mean squareerror /（°）
Undivided Divided Undivided Divided
25 1.78 1.56 1.10 0.66
40 2.20 2.36 1.35 0.79
55 3.80 2.59 2.32 0.80
70 4.09 1.61 2.59 0.90
85 7.83 2.25 5.42 0.94

Table 3. Maximum error and root mean square error of heading angle under obstruction of trees and buildings

View table

Table 3. Maximum error and root mean square error of heading angle under obstruction of trees and buildings

Occlusion rate /%	Maximumerror /（°）		Root mean squareerror /（°）
Occlusion rate /%	Undivided	Divided	Undivided	Divided
17	181.14	1.25	78.17	0.60
35	180.06	1.51	107.95	0.72
45	4.13	1.56	1.49	0.73
67	16.55	1.58	6.04	0.75
79	9.76	1.27	5.78	0.52
83	35.90	1.33	12.83	0.64

Tools

Get Citation

Copy Citation Text

Yucong Zhou, Chao Ye, Zihan Lin, Chunlian Zhan, Han Gao. Self-Adaptive Image Segmentation Algorithm for Polarization Navigation under Complex Scenes[J]. Acta Optica Sinica, 2024, 44(19): 1910002

Download Citation

EndNote(RIS)BibTex Plain Text

Set citation alerts for article

Save article for my favorites

Paper Information

Category: Image Processing

Received: Mar. 15, 2024

Accepted: May. 20, 2024

Published Online: Oct. 12, 2024

The Author Email: Gao Han (gaohan@cjlu.edu.cn)

DOI:10.3788/AOS240735

CSTR:32393.14.AOS240735

Topics

laser devices and laser physics

Lasers and Laser Optics

Laser physics

laser manufacturing

Instrumentation, Measurement and Metrology

Table 1. Comparison of test results between two algorithms

Table 1. Comparison of test results between two algorithms

Table 2. Maximum error and root mean square error of heading angle under obstruction of baffle

Table 2. Maximum error and root mean square error of heading angle under obstruction of baffle

Table 3. Maximum error and root mean square error of heading angle under obstruction of trees and buildings

Table 3. Maximum error and root mean square error of heading angle under obstruction of trees and buildings