Dual-focal-plane polarizer-based full stokes image acquisition device and method

Abstract

The application discloses a full-Stokes image acquisition device and method based on a double-focal-plane polarizer, relates to the technical field of polarization imaging, and comprises two modified focal-plane imaging polarizers, namely DoFP-1 and DoFP-2, and two quarter-wave plates. The application adopts two quarter-wave plates which are respectively fixed in front of the lens faces of the two modified focal-plane imaging polarizers, so that the modified focal-plane imaging polarizers can detect circular polarization. The two focal-plane imaging polarizers complement each other to solve the problems that the DoFP-2 focal-plane imaging polarizer cannot detect Stokes linear components S0 and S1 and the DoFP-1 focal-plane imaging polarizer cannot detect Stokes linear component S2, so that full-Stokes detection is realized. The application can quickly obtain full-Stokes images of a target in the upper hemispherical space and can be used for target full-polarization characteristic analysis and three-dimensional polarization reconstruction.

Description

Technical Field

[0001] This invention relates to the field of polarization imaging technology, and in particular to a full Stokes image acquisition device and method based on a dual-focus plane polarizer. Background Technology

[0002] Polarization measurements are used to determine the polarization state of light when it interacts with matter. This can be achieved using four Stokes parameters (linear components S0 to S2 and circular components). The focal plane (DoFP) imaging polarizer uses an array of micro-polarizers oriented at 0°, 45°, 90°, and 135°, connected to the pixels of the detector array. It can obtain the first three linear components of the Stokes parameter in a single snapshot, but cannot obtain the circular component. This imaging system is the most compact of all types of polarization imaging systems, possessing irreplaceable advantages such as small size, light weight, and strong environmental adaptability.

[0003] However, circular components Carrying unique information distinct from linear components, these parameters provide essential data for specific applications, including object recognition, underwater and medical diagnostics, and can overcome the dependence on linear components in many imaging or display systems. Therefore, methods for real-time acquisition of all Stokes parameters have received considerable attention in recent years.

[0004] Currently, there are three main methods for real-time acquisition of the full Stokes parameter: The first method combines an unpolarized beamsplitter with a waveplate to achieve multipath measurement. However, the presence of the beamsplitter and optical support limits the back focal distance of the objective lens, and the reflection from the beamsplitter causes multiple sub-images to overlap on the main image. The second method connects a micro-retarder array to a micro-polarizer array and a CCD sensor, but this requires pre-calibrating individual micro-retarders and aligning them with the micro-polarizers. The last method is the metasurface method, which involves reflecting and diffracting different polarized light through a metasurface that diffuses in different directions. By analyzing the diffracted light, the complete Stokes parameter can be obtained. The challenges are diffraction efficiency, wavelength dependence, high process requirements, and cost, making large-scale production difficult. Reference [Y. Lin, H. Huang, Y. Wang, et al., “Image-based polarization detection and material recognition,” Opt. Express 30, 39234 (2022).] demonstrates a single-shot polarization detection system that uses four electrically tunable liquid crystal waveplates, four polarizers, and four camera modules to acquire full Stokes images. Reference [PG Sara, BG Maria, GG Pablo, et al., “Quantitative demonstration of the superiority of circularly polarized light in fog environments,” Opt. Lett. 47, 15 (2022).] proposes a full Stokes image acquisition device and method based on a dual-focal-plane polarizer by placing a rotating retarder in front of the focal plane imaging polarizer to obtain the circular component. Summary of the Invention

[0005] The present invention aims to provide a full Stokes image acquisition device and method based on a dual-focus plane polarizer, so as to acquire full Stokes parameters in real time using the dual-focus plane polarization imaging method.

[0006] To achieve the above objectives, the present invention provides the following technical solution: A full Stokes image acquisition device and method based on a dual-focus plane polarizer includes a first focal plane imaging polarizer, a second focal plane imaging polarizer, a first quarter-wave plate, and a second quarter-wave plate. The first quarter-wave plate is fixed in front of the lens of the first focal plane imaging polarizer. The angle between the fast axis of the first quarter-wave plate and the micro-polarizer of the first focal plane imaging polarizer is 0°. The 45° direction of the micro-polarizer of the first focal plane imaging polarizer is changed to detect right-hand circular polarization, and the 135° direction of the micro-polarizer of the first focal plane imaging polarizer is changed to detect left-hand circular polarization. The second quarter-wave plate is fixed in front of the lens of the second focal plane imaging polarizer. The angle between the fast axis of the second quarter-wave plate and the micro-polarizer of the second focal plane imaging polarizer is 45°. The 0° direction of the micro-polarizer of the second focal plane imaging polarizer is changed to detect right-handed circular polarization, and the 90° direction of the micro-polarizer of the second focal plane imaging polarizer is changed to detect left-handed circular polarization.

[0007] Furthermore, the first quarter-wave plate and the second quarter-wave plate are non-color.

[0008] The full Stokes image acquisition method based on a dual-focal plane polarizer includes the following steps: S1. Perform polarization measurements on the first and second focal plane imaging polarizers respectively to obtain polarization images; S2. Align the polarization images obtained by the first focal plane imaging polarizer and the second focal plane imaging polarizer. S3. Calculate the corresponding polarization vector image, and then calculate the linear polarization degree (DoLP), circular polarization degree (DoCP), polarization degree (DoP), linear polarization angle (AoLP), and circular polarization angle (AoCP).

[0009] Furthermore, in step S1, the measurement process of a single polarizer can be represented as follows:

[0010] in, The light intensity is at four angles. To input the Stokes vector, This is the focal plane (DoFP) matrix that converts input polarization into intensity information; When quarter-wave plates are added in front of the lenses of the two focal plane imaging polarizers, two new focal plane (DoFP) matrices are obtained, namely... and :

[0011] Furthermore, in step S2, a feature detection algorithm is used to detect and match the corresponding feature points of the image, thereby further calculating the rotation matrix and aligning the polarization image.

[0012] Furthermore, in step S2, one of the feature detection algorithms Harris, SIFT, SURF, and KAZE is used to detect and match the corresponding feature points of the image.

[0013] Furthermore, in step S3, the formula for calculating the polarization vector is:

[0014] The formulas for calculating the degree of linear polarization (DoLP), degree of circular polarization (DoCP), degree of polarization (DoP), angle of linear polarization (AoLP), and angle of circular polarization (AoCP) are as follows:

[0015] DoLP, DoCP, and DoP represent the proportions of linearly polarized light, circularly polarized light, and total polarized light to the total light intensity, respectively.

[0016] The principle and beneficial effects of this technical solution: This invention utilizes two focal-plane imaging polarimeters and a fixed-angle retarder to obtain a complete real-time Stokes image. Furthermore, it proposes a solution to the problem of different fields of view in the obtained Stokes images due to different polarimeter positions. The advantages of the dual focal-plane imaging polarimeter are that it avoids too many optical paths and uses algorithms to solve for different fields of view, avoiding the large and complex system associated with beam splitters. Simultaneously, this invention can quickly obtain a full Stokes image of the target in the upper hemisphere, which can be used for target full polarization feature analysis and three-dimensional polarization reconstruction. Attached Figure Description

[0017] Figure 1 (a) is a schematic diagram showing the changes in the actual detection angles of DoFP-1 and DoFP-2 in four directions before and after the addition of a quarter-wave plate; Figure 1 (b) is a schematic diagram of a focal plane polarizer with a quarter-wave plate applied to a multi-angle measurement device; Figure 2 (a) Schematic diagram of a device for verifying the ability of a single focal plane polarizer with a quarter-wave plate added to acquire circular components; Figure 2 (b) is a schematic diagram of the verification results of DoFP-1; Figure 2 (c) is a schematic diagram of the verification results of DoFP-2; Figure 2 (b) and (c) show the polarization state of the laser in front of the quarter-wave plate 2 (QWP2) in the first row; the second row shows the images of the four channels of the focal plane (DoFP); and the third line shows the theoretical values ​​(color bars) and actual values ​​(black dots) of the four channels. Figure 3 For the whole Stokes vector ( The images of DoLP, DoCP, DOP, AoLP and AoCP are calculated by equation (6). Detailed Implementation

[0018] The present invention will now be described in further detail with reference to the accompanying drawings and embodiments: like Figure 1 As shown, it includes a first focal plane imaging polarizer (denoted as DoFP-1), a second focal plane imaging polarizer (denoted as DoFP-2), a first quarter-wave plate, a second quarter-wave plate, and a fixed-angle retarder. The first focal plane imaging polarizer and the second focal plane imaging polarizer are mounted on the fixed-angle retarder. The fixed-angle retarder is used to fix the angle between the first focal plane imaging polarizer and the second focal plane imaging polarizer and to provide a stable observation path. The first quarter-wave plate is fixed in front of the lens of the first focal plane imaging polarizer. The angle between the fast axis of the first quarter-wave plate and the micro-polarizer of the first focal plane imaging polarizer is 0°. The 45° direction of the micro-polarizer of the first focal plane imaging polarizer is changed to detect right-handed circular polarization, and the 135° direction of the micro-polarizer of the first focal plane imaging polarizer is changed to detect left-handed circular polarization. The second quarter-wave plate is fixed in front of the lens of the second focal plane imaging polarizer. The fast axis of the second quarter-wave plate is at an angle of 45° to the micro-polarizer of the second focal plane imaging polarizer. The 0° direction of the micro-polarizer of the second focal plane imaging polarizer is changed to detect right-handed circular polarization, and the 90° direction of the micro-polarizer of the second focal plane imaging polarizer is changed to detect left-handed circular polarization.

[0019] The polarizer consists of two focal plane (DoFP) imaging polarimeters (MER2-502-79U3M Pol) and two lenses (XHF25XA-5 M). The polarimeters utilize a Sony IMX250MYR CMOS sensor, which comprises a set of micro-polarizers facing four different angles. The measurement process of a single polarimeter can be represented as follows:

[0020] In the formula, The light intensity is at four angles. To input the Stokes vector, This is the focal plane (DoFP) matrix that converts input polarization into intensity information.

[0021] DoFP matrix It can detect linear polarization but not circular polarization. To detect linear and circular polarization in real time, two non-color quarter-wave plates (QWP; Thorlabsaqwp10m-580) are fixed in front of the lens of the focal plane imaging polarizer. The first focal plane imaging polarizer is denoted as DoFP-1, and the second focal plane imaging polarizer is denoted as DoFP-2.

[0022] When quarter-wave plates are added in front of the lenses of the two focal plane imaging polarizers, two new focal plane (DoFP) matrices are obtained through the Müller matrix. and Theoretically, it can be expressed as:

[0023] It should be noted that the modified DoFP-1 cannot detect Stokes linear components due to the missing 45° and 135° directions. DoFP-2 cannot detect Stokes linear components due to the lack of 0° and 90° directions. and Therefore, in order to achieve real-time full Stokes detection, two modified focal plane (DoFP) imaging polarizers, DoFP-1 and DoFP-2, are needed to compensate for each other's shortcomings.

[0024] Establish a system to verify the ability of a single modified DoFP to measure linear and circular components, such as... Figure 2 As shown in (a), the He-Ne laser emits linearly polarized light, which is then passed through QWP1 to generate circularly polarized light. PG produces linearly polarized light of equal intensity in all directions. Subsequently, the system generates horizontal, vertical, right-handed, and left-handed polarized light to verify DoFP-1; then it generates 45-degree, 135-degree, right-handed, and left-handed polarized light to verify DoFP-2, as shown in (a). Figure 2 As shown in (b) and (c).

[0025] It should be noted that the images of these four channels are normalized, therefore the actual value of the channel matching the incident light direction is 1. Due to the extinction ratio of the micropolarizer, a channel with a theoretical value of 0 can still detect a small light intensity, and a channel with a theoretical value of 0.5 can detect a slightly larger light intensity, but their consistency does not affect the calculation of the Stokes vector.

[0026] Because the two focal planes (DoFP) are in different positions, the resulting polarization images are not aligned, making it impossible to further calculate the Stokes vector image. Common image registration methods include: 1) calculating the rotation matrix between the two shooting positions using a calibration board; 2) using feature detection algorithms to detect and match corresponding feature points in the image, and then calculating the rotation matrix. The first method requires relatively fixed shooting positions; considering the adaptability of the method, the second method is adopted here.

[0027] Among numerous feature detection algorithms such as Harris, SIFT, SURF, and KAZE, KAZE exhibits a significant advantage in detecting polarimetric images. This is because polarimetric imaging utilizes the acquired polarimetric information to reconstruct and enhance the target, creating a sharp contrast between the target and the background and enhancing image edge information. The KAZE algorithm employs the AOS (Additive Operator Splitting) algorithm for nonlinear diffusion filtering, constructing a stable, arbitrary-step-size nonlinear scale space that replaces the traditional linear Gaussian scale space, thus preserving image edges and details. The expression for the nonlinear scale space obtained by the AOS algorithm is as follows:

[0028] In the formula, Image brightness; Representing each dimension The conductivity matrix of the image; Indicates the step size; It is a dimension The column and column product; It is an image sequence; It is an identity matrix.

[0029] Then, the KAZE algorithm detects feature points by finding local maxima in the Hessian matrix. The formula for calculating the Hessian matrix is:

[0030] in, Rounding according to scale parameters; and It is the first-order derivative of all neighboring points within the retrieval domain. Each pixel is compared with 26 pixels at the same scale and adjacent scales above and below, and the extreme point obtained should be greater than all neighboring points.

[0031] Table 1 shows a comparison of image registration using different feature detection algorithms. The KAZE algorithm detected the most feature points. Furthermore, the first three algorithms use the Brute-Force (BF) matcher, while the last algorithm uses the FLANN (Fast Library for Approximate Nearest Neighbors) matcher. A threshold is set to filter the number of correctly matched feature points; this threshold is determined by the number of feature points without obvious incorrect matches.

[0032] Table 1 Comparison of image registration using different feature detection algorithms

[0033] After solving the problems for different fields of view, the corresponding polarization vector images can be calculated according to equation (3), and then the linear polarization degree (DoLP), circular polarization degree (DoCP), polarization degree (DoP), linear polarization angle (AoLP), and circular polarization angle (AoCP) can be calculated:

[0034] DoLP, DoCP, and DoP represent the proportions of linearly polarized light, circularly polarized light, and total polarized light to the total light intensity, respectively.

[0035] Figure 3 The full Stokes vector and polarization image are calculated by equation (6). When the incident zenith angle is 50°, the observation azimuth angle is 180°, the observation zenith angle is 0~50°, and the interval is 10°.

[0036] By combining the theoretical values ​​of the parameters with the actual values ​​of the image, the display range of grayscale from 0 to 2 is obtained. image. The image displays a color range of -1 to 1. The grayscale range of DoLP, DoCP, and DoP images is 0 to 1. The color ranges of AoLP and AoCP images are −π / 2 to π / 2 and −π / 4 to π / 4, respectively. It can be observed that... The image reflects the overall appearance of the object, becoming brighter as the observation zenith angle increases, reaching its maximum at the specular reflection position (50°). In the image, linear polarization The value is not only greater than that of circular polarization ( Furthermore, the increase is more pronounced, which can be further reflected in DoLP and DoCP images. The main reason is the polarization caused by specular reflection. Therefore, when the observed zenith angle is close to the specular reflection position, DoLP and DoP increase. AoLP and AoCP images are mainly used to reflect surface orientation and highlight the target outline.

[0037] In summary, this invention proposes a full Stokes image acquisition device and method based on a dual-focal plane polarizer, and verifies the feasibility of the method through measurement results of light polarized in different directions. Then, several commonly used registration algorithms are compared, and it is found that KAZE-FLANN is best suited for solving different fields of view using polarized images. Finally, a set of multi-angle fully polarized images is obtained, and the image characteristics are analyzed. This method has great application potential in full polarization detection, analysis of multi-angle polarization characteristics of targets, and polarization-based 3D imaging.

[0038] It should be clarified that the present invention is not limited to the specific configurations and processes described above and shown in the figures. For the sake of brevity, detailed descriptions of known methods are omitted here. In the above embodiments, several specific steps are described and shown as examples. However, the method process of the present invention is not limited to the specific steps described and shown. Those skilled in the art can make various changes, modifications, and additions, or change the order of steps, after understanding the spirit of the present invention.

[0039] In this invention, features described and / or illustrated for one embodiment may be used in the same or similar manner in one or more other embodiments, and / or combined with or in place of features of other embodiments.

Claims

1. A full Stokes image acquisition device based on a dual-focus plane polarizer, characterized in that, include: Fixed-angle retarder, first focal plane imaging polarizer, second focal plane imaging polarizer, first quarter-wave plate and second quarter-wave plate; The first and second focal plane imaging polarizers are mounted on a fixed-angle retarder. The fixed-angle retarder is used to fix the angle between the first and second focal plane imaging polarizers and to provide a stable observation path. The first quarter-wave plate is fixed in front of the lens of the first focal plane imaging polarizer. The angle between the fast axis of the first quarter-wave plate and the micro-polarizer of the first focal plane imaging polarizer is 0°. The 45° direction of the micro-polarizer of the first focal plane imaging polarizer is changed to detect right-hand circular polarization, and the 135° direction of the micro-polarizer of the first focal plane imaging polarizer is changed to detect left-hand circular polarization. The second quarter-wave plate is fixed in front of the lens of the second focal plane imaging polarizer. The angle between the fast axis of the second quarter-wave plate and the micro-polarizer of the second focal plane imaging polarizer is 45°. The 0° direction of the micro-polarizer of the second focal plane imaging polarizer is changed to detect right-handed circular polarization, and the 90° direction of the micro-polarizer of the second focal plane imaging polarizer is changed to detect left-handed circular polarization. The first and second quarter-wave plates are non-color. The measurement process of a single polarizer can be represented as follows: Where I represents the light intensity at the four angles, and S... in For the input Stokes vector, A is the input bias. Vibration is converted into a focal plane (DoFP) matrix that provides intensity information; When quarter-wave plates are added to the imaging polarizers at the two focal planes respectively, two new focal plane (DoFP) matrices are obtained through the Müller matrix, namely A DoFP-1 and A DoFP-2 : ； The first focal plane imaging polarizer is missing in the 45° and 135° directions and cannot detect the Stokes linear component S2. The second focal plane imaging polarizer is missing in the 0° and 90° directions and cannot detect the Stokes linear component S1. The two focal plane imaging polarizers compensate for each other's deficiencies to achieve real-time full Stokes detection.

Images