Research on polarization imaging control system based on LabVIEW

Abstract

With the continuous research and development of remote sensing, it has become the consensus of the industry that spectral and polarization remote sensing can provide more information than ordinary imagery. In particular, it has obvious advantages in the analysis of the changes in “water bodies”, “gases” and “vegetation”, it also can be used for more accurate detection and analysis, so it is increasingly used in remote sensing and military investigations. Based on the study of polarization imaging mechanism, this paper proposes a single polarizer time-lapse imaging method to acquire polarization images, we complete the control of polarization imaging and polarization image processing using LabVIEW, achieve polarization information fusion, and finally improve the contrast between the target and the background.

Keywords

Polarization image virtual instrument image fusion

1. Introduction

According to the electromagnetic theory of light waves, the main characteristics of light waves are: light intensity, wavelength, phase and polarization state. Optical radiation imaging is used to acquire the image by acquiring the light wave characteristics of the target and inverting the target nature parameters. Among them, the polarization property is closely related to the material properties and is the main information parameter to be acquired by remote sensing. In the optical band, whether it is visible light or other spectral bands, different targets have their own respective polarization characteristics. In recent years, the use of this feature to obtain the measured target information, and then the polarization imaging research has become a hot spot.

Polarization imaging has passive work and active work [9]. Passive work has the advantages of good covertness, but the imaging effect and distance are subject to meteorological conditions, the target temperature contrast, sky background illumination and other factors. Active work uses the laser illumination polarization imaging technology. Relative to passive imaging, this technique relies on emitted laser light as a source of illumination, independent of the target’s own radiation and target’s reflection of secondary light sources such as the sun or the moon. Extracting target information by reflecting or scattering photons from the detected target. It overcomes the shortcomings of passive imaging and is not affected by meteorological conditions, target temperature and background illumination. It has wide application prospects into long-distance dark target detection, underwater detection, remote sensing and military target recognition [9].

After years of development, both passive and active working methods based on imaging sensors and working principle of polarization imaging can be broadly divided into five categories [13]: (1) Mechanical rotating polarization optics imaging technology; (2) Sub-amplitude polarization imaging technology; (3) Liquid crystal tunable filter polarization imaging technology; (4) Sub-aperture polarization imaging; (5) Focal plane and channel modulation polarization imaging technology.

In this paper, using a virtual instrument development platform, a rotating polarization polarimetric polarization imaging system for passive imaging of polarization imaging is designed and fabricated. The sensor uses a mechanically rotating polarizer to project the polarization state of the measured target’s information to the CCD camera. The computer uses the computer to combine the polarization image with the extracted image Stokes parameters to mine the polarization information of the image and provide more dimensional goals. Information eventually reaches the goal of improving the detection and identification capabilities of photodetection equipment.

2. Polarization imaging basic principle

The light wave is a transverse wave, its electric vector vibration plane and the propagation direction perpendicular to each other, the electric field vibration direction relative to the propagation direction asymmetry is called polarization, the polarization of light can be defined by the polarization degree $P$ [12]:

$\displaystyle{P}=\frac{I_{\max}-I_{\min}}{I_{\max}+I_{\min}}$ (1)

Define the degree of polarization, $I_{\min}$ and $I_{\max}$ are the minimum and maximum light intensities of a beam of light in two perpendicular polarization directions. Obviously, this definition has limitations in indicating the degree of polarization of circularly polarized or elliptically polarized light, which cannot be effectively distinguished from part of the polarized and natural light. Ideally, a completely smooth surface is reflected by a single light source, and the polarization degree will be maximized at a certain angle. Therefore, in practice, the surface of a relatively smooth object has a high degree of polarization, and if the surface of the object is rough, the degree of polarization will be small.

In the field of optical polarization remote sensing, Stokes vector method is usually used to express the polarization state of light [7]:

$\displaystyle{S}=\left[{\begin{array}[]{l}I\\ Q\\ U\\ V\\ \end{array}}\right]=\left[{\begin{array}[]{l}I_{0^{\circ}}+I_{90^{\circ}}\\ I_{0^{\circ}}-I_{90^{\circ}}\\ I_{+45^{\circ}}-I_{-45^{\circ}}\\ I_{r}+I_{Z}\\ \end{array}}\right]$ (2)

Degree of polarization is

$\displaystyle{P}=\frac{\sqrt{Q^{2}+U^{2}+V^{2}}}{I}$ (3)

Polarization azimuth angle is

$\displaystyle\theta=\frac{1}{2}\text{arctan }(U/Q)$ (4)

In the formula, $I$ is non-polarized light intensity, $Q$ , $U$ respectively represent the linear polarization intensity in two directions, $V$ represents the circularly polarized light intensity, ${I}_{0^{\circ}}$ , ${I}_{90^{\circ}}$ , ${I}_{+45^{\circ}}$ , ${I}_{-45^{\circ}}$ , ${I}_{r}$ , ${I}_{Z}$ stand for linearly polarized light in the direction of 0 ${}^{\circ}$ , 90 ${}^{\circ}$ , $+$ 45 ${}^{\circ}$ , $-$ 45 ${}^{\circ}$ of an ideal polarizer placed in the light propagation path and circularly polarized light intensity of left-hand (z) and right-hand circularly polarized light (r).

When the beam interacts with the material, the four Stokes parameters of the outgoing beam are linearly related to the four Stokes parameters of the incident beam. Writing in matrix form is

$\displaystyle{S}_{\textit{out}}=MS_{\textit{in}}$ (5)

$M$ is known as the Miller matrix, which is a 4 $\times$ 4 matrix, it represents the material properties and orientation [10]. The Miller matrix is an ideal way to describe the optical changes to the beam (including light attenuation and depolarization). If the polarization information of the incident light is completely known, which means the feature of the bidirectional reflection of the Stoke vector is known, then transforming it by the Miller matrix, the output Stokes vector after the light beam passes through the polarizer can be obtained. Through the conversion of the Miller matrix, the light intensity of different polarization azimuths of the light beam after passing through the polarizer can be calculated to establish the model to realize the mutual conversion between the directional reflection and the multi-angle polarization reflection.

Considering that the Stokes parameter is a quaternion vector, at least four equations are needed to solve for the complete Stokes parameter of a target point to be measured. When incident light passes through an ideal polarizer, the Miller matrix of an ideal polarizer that forms a p-angle with the reference direction is

$\displaystyle{M}_{p}=\left[{{\begin{array}[]{*{20}c}1&{\cos 2\theta}&{\sin 2% \theta}&0\\ {\cos 2\theta}&{\cos^{2}2\theta}&{\cos 2\theta\sin 2\theta}&0\\ {\sin 2\theta}&{\cos 2\theta\sin 2\theta}&{\sin^{2}2\theta}&0\\ 0&0&0&0\\ \end{array}}}\right]$ (6)

The Stokes parameter $S_{\textit{out}}$ of the outgoing light is

$\displaystyle{S}_{\textit{out}}=\left[{{\begin{array}[]{*{20}c}{S^{\prime}_{0}% }\\ {S^{\prime}_{1}}\\ {S^{\prime}_{2}}\\ {S^{\prime}_{3}}\\ \end{array}}}\right]=M_{P}\left[{{\begin{array}[]{*{20}c}{S_{0}}\\ {S_{1}}\\ {S_{2}}\\ {S_{3}}\\ \end{array}}}\right]$ (7)

The intensity of transmitted light can be derived as

$\displaystyle I=S^{\prime}_{0}=\frac{1}{2}(S_{0}+\cos 2\theta\cdot S_{1}+\sin 2% \theta\cdot S_{2})$ (8)

In this way, three independent equations can be obtained in 0 ${}^{\circ}$ , 90 ${}^{\circ}$ and 45 ${}^{\circ}$ directions so as to deduce the complete polarization information of the target to be measured.

Light polarization state is divided into non-polarization, circular polarization, linear polarization, partial polarization and elliptical polarization [15]. In order to reverse the multi-angle polarization reflection by using two-way reflection, the polarization state and the light intensity of these five polarization states after passing through the polarizer can be obtained. According to Marius’s law, the intensity of two polarized light through the analyzer, the transmitted light intensity is

$\displaystyle{I}=I_{0}\cos^{2}\alpha$ (9)

Among them, $\alpha$ is the angle between the polarization direction of the analyzer and the polarization direction of the incident polarized light. A beam of light through the polarizer after the light intensity, unpolarized light, circularly polarized light becomes half of the original and linearly polarized light, partially polarized light and elliptically polarized light is a relation of $\alpha$ parameter, and then you can reverse the two polar polarization distribution function.

3. Polarization imaging system structure

A typical polarization imaging system consists of a rotating polarizer (mechanical rotating polarizer), an optical lens, a CCD camera and a polarization analyzer, as shown in Fig. 1. According to different polarization methods, the polarization imaging system can be divided into time-sharing polarization imaging system and synchronous polarization imaging system [11], this article uses the time-sharing polarization imaging system.

Figure 1.

Polarization imaging principle.

The polarizer is a film made by artificial method, and special selective crystal particles with strong selective absorption are regularly arranged in the transparent adhesive layer, which allows the light transmitted through the polarization direction to absorb the light with its vertical vibration, and has dichroism. Therefore, after the natural light passes through the polarizer, the transmitted light becomes substantially plane polarized light. Obtaining an angle image requires the use of a polarizer. As the rotating polarizer adopts a mechanical rotating structure, the stepping motor is used to control the rotating angle of the deflector. Polarization analyzer takes advantage of the software methods to analyze and process the image information collected by CCD camera. Imaging circuit block diagram shown in Fig. 2.

Figure 2.

Polarization imaging circuit block diagram.

3.1 Polarization analyzer

The measurement of light polarization can be done with the Stokes parameter, measurement accuracy depends on the space and time resolution of the measurement. According to the response speed of the detection system, the polarization analysis methods are mainly divided into two categories: one is the polarization modulation method, the introduction of the polarizer and the phase retarder in the optical path to be measured; the other is the sub-amplitude, sub-aperture or sub- Focal plane imaging method.

3.2 System calibration method

Calibration is the use of the known parameters of the polarized light measurements to obtain the matrix of the instrument. Each working wavelength need only be calibrated once, and after calibration the instrument matrix will be stored for later use in the measurement of the polarization state. The system will be placed in the standard state of polarization measuring instrument at the CCD camera at a rotational angle control of the rotating polarizer horizontal, vertical, 45 ${}^{\circ}$ , $-$ 45 ${}^{\circ}$ four sets of polarizer, and the reading is stored. The calibration using blue and green light source.

3.3 Rotary polarizer control

In this paper, the rotating polarizer is PRM1/MZ8 rotating polarizer manufactured by THORLABS company, as shown in Fig. 3.

Figure 3.

Rotary polarizer.

The rotating polarizer is controlled by a stepping motor. The principle and procedure of the stepping motor control are shown in Fig. 4.

Figure 4.

Stepper motor control block diagram.

4. Polarization imaging system software and image fusion design

4.1 Polarization imaging system software design

In this paper, the final PC platform is built by designing the program based on LabVIEW platform for image acquisition, Stokes image calculation, polarization image preprocessing and image fusion. LabVIEW is a virtual instrument simulation application with multiple functions. LabVIEW uses a graphical editing G language program, and the resulting program is in the form of a block diagram. LabVIEW has a wealth of library functions for signal processing, real-time control, and data acquisition. Based on the above characteristics of LabVIEW, it is used to write software to realize the driving control of the rotating polarizer, subsequent image acquisition and image processing.

Polarimetric imaging system software block diagram is shown in Fig. 5, which divided into four parts: image acquisition, image preprocessing, polarization image computing (Stokes parameter extraction) and image fusion.

Figure 5.

Polarization image acquisition system design block diagram.

4.2 Image acquisition part

The image acquisition is the first step in the entire system. It is mainly to make the image of polarized light which contains measured polarization information on the CCD camera by adjusting the rotating polarizer and the Stokes parameters at 0 ${}^{\circ}$ , 90 ${}^{\circ}$ , $+$ 45 ${}^{\circ}$ and $-$ 45 ${}^{\circ}$ can be read. Read the block diagram shown in Fig. 6.

Figure 6.

Get S0 image flow chart.

The polarization degree image DOLP and the polarization angle image AOP are obtained by calculation. Polarization parameter images are mainly based on the Eqs (6)–(8) derived Stokes vector four parameters of the calculation equation, through the joint solution can be calculated S0, S1, S2, and then get the polarization image and polarization angle of the image.

Taking the S0 image as an example, the process of obtaining S0, S1 and S2 images is illustrated. The S0 image is obtained by adding the 0 ${}^{\circ}$ image to the 90 ${}^{\circ}$ image. By reading the 0 ${}^{\circ}$ image and the 90 ${}^{\circ}$ image pixel, the corresponding pixel points are added according to each line, and then the S0 image is obtained.

This article uses the corner of the laboratory as a scenario and experimented with the LabVIEW platform, the results are shown in Fig. 7. S0, S1, S2, DOLP and AOP images are obtained after the images are acquired, where S1 and S2 contain the more abundant polarization information in the target object. DOLP and AOP are more obvious description of the target contour. Using the results obtained, the next step based on wavelet transform image fusion.

Figure 7.

Polarization image display.

4.3 Polarization analyzer

Compared with traditional optical imaging, polarized target imaging can provide more information about the surface roughness, texture, physical and chemical properties of materials, but it is easy to lose the details of the image information and does not fully reflect the polarization information of the target reflected light. By using information redundancy and complementarity between polarized images, the obtained image polarization information and intensity information are fused to enhance the contrast of the image and improve the original target information.

Firstly, obtain the P image though fusion the polarization degree image and the polarization angle image, then fusion polarization information. Secondly, the P image and the S0 image are read for wavelet transformation, then fusion the polarization information and the intensity information. Figure 8a and b are the flow charts of P image and wavelet transform respectively.

Figure 8.

(a) Find P image flow chart; (b) Find Y image flow chart.

5. Analysis and evaluation of experimental results

The actual effect after fusion is shown in Fig. 9, and the P image is compared with the S0 image as shown in Fig. 10. The S0 image contains intensity information, the P image contains polarization information, and the Y image is a fused image. The result shows that the details of the image are presented.

Figure 9.

Wavelet transform fusion image display.

Figure 10.

S0 image, polarization fusion image and wavelet transform image display.

To measure the quality of fused images, qualitative and quantitative analysis must be performed to determine whether they meet the requirements or reflect the merits of the algorithm. For the experimental results in this paper, subjective and objective analysis methods are used respectively. The subjective evaluation method is simple and intuitive, relying solely on human visual characteristics to quickly evaluate the quality of the image. The disadvantage is that the method is one-sided, has a strong subjectivity and uncertainty, and cannot stand repeated inspections, when the difference between the fusion images are not obvious or the observation object and the objective environment change, subjective evaluation methods are not given to make an accurate judgment. Therefore, a set of objective quality evaluation criteria is needed. In practice, we often use a combination of subjective and objective evaluation methods for quality assessment.

For the subjective analysis, as shown in Fig. 10, the serious reflective window frame and the flowerpot in the original image can be clearly observed, and after the image processing, the detail information is obviously increased, and the outline and the window frame of the flowerpot are presented, achieve the desired effect. In the objective analysis method, the standard deviation, gradient value and entropy of the original image and the fused image are calculated respectively. Specific calculation of indicators as shown in Table 1.

Table 1

Fusion evaluation index

	Standard deviation	Gradient value	Entropy
Original image	50.8012	0.978082	6.03528
Fusion image	95.9269	3.85322	6.94574

As can be seen from Table 1, the standard deviation of the fused image is significantly higher than the standard deviation of the original image, indicating that the gray level of the image is dispersed and the contrast of the image is improved. After fusion, the gradient value of the image is also higher than that of the original image, indicating that the image has more layers and the contrast and texture details also increase. The entropy of the fused image is higher than that of the original image, which indicates that the fused image contains more information. Based on the above subjective analysis and objective analysis, it is shown that the contrast effect of the fused image is better than the original image.

6. Conclusion

Based on LabVIEW platform, we design and fabricate a polarization automatic imaging system for the rotating polarizer. By analyzing the detailed information in the S0 image and the edge information in the DOLP and AOP images, a multi-resolution wavelet fusion algorithm based on polarization parameters is proposed, and the background information is weakened by using a combination of polarization imaging and intensity image processing techniques. The numerical values of DOLP and AOP correspond to the intensity space. Experimental results show that the algorithm greatly weakens the background component information, improves the clarity of the target image contour, achieve the improvement of the target background contrast.

References

D.M.

Liu

and Niu

G.C.

, Visible light polarization imaging system detects low contrast targets [J], Laser & Optoelectronics Progress (6) (2017).

Zhang

D.X.

Yuan

B.H.

and Zhang

J.J.

, Research on polarization image fusion using tetrolet transform [J], Applied Mechanics & Materials 713–715 (2014), 1859–1862.

Zhang

D.X.

Wang

H.H.

and Feng

, Fusion of polarization image based on curvelet transform [J], Applied Mechanics and Materials 3138(536) (2014).

Dai

, Polarization Detection and Imaging System Research and Optimization [D]. Graduate University of Chinese Academy of Sciences (Changchun Institute of Optics, Fine Mechanics and Physics), 2015.

Hou

J.F.

T.X.

Wang

D.G.

et al., Study on compensation method of beam deviation in division of time imaging polarimetry [J], Acta Physica Sinica 64(6) (2015), 60701–060701.

Liang

et al., Research and design of prefixing objective with large aperture in imaging spectrometer [J], Acta Optica Sinica (2) (2015), 268–275.

Yang

et al., Principle and simulation of a spatially modulated full stokes parameters polarization imaging system [J], Acta Optica Sinica 35(2) (2015).

Liu

, Research on Key Control Techniques in Spectral Total Polarization Imaging System [D], North University of China, 2017.

S.J.

Jiang

H.L.

Zhu

J.P.

et al., Development status and key technologies of polarization imaging detection [J], Chinese Optics 6(6) (2013), 803–809.

10.

Liu

Wang

Z.B.

Xue

et al., New method for obtaining stokes parameters of hyperspectral polarization imaging (English) [J], Spectroscopy and Spectral Analysis (12) (2017).

11.

Wang

Y.H.

, Dual-CCD simultaneous polarization imaging system and its polarization information processing software system [D], Beijing Institute of Technology, 2016.

12.

Zhang

Y.M.

, Research on Stokes Polarization Imaging Technology [D], Nanjing University of Science and Technology, 2015.

13.

Wang

Liu

Jin

et al., Polarization information processing and software system design for simultaneously imaging polarimetry [J], Proceedings of SPIE – The International Society for Optical Engineering, 2015, 9622.

14.

Liu

, Research on Target Detection Technology Based on Visible Light Polarization Imaging [D], Graduate School of the Chinese Academy of Sciences (Institute of Optoelectronic Technology), 2016.

15.

Liu

, Visible polarization image fusion with non-sampling Shearlets, in: The 2th Seminar on Novel Detection Technology and Application, Proc. of SPIE Vol. 9795-145, 2015.