Method for constructing polarization calibration precision model of focal plane visible light polarization detector and application

Abstract

The invention discloses a method for constructing a polarization calibration precision model of a focal plane visible light polarization detector and application. According to the method, a comprehensive matrix form is formed by combining response models of four polarization channels of the detector, and a final target model is obtained through verification and continuous correction. The matrix method can more systematically describe the mutual relation among the channels, so that quantization characteristics such as the degree of polarization (DoLP) can be more accurately calculated through a plurality of response values, and higher calculation efficiency and higher adaptability are achieved.

Description

Technical Field

[0001] The invention relates to the field of photoelectric detection technology, and in particular to a polarization calibration accuracy model method and application of a split-focal-plane visible light polarization detector. Background Art

[0002] Compared with traditional optical imaging, polarization imaging, as an important supplement to optical imaging methods, can not only obtain the intensity information of the target, but also obtain the polarization information of the target. Currently, polarization imaging is widely used in target detection and identification, and has great application value in military and civilian applications.

[0003] Polarization imaging methods include division of time (DoT), division of amplitude (DoAM), division of aperture (DoAP) and division of focal plane (DoFP). The focal plane imaging method closely couples each polarization direction device with a single pixel, greatly reducing the volume, weight and power consumption of the system, and preliminarily solving the problem of system miniaturization and low cost. However, due to its system structure, the focal plane imaging method has polarization information errors, so many researchers at home and abroad have conducted research on the calibration of focal plane polarization detectors.

[0004] In 2020, Wu Rong and his team constructed an experimental system for verifying the polarization pixel response model of the focal plane polarization detector. Based on the previous research, they discussed the pixel response model and calibrated the polarization performance parameters. The proposed model can accurately describe the response characteristics of the focal plane polarization detector. In 2022, Bai Yang proposed a method based on measuring the phase difference of the curve fitted by the average value of the pixel output in each polarization main axis direction in the focal plane polarization image sensor. This method has high calibration accuracy and strong operability, and can effectively complete the calibration task of the polarization main axis direction of the focal plane polarization image sensor. In 2022, Zhao Yonggang analyzed the imaging error of the focal plane polarization detector and numerically simulated the related factors causing the error. The dependence of the imaging error of the focal plane polarization detector on each influencing factor was obtained. In 2023, Wang Wei and others theoretically analyzed the influence of relative transmittance and polarization azimuth on the accuracy of polarization degree measurement, and verified the analysis results through experiments. The following are the main problems of the focal plane polarization detector:

[0005] Traditional polarization calibration methods have many limitations. They usually assume that the detector responds uniformly and linearly in all directions, without considering the impact of the detector's non-uniform response and nonlinear factors (such as light intensity, polarization angle, transmittance, etc.) on the detector's output signal. In addition, many traditional methods rely on simplified response models and ignore the complexity of actual detectors, such as the nonlinearity of the response curve and the spatial variation of the photoelectric conversion efficiency. Therefore, these methods are difficult to accurately reflect the true response of the detector at different positions or angles in practical applications, and the polarization calibration accuracy is low, which is easy to introduce errors. Especially in high polarization or complex environments, the simple linear relationship between the degree of polarization (DoLP) and the angle of polarization (AoP) assumed by traditional methods cannot accurately describe the response of the actual detector.

[0006] Based on the optical characteristics of the sub-focal plane visible light polarization detector, the present invention introduces key parameters such as the detector's photoelectric conversion coefficient, electronic gain, detection efficiency non-uniformity and nonlinear index, and establishes a polarization response model of the sub-focal plane visible light polarization detector. The model includes the Mueller matrix of the detector, and constructs a polarization calibration accuracy model of the visible light polarization detector. The calibration accuracy model can provide a more accurate optical system characterization, help to better compensate for the errors of the optical system, improve the accuracy and stability of the sub-focal plane visible light polarization detector, and provide a theoretical basis for error correction of the sub-focal plane type polarized visible light detector. Summary of the invention

[0007] The technical problem to be solved by the present invention is how to combine relevant physical parameters to provide a more comprehensive and accurate calibration model.

[0008] The present invention solves the above technical problems through the following technical means:

[0009] The method for constructing a polarization calibration accuracy model of a focal plane visible light polarization detector comprises the following steps:

[0010] S01. Construct a polarization calibration accuracy model for visible light polarization detectors:

[0011]

[0012] Among them, C1 to C4 correspond to the response values ​​of the four polarization directions respectively; τ1 to τ4 represent the relative transmittance of P1 to P4; θ k represents the polarization azimuth angle; α and R are the photoelectric conversion coefficient and the electronic gain respectively; ρ(x,y) is the detection efficiency non-uniformity at position (x,y), that is, the response difference at different positions or directions; γ is a nonlinear index, which indicates the nonlinear degree of the response; c is the dark current response, which indicates the signal generated by the detector due to noise and other reasons when there is no light intensity incident;

[0013] S02. Calculation of polarization calibration accuracy model parameters: Experimentally measure the responses of light sources with different known polarization states in the visible light polarization detector, and use these measurement data to infer the parameters of the polarization calibration accuracy model;

[0014] S03. Verify the polarization calibration accuracy model after the parameter calibration in step S02 to obtain the polarization error of the visible light polarization detector;

[0015] S04, updating the polarization calibration accuracy model according to the polarization degree error obtained in step S03, thereby optimizing the performance of the visible light polarization detector.

[0016] Furthermore, the construction process of the polarization calibration accuracy model in step S01 is as follows:

[0017] Step 1: Construct the Mueller matrix of the visible light polarization detector

[0018] Assuming that the incident light beam corresponding to a certain pixel can be decomposed into mutually orthogonal p light and s light, the surface composed of p and s is parallel to the CMOS photosensitive surface, the transmittance of the microlens to p light and s light is η1 and η2 respectively, and the linear bidirectional attenuation of the microlens is ε, then

[0019]

[0020] Thus, the Mueller matrix M of the microlens array in the visible light polarization detector is lens :

[0021]

[0022] Among them: η1 and η2: are the transmittance coefficients of the microlens array for two orthogonal polarized lights, reflecting the intensity change of polarized light; ε: is the cross-coupling effect, that is, the interaction between p-light and s-light; Used to ensure that the total intensity of light does not change with changes in polarization; and Determines the intensity attenuation of light and the propagation of polarized light; since η1 and η2 are the transmittances of p-light and s-light respectively, their square sum reflects the total transmittance; and It involves the cross-coupling effect between p-light and s-light, and ε is the difference in transmittance between polarized lights; To deal with the nonlinear attenuation of polarized light; Adjusted the propagation intensity of polarized light;

[0023] Ignoring the polarization effect of CMOS, the Mueller matrix M of the visible light polarization detector includes the Mueller matrix of the polarizer and the Mueller matrix M of the microlens lens ;

[0024]

[0025] The micro-polarization array is approximately regarded as an ideal linear polarization device, and the Mueller matrix of the θ polarization azimuth

[0026]

[0027] In an optical system, when light passes from one element to the next, the Mueller matrix of the latter element will act on the polarization state of the light output by the previous element. The microlens array first acts on the incident light to change the propagation direction and polarization state of the light; then, the micropolarizer array will act on the polarization state of the light output by the microlens, further changing the polarization characteristics of the light; so the Mueller matrix M of the detector is:

[0028]

[0029] After the incident light passes through the microlens and micropolarizer array, the Stokes vector reaching the CMOS can be represented by the Mueller matrix M and the Stokes vector of the incident light, and the physical quantity that the CMOS can detect is the total light intensity in the Stokes vector; the first term of the light Stokes vector is the light intensity, so the light intensity detected by the CMOS is I CMOS It is expressed as:

[0030]

[0031] After passing through the microlens and micropolarizer array, the incident light finally reaches the CMOS detector;

[0032] Step 2: Construct a visible light polarization detector response model

[0033] The response count value C of the visible light polarization detector and the incident light intensity I CMOS The relationship is:

[0034]

[0035] Among them: α is the photoelectric conversion coefficient, which indicates the efficiency of converting photons into electronic signals in the detector; R is the electronic gain, which indicates the electronic amplification factor of the detector; ρ(x,y) is the detection efficiency non-uniformity at position (x,y), that is, the response difference at different positions or directions; γ is a nonlinear index, which indicates the nonlinear degree of the response; c is the dark current response, which indicates the signal generated by the detector due to noise and other reasons when there is no light intensity incident;

[0036] Substituting the light intensity of the visible light polarization detector into the response model of the visible light polarization detector, we finally get the expression of the response model of the visible light polarization detector:

[0037]

[0038] After passing through the microlens and micropolarizer array, the incident light finally reaches the CMOS detector. Here, the Stokes vector of the light will be affected by the polarization filter array. Specifically, the CMOS detector can detect the total light intensity part of the modulated Stokes vector. The specific form provided in the formula is:

[0039]

[0040] The response value of each channel can be expressed by the following formula:

[0041]

[0042] Among them, τ k is the transmittance of each channel, θ k is the polarization azimuth corresponding to each channel;

[0043] According to the formula:

[0044] Q=DoLP·I·cos(2AoP)

[0045] U=DoLP·I·sin(2AoP)

[0046] Substituting Q and U into the equation:

[0047]

[0048] The response models C(x, y) of the four polarization channels are combined into one matrix to obtain the polarization calibration accuracy model of the visible light polarization detector.

[0049] Furthermore, the specific process in step S02 is: first obtain the relevant parameters of the relative transmittance, polarization azimuth, photoelectric conversion coefficient, electronic gain, detection efficiency non-uniformity and nonlinear index of the visible light polarization detector, and then substitute the above 7 kinds of data into the polarization calibration accuracy model to obtain the calibrated polarization calibration accuracy model.

[0050] Furthermore, the verification method of the polarization calibration accuracy model in step S03 is:

[0051] Start the visible light polarization detector, adjust the adjustable polarization light source to output 8 states, with polarization degrees of 0, 5%, 10%, 15%, 20%, 25%, 30% and 45% respectively; each polarization channel of the visible light polarization detector measures the 8 states 10 times respectively;

[0052] Calculate the reference value of the measured polarization degree under different bands and different polarization conditions and the detection unit measurement value Statistical deviations;

[0053] The measured value of the visible light polarization detector is inverted by the following formula.

[0054]

[0055] The deviation between the measured value of the visible light polarization detector and the theoretical reference value of the adjustable polarization light source VPOLS-II is compared by the following formula, and the statistical deviation is calculated.

[0056]

[0057] Calculate E according to the universal comparison standard for the qualification assessment of measurement uncertainty k ;

[0058]

[0059] in Calculated at 0.5%, Calculated at 2%.

[0060] Furthermore, the polarization calibration accuracy model error correction in step S04 includes:

[0061] Step 1, comparing the measured polarization degree with the theoretical polarization degree and calculating the error;

[0062] Step 2: Based on the characteristics of the error, use the error model to describe the error and find the appropriate error correction parameters;

[0063] Step 3, using the error model to adjust the polarization calibration accuracy model of the visible light polarization detector;

[0064] Step 4, re-measure the polarized light and calculate whether the corrected polarization calibration accuracy model can accurately restore the known polarization degree; if the error is reduced, it indicates that the correction is effective; if the error is still large, re-check the error model or correction method.

[0065] Furthermore, the error model is a least squares method.

[0066] The present invention also provides a polarization calibration accuracy model for a focal plane visible light polarization detector. The polarization calibration accuracy model expression is as follows:

[0067]

[0068] Furthermore, the method for constructing the polarization calibration accuracy model includes:

[0069] Step 1: Construct the Mueller matrix of the visible light polarization detector

[0070] Assuming that the incident light beam corresponding to a certain pixel can be decomposed into mutually orthogonal p light and s light, the surface composed of p and s is parallel to the CMOS photosensitive surface, the transmittance of the microlens to p light and s light is η1 and η2 respectively, and the linear bidirectional attenuation of the microlens is ε, then

[0071]

[0072] Thus, the Mueller matrix M of the microlens array in the visible light polarization detector is lens :

[0073]

[0074] Among them: η1 and η2: are the transmittance coefficients of the microlens array for two orthogonal polarized lights, reflecting the intensity change of polarized light; ε: is the cross-coupling effect, that is, the interaction between p-light and s-light; Used to ensure that the total intensity of light does not change with changes in polarization; and Determines the intensity attenuation of light and the propagation of polarized light; since η1 and η2 are the transmittances of p-light and s-light respectively, their square sum reflects the total transmittance; and It involves the cross-coupling effect between p-light and s-light, and ε is the difference in transmittance between polarized lights; To deal with the nonlinear attenuation of polarized light; Adjusted the propagation intensity of polarized light;

[0075] Ignoring the polarization effect of CMOS, the Mueller matrix M of the visible light polarization detector includes the Mueller matrix of the polarizer and the Mueller matrix M of the microlens lens ;

[0076]

[0077] The micro-polarization array is approximately regarded as an ideal linear polarization device, and the Mueller matrix of the θ polarization azimuth

[0078]

[0079] In an optical system, when light passes from one element to the next, the Mueller matrix of the latter element will act on the polarization state of the light output by the previous element. The microlens array first acts on the incident light to change the propagation direction and polarization state of the light; then, the micropolarizer array will act on the polarization state of the light output by the microlens, further changing the polarization characteristics of the light; so the Mueller matrix M of the detector is:

[0080]

[0081] After the incident light passes through the microlens and micropolarizer array, the Stokes vector reaching the CMOS can be represented by the Mueller matrix M and the Stokes vector of the incident light, and the physical quantity that the CMOS can detect is the total light intensity in the Stokes vector; the first term of the light Stokes vector is the light intensity, so the light intensity detected by the CMOS is I CMOS It is expressed as:

[0082]

[0083] After passing through the microlens and micropolarizer array, the incident light finally reaches the CMOS detector;

[0084] Step 2: Construct a visible light polarization detector response model

[0085] The response count value C of the visible light polarization detector and the incident light intensity I CMOS The relationship is:

[0086]

[0087] Among them: α is the photoelectric conversion coefficient, which indicates the efficiency of converting photons into electronic signals in the detector; R is the electronic gain, which indicates the electronic amplification factor of the detector; ρ(x,y) is the detection efficiency non-uniformity at position (x,y), that is, the response difference at different positions or directions; γ is a nonlinear index, which indicates the nonlinear degree of the response; c is the dark current response, which indicates the signal generated by the detector due to noise and other reasons when there is no light intensity incident;

[0088] Substituting the light intensity of the visible light polarization detector into the response model of the visible light polarization detector, we finally get the expression of the response model of the visible light polarization detector:

[0089]

[0090] After passing through the microlens and micropolarizer array, the incident light finally reaches the CMOS detector. Here, the Stokes vector of the light will be affected by the polarization filter array. Specifically, the CMOS detector can detect the total light intensity part of the modulated Stokes vector. The specific form provided in the formula is:

[0091]

[0092] Polarization detectors generally use multiple polarization directions (such as P1 to P4) to measure polarization information at different angles. The response value of each channel can be expressed by the following formula:

[0093]

[0094] Among them, τ k is the transmittance of each channel, θ k is the polarization azimuth corresponding to each channel;

[0095] According to the formula:

[0096] Q=DoLP·I·cos(2AoP)

[0097] U=DoLP·I·sin(2AoP)

[0098] Substituting Q and U into the equation:

[0099]

[0100] The response models of the four polarization channels are combined into a matrix form, thus obtaining the polarization calibration accuracy model of the visible light polarization detector.

[0101] Furthermore, the specific process of parameter calibration of the polarization calibration accuracy model is: firstly obtain the relevant parameters of the relative transmittance, polarization azimuth, photoelectric conversion coefficient, electronic gain, detection efficiency non-uniformity and nonlinear index of the visible light polarization detector, and then substitute the above 7 kinds of data into the polarization calibration accuracy model to obtain the calibrated polarization calibration accuracy model.

[0102] The present invention also provides a method for calculating the degree of polarization DoLP of a visible light polarization detector using the above-mentioned polarization calibration accuracy model, and the calculation formula is:

[0103]

[0104] The advantages of the present invention are:

[0105] The present invention combines the response models of the four polarization channels of the detector to form a comprehensive matrix form. This matrix method can more systematically describe the relationship between the channels, so that quantitative characteristics such as the degree of polarization (DoLP) can be more accurately inferred through multiple response values, with higher calculation efficiency and stronger adaptability.

[0106] The response model of the present invention not only takes into account the light intensity signal, but also integrates factors such as the polarization direction and degree of polarization (DoLP). In particular, it introduces key parameters such as the detector's photoelectric conversion coefficient (α), electronic gain (R), detection efficiency non-uniformity (ρ(x,y)) and nonlinear index (γ), which can more accurately reflect the actual detector performance and response than traditional methods.

[0107] In the response model, the nonlinear exponent (γ) and detection efficiency inhomogeneity (ρ(x,y)) are used to compensate for the nonlinear response and spatial position differences of the detector. This allows the method to adapt to the characteristics of different detectors rather than just being based on linear assumptions, thereby enhancing the robustness of the calibration method.

[0108] The DoLP calculation formula proposed in this invention integrates the differences of the four polarization channels, can more accurately measure and evaluate the degree of polarization, and adapt to complex detection environments. Traditional methods usually estimate DoLP through linear combinations of four channels or simple geometric relationships, while the model of this patent can dynamically consider various factors (such as nonlinearity and inhomogeneity), improving the accuracy of polarization calculation.

[0109] In summary, the present invention proposes a more comprehensive and accurate calibration method by innovating the traditional polarization calibration method, which is particularly suitable for complex environments considering non-uniform responses and nonlinear effects, and can significantly improve the accuracy and reliability of polarization detection. BRIEF DESCRIPTION OF THE DRAWINGS

[0110] Figure 1 It is a flow chart of a method for constructing a polarization calibration accuracy model of a focal plane visible light polarization detector in an embodiment of the present invention;

[0111] Figure 2 Schematic diagram of polarization pixel array of split-focus plane polarization imaging detector in an embodiment of the present invention;

[0112] Figure 3 A flow chart for constructing the Mueller matrix of a detector in an embodiment of the present invention;

[0113] Figure 4 The signal transmission process of the focal plane polarization imaging detector in the embodiment of the present invention;

[0114] Figure 5 The basic process of calibrating the parameters of the polarization calibration accuracy model in the embodiment of the present invention;

[0115] Figure 6 This is a schematic diagram of relative transmittance testing in an embodiment of the present invention;

[0116] Figure 7 This is the basic process for verifying the polarization calibration accuracy model in the embodiment of the present invention.

[0117] Figure 8 It is a correction flow chart of the polarization calibration accuracy model in the embodiment of the invention. DETAILED DESCRIPTION

[0118] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in combination with the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0119] This embodiment describes a method for constructing a polarization calibration accuracy model for a focal plane visible light polarization detector, comprising the following steps:

[0120] Step 1: Build a mathematical model to describe the performance of the polarization detector. The goal is to define the detector's response characteristics to light of different polarizations, including error sources, etc. This model includes the detector's polarization sensitivity, the response to various polarization states, etc.

[0121] Step 2, polarization calibration accuracy model parameter calculation: experimentally measure the response of light sources with different known polarization states in the visible light polarization detector, and use these measurement data to infer the parameters of the polarization calibration accuracy model;

[0122] Step 3: Verify the polarization calibration accuracy model after the parameter calibration in step S02 to obtain the polarization degree error of the visible light polarization detector. By comparing the theoretical and actual polarization degrees, analyze the error source and make corrections.

[0123] Step 4: Based on the error found in step 3, the model is corrected and the polarization calibration accuracy model is updated to optimize the performance of the visible light polarization detector. The Mueller matrix after this adjustment should be able to accurately invert the polarization characteristics of the target object.

[0124] 1. Constructing a polarization calibration accuracy model for visible light polarization detectors

[0125] like Figure 2 As shown in the figure, split-focal plane polarization imaging is a method of integrating the pixel and the micro-polarizer on the same focal plane, and photolithography of a metal wire grid is performed on each pixel, so that each pixel corresponds to a micro-polarizer in one direction. Figure 1 As shown in the figure, on the focal plane, four adjacent pixels form a polarization pixel of different sizes, corresponding to four micro-polarizers at 0°, 45°, 90° and 135°. Each polarization pixel can simultaneously obtain polarization information in four directions (the four directions are referred to as polarization directions). The information in the four polarization directions can be calculated to obtain polarization parameters, which can reflect the polarization characteristics of the target, thus completing the polarization imaging process.

[0126] When describing the process of light passing through multiple optical elements in an optical system, the Mueller Matrix is ​​used to represent how the polarization state of light changes under the influence of these elements. Each optical element will affect the polarization state of light, and the influence of these elements can usually be described by the Mueller Matrix. The Stokes vector S of the incident wave multiplied by the Mueller matrix M of the detector on the left can obtain the Stokes vector S' of the outgoing wave, that is:

[0127] S'=M*S (1)

[0128] If the incident light passes through multiple optical devices or media in sequence (the Mueller matrices of these optical devices or media are M1, M2, ..., M n ), then the Stokes vector S' of the outgoing light is:

[0129] S'=M n …M3M2M1S (2)

[0130] The Stokes vector S is usually expressed as:

[0131]

[0132] [IQ UV] T , where I is the total light intensity, Q is the horizontal or vertical linear polarization component; U is the +45° or -45° linear polarization component; V is the left-handed or right-handed circular polarization component. Since the amount of circularly polarized radiation in natural scene radiation is extremely small, V is usually ignored. Through the intensity information of the four polarization directions, the Stokes parameters, polarization degree and polarization angle of the incident light can be calculated.

[0133]

[0134] Where: DoLP represents the degree of polarization; AoP represents the angle of polarization.

[0135] like Figure 3 As shown, the process of constructing the Mueller matrix is ​​as follows:

[0136] Step 1: Construct the Mueller matrix of the detector

[0137] If the incident light first passes through the microlens array and then the micropolarizer array, then we should combine the Mueller matrices of these two elements according to the propagation order of the light. That is, the light first passes through the microlens array and then the micropolarizer array. Therefore, the total Mueller matrix of the system is the product of the Mueller matrix of the microlens and the Mueller matrix of the micropolarizer. Let's first analyze the Mueller matrix of the microlens array.

[0138] For microlens arrays, we are usually more concerned about the intensity and polarization state of light, without considering elliptical polarization (i.e., V component). In many polarization detection systems, the polarization characteristics of the light source are often linear polarization, or the system design only considers the linear polarization components (i.e., Q and U). In this case, the Mueller matrix can be simplified to a 3×3 matrix without considering the change of V.

[0139] Assuming that the incident light beam corresponding to a certain pixel can be decomposed into mutually orthogonal p light and s light (the surface composed of p and s is parallel to the CMOS photosensitive surface), the transmittance of the microlens to p light and s light is η1 and η2 respectively, and the linear bidirectional attenuation of the microlens is ε, then

[0140]

[0141] Therefore, the Mueller matrix M of the microlens array in this system is lens :

[0142]

[0143] in:

[0144] η1 and η2: These are the transmittance coefficients of the microlens array for two orthogonal polarizations, reflecting the intensity variation of polarized light.

[0145] ε: This element describes the cross-coupling effect, i.e. the interaction between p-light and s-light. It indicates the degree of coupling between the two polarization directions, affecting the polarization state of the light.

[0146] This element ensures the conservation of the light intensity of the matrix (i.e. ensures that the total light intensity does not change). It is derived from the identity matrix and is used to ensure that the total intensity of the light does not change with changes in polarization.

[0147] and These two elements determine the intensity attenuation of light and the propagation of polarized light. Since η1 and η2 are the transmittances of p-light and s-light respectively, their sum of squares reflects the total transmittance.

[0148] and This term relates to the cross-coupling effect between p- and s-light, and ε is the difference in transmittance between polarizations. Through this term, the matrix takes into account the interaction between p- and s-light, which usually occurs in optical components with polarization dependence.

[0149] This term deals with the "non-linear" attenuation of polarized light. It adjusts the propagation intensity of polarized light, taking into account the relative intensity attenuation between different polarization states.

[0150] Ignoring the polarization effect of CMOS, the Mueller matrix M of the instrument mainly includes the Mueller matrix of the polarizer and the Mueller matrix M of the microlens lens .

[0151]

[0152] The micro-polarization array can be approximately regarded as an ideal linear polarization device, and the Mueller matrix of the polarization azimuth angle θ

[0153]

[0154] In an optical system, when light passes from one element to the next, the Mueller matrix of the latter element will act on the polarization state of the light output by the previous element. The microlens array first acts on the incident light to change the propagation direction and polarization state of the light. Subsequently, the micropolarizer array will act on the polarization state of the light output by the microlens, further changing the polarization characteristics of the light. Therefore, the Mueller matrix M of the detector is:

[0155]

[0156] After the incident light passes through the microlens and micropolarizer array, the Stokes vector reaching the CMOS can be represented by the instrument's Mueller matrix M and the Stokes vector of the incident light, and the physical quantity that the CMOS can detect is the total light intensity in the Stokes vector. The first term of the light Stokes vector is the light intensity, so the light intensity detected by the CMOS is I CMOS It is expressed as:

[0157]

[0158] After passing through the microlens and micropolarizer array, the incident light finally reaches the CMOS detector. Here, the Stokes vector of the light is affected by the polarization filter array. Specifically, the CMOS detector can detect the total light intensity part of the modulated Stokes vector.

[0159] Step 2: Build a detector response model

[0160] like Figure 4As shown in the figure, when light enters the polarization imaging system, it is first processed by the optical system and components, and then transmitted to the surface of the detector. The photosensitive surface of the detector can capture the radiation energy of the light signal. Once the photosensitive surface captures this energy, the detector starts its function and uses the internal photoelectric effect to convert the captured radiation energy into an analog electrical signal. Subsequently, the electronic system processes and processes the detection signal converted by the photoelectric imaging system in various signal processing circuits according to different system function requirements, including amplification, correction and compensation steps. Finally, the processed analog signal is converted into a digital signal through A / D conversion and output for display.

[0161] The light intensity I passing through the detector CMOS It will be detected by the detector and converted into an electronic signal. The detector's response count value C is related to the incident light intensity I CMOS The relationship is given by the following expression:

[0162]

[0163] in:

[0164] α is the photoelectric conversion coefficient, which indicates the efficiency of converting photons into electronic signals in the detector.

[0165] R is the electronic gain, which represents the electronic amplification factor of the detector (the signal gain increased during the electronic processing).

[0166] ρ(x,y) is the detection efficiency non-uniformity at position (x,y), that is, the difference in response at different positions or directions.

[0167] γ is a nonlinear index that indicates the degree of nonlinearity of the response.

[0168] c is the dark current response, which indicates the signal generated by the detector due to noise and other reasons when there is no incident light.

[0169] The signal intensity is adjusted by the photoelectric conversion coefficient α and the electronic gain R, and a constant term c is added to compensate for dark current or background noise.

[0170] Substitute the light intensity of the detector into the response model of the detector, and finally obtain the expression of the response model of the detector:

[0171]

[0172] After passing through the microlens and micropolarizer array, the incident light finally reaches the CMOS detector. Here, the Stokes vector of the light will be affected by the polarization filter array. Specifically, the CMOS detector can detect the total light intensity part of the modulated Stokes vector. The specific form provided in the formula is:

[0173]

[0174] Polarization detectors generally use multiple polarization directions (such as P1 to P4) to measure polarization information at different angles. The response value of each channel can be expressed by the following formula:

[0175]

[0176] Among them, τ k is the transmittance (or gain) of each channel, θ k is the polarization azimuth corresponding to each channel (for example, θ1 corresponds to P1, θ2 corresponds to P2, and so on), and other symbols are the same as the above description.

[0177] According to the formula:

[0178] Q=DoLP·I·cos(2AoP) (16)

[0179] U=DoLP·I·sin(2AoP)

[0180] Substituting Q and U into the equation:

[0181]

[0182] Combining the responses of all four polarization channels into a matrix form, the detector response model is:

[0183]

[0184] Among them, C1 to C4 correspond to the response values ​​of four polarization directions respectively;

[0185] τ1 to τ4 represent the relative transmittance of P1 to P4;

[0186] θ k Indicates the polarization azimuth angle;

[0187] The calculation of polarization degree is:

[0188]

[0189] 2. Polarization calibration accuracy model parameter calibration

[0190] Polarization calibration accuracy model parameter calibration refers to the calibration of polarization instruments or optical systems to ensure that the polarization information they measure is accurate and reliable, aiming to eliminate systematic errors in polarization signals and ensure comparability under different experimental environments. According to the main work content of model parameter calibration, the basic process of polarization calibration accuracy model parameter calibration is as follows: Figure 5 shown.

[0191] (1) Relative transmittance test

[0192] The relative transmittance of the visible light defocusing plane polarization imaging detection system is the ratio of the response value in each polarization direction to the response value in the reference polarization direction. The response values ​​in different polarization directions are measured using an integrating sphere, and the ratio of the response value in each polarization direction to the response value in the reference polarization direction is calculated to obtain the relative transmittance of the visible light defocusing plane polarization imaging detection system. The relative transmittance test diagram is shown in Figure 6 shown.

[0193] Step 1: The specific measurement steps and data processing method are as follows: adjust the detection unit to face the adjustable polarization light source.

[0194] Step 2, turn on the detection unit, adjust the spectral radiance output of the integrating sphere radiation source to a natural light state of 75% of the effective maximum response of the detection unit; preheat the detection unit and the adjustable polarization light source to a stable state.

[0195] Step 3: Calculate the relative transmittance T by detecting the responses of different polarization channels of the unit. k,a , as shown below.

[0196] T k,a =DN k,a / DN k,1 (20)

[0197] Where: T k,a ——Relative transmittance of different polarization channels in k-band;

[0198] DN k,a ——Response values ​​of different polarization channels in the k-band after deducting the background.

[0199] a——represents the polarization channel number (0, 1, 2 and 3);

[0200] (2) Pixel polarization direction test

[0201] In order to improve the accuracy of the data obtained by the polarization imaging detection system, the four polarization transmission axis directions of the system are measured to check whether the directions of the four polarizers are fixed and corrected according to the originally designed 0°, 45°, 90° and 135°. The angle interval between the transmission axis directions of the polarizers with four different polarization directions in the focal plane polarization imaging detection unit is used, and one of the directions is used as the 0-degree reference transmission axis direction. The curve fitting method is selected to detect the angle error of the polarizer direction. The schematic diagram of the pixel polarization direction test is shown in the figure. Figure 6 shown.

[0202] The polarization azimuth of the focal plane polarization imaging detection unit is obtained by a test method of rotating the reference polarization. The Malus fitting model is used for polarization azimuth data processing, as shown in formula (7).

[0203] y i =y 0,i +A i ·sin 2 (xx c,i ) (twenty one)

[0204] Set the P1 channel as reference 0°, x c,2 -x c,1 The value is the direction of the P2 channel relative to the transmission axis, x c,3 -x c,1 The value is the relative resolution direction of the P3 channel, calculate x c,4 -x c,1 The value is the relative resolution direction of the P4 channel.

[0205] Step 1, follow Figure 6 Install the measuring instrument as shown.

[0206] Step 2: The detection unit is turned on and preheated, and the precision electric control turntable drives the reference polarizer to rotate so that the response of the load in the 0° polarization resolution direction is close to the minimum value.

[0207] Step 3: Rotate the precision turntable within the range of 0 to 360°. The detection unit records the response value of the 0° polarization resolution direction every 0.25°. Perform sine function fitting on the turntable angle and the response value, and record the corresponding turntable angle reading φ o is the initial angle.

[0208] Step 4: The high-precision turntable drives the reference polarizer to rotate in the range of 0° to 360°. The detection unit records the response value of each polarization resolution direction every 0.25°, performs sine function fitting on the turntable angle and the response value, and records the corresponding turntable angle reading φ 60 , measure the turntable angle reading φ in the same way 120 .

[0209] Step 5: Calculate and record the polarizer transmission axis direction aπ / 3+α of the detection unit k,a .

[0210] Step 6, repeat test steps ② to ④ to measure the transmission axis direction of the polarizer of other bands of the detection unit.

[0211] (3) Photoelectric conversion coefficient measurement

[0212] Step 1: Prepare a light source of known intensity, such as a laser or a standard light source, and ensure that the light source is stable and its output power is controllable.

[0213] Step 2, irradiate the light source onto the detector surface and record the output signal (current or voltage) of the detector.

[0214] Step 3: Use a photometer or power meter to measure the optical power (P_in) irradiating the detector.

[0215] Step 4: Calculate the photoelectric conversion coefficient based on the ratio of the known input optical power to the detector output signal:

[0216]

[0217] (4) Electronic gain measurement:

[0218] Step 1: Apply a known electrical signal, usually a constant current signal, to the detector.

[0219] Step 2, measure the output signal (current or voltage) of the detector.

[0220] Step 3, calculate the electronics gain using the known current value (I_in) and the ratio of the output signal:

[0221]

[0222] Step 4: Determine the gain stability by measuring the output signal under different input currents for multiple times.

[0223] (5) Detection efficiency non-uniformity measurement (including dark current measurement)

[0224] When the photoelectric detection unit images the target, if the field of view target is uniform, then the ideal image output signal is also uniform. However, many factors make the image often uneven: the unevenness of some optical lenses and films during the manufacturing process of the detector, the unevenness of the filter transmittance within a certain spectral range; the unevenness of the photosensitive elements of the detector imaging device and the unevenness of the detector, etc. The unevenness of the photosensitive elements is caused by the unevenness of the process and materials. The larger the device, the more prominent the uniformity problem. The unevenness of the detector is mainly caused by the inconsistency of the transfer efficiency. All these unevenness problems can be solved by uniformity calibration correction.

[0225] Within the linear response range of the photoelectric detection unit, the ratio of the standard deviation to the mean of the collected image is used to represent the non-uniformity of the detector. The specific formula is shown in Equation 3:

[0226]

[0227] Where M×N is the total number of detectors. represents the average value of all pixel responses, and ρ is the inhomogeneity parameter of the detector response.

[0228] Step 1: Block the light entrance hole of the detector and perform 50 measurements. The grayscale values ​​of these images are averaged to obtain the dark current.

[0229] Step 2: Align the light entrance hole of the multi-band polarization CCD camera with the light source and perform 50 measurements. Analyze the DN value from the obtained image and calculate its average value and variance:

[0230]

[0231] In the formula, It represents the average DN value of the image obtained after 50 measurements. DN represents the average value of dark current;

[0232] Step 3, the radiation calibration coefficient is obtained as:

[0233]

[0234] Step 4, each pixel is multiplied by the corresponding calibration coefficient, thus completing the correction of each channel image;

[0235] Step 5: Calculate the non-uniformity parameter ρ of the processed image using equation (3) to verify the image plane uniformity correction result.

[0236] (6) Nonlinear index measurement

[0237] The nonlinear index is mainly used to understand whether the detector's response to different light intensities follows a linear relationship. The nonlinear index describes the degree of nonlinearity in the relationship between the detector output signal and the input light intensity. When γ = 1, the response is linear; when γ> 1, the response is superlinear; when γ< 1, it is sublinear. Introducing this power relationship helps to more accurately simulate the saturation phenomenon at high light intensity or the compression phenomenon at low light intensity.

[0238] Step 1: Prepare an integrating sphere with adjustable light intensity, collimate the light source and illuminate it onto the light polarization detector to ensure that the detector surface can evenly receive the radiation of the light source.

[0239] Step 2: Adjust the number of lights on in the integrating sphere to change the output brightness of the integrating sphere, and measure the nonlinear error when the instrument response is close to the full scale and one-third of the full scale.

[0240] Step 3: For each light intensity level, record the input light intensity (usually measured using an optical power meter) and the detector output signal (current or voltage). When recording, try to minimize ambient light and other interference during the measurement process.

[0241] Assume that the relationship between the output signal and the light intensity is:

[0242]

[0243] Among them, I out is the output signal of the detector, I in is the input light intensity, β is the proportionality coefficient, and γ is the nonlinear exponent.

[0244] Step 4: To extract the nonlinear index γ, the above relationship can be logarithmized to obtain:

[0245] log(I out )=log(β)+γlog(I in ) (28)

[0247] After logarithmic transformation of the input and output data under different light intensities, there should be a linear relationship between the logarithmic value of the output signal and the logarithmic value of the input light intensity.

[0248] Step 5, use the linear regression method to fit the above logarithmic transformed data. The slope of the fitting result is the nonlinear index γ. If the relationship between the logarithm of the output signal and the logarithm of the input light intensity is linear, then the slope of the fitting curve will be the nonlinear index γ. The slope (γ) in the linear regression fitting result is the nonlinear index of the detector.

[0249] (7) Fitting system polarization calibration accuracy model

[0250] The relevant parameters measured by polarization calibration, such as relative transmittance, polarization azimuth, photoelectric conversion coefficient, electronic gain, detection efficiency non-uniformity and nonlinear index, are substituted into formula 18 to obtain the polarization calibration accuracy model of the visible light focal plane polarization imaging detection system.

[0251] 3. Polarization calibration accuracy model verification

[0252] Polarization measurement accuracy refers to the degree of deviation of the polarization measurement value of the focal plane polarization imaging detection unit from the reference value of the measured target after polarization calibration. The polarization calibration accuracy test uses the adjustable polarization light source VPOLS-II to generate a polarized light with a known polarization degree. By measuring and comparing the theoretical reference value of the adjustable polarization light source VPOLS-II, the polarization measurement accuracy of the visible light focal plane polarization imaging detection system is tested. The theoretical reference value of the adjustable polarization light source VPOLS-II is calculated through the transmittance data and rotation angle data of the flat glass in its polarization box. The deviation of the polarization measurement value from the reference value and the measurement uncertainty are used to determine whether the polarization measurement accuracy meets the use requirements.

[0253] like Figure 7 The specific measurement steps are as follows.

[0254] Step 1: Adjust the internal focusing collimator to align it with the two glass plates (initial position) of the polarization regulator.

[0255] Step 2: Turn on the detection unit and collect the cross-wire image signal through the internal focusing parallel light tube to find its central field of view.

[0256] Step 3, remove the internal focusing collimator, light up the adjustable polarization light source and preheat for 20 minutes;

[0257] Step 4, adjust the adjustable polarization light source to output 8 states, with polarization degrees of 0, 5%, 10%, 15%, 20%, 25%, 30% and 45% respectively. Each polarization channel of the detection unit measures the 8 states 10 times respectively.

[0258] Step 5: Calculate the reference value of the measured polarization degree under different bands and different polarization degrees (different rotation angles) and the detection unit measurement value statistical deviation.

[0259] The measured value of the visible light focal plane polarization imaging detection system is inverted by the following formula.

[0260]

[0261] The deviation between the measured value of the visible light split-focus plane polarization imaging detection system and the theoretical reference value of the adjustable polarization light source VPOLS-II is compared by the following formula, and the statistical deviation is calculated.

[0262]

[0263] According to the universal comparison standard for measurement uncertainty qualification assessment, E is calculated by formula (6-19): k .

[0264]

[0265] in Calculated at 0.5%, Calculated at 2%.

[0266] 4. Polarization calibration accuracy model error correction

[0267] To correct the Mueller matrix of the polarization detector, the main purpose is to solve the deviation between the detector measurement results and the theoretical polarization information. To this end, correction can be performed through the following steps.

[0268] Step 1: Compare the measured polarization degree with the theoretical polarization degree and calculate the error. The error is usually the difference between the measured polarization degree and the theoretical polarization degree.

[0269] Step 2: Select a suitable model (such as a linear or nonlinear model) to describe the error based on the characteristics of the error. The most common method is to use the least squares method to fit and find the appropriate error correction parameters.

[0270] Step 3: Use the error model to adjust the polarization calibration accuracy model of the detector. The specific method may be to update the polarization calibration accuracy model through matrix multiplication, weighted average or other optimization algorithms to reduce the error.

[0271] Step 4: Re-measure the polarized light and calculate whether the corrected polarization calibration accuracy model can accurately restore the known polarization degree. If the error is reduced, it indicates that the correction is effective; if the error is still large, it may be necessary to re-examine the error model or correction method.

[0272] The above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features may be replaced by equivalents. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for constructing a polarization calibration accuracy model for a focal plane visible light polarization detector, characterized in that: The following steps are involved: S01. Construct a polarization calibration accuracy model for visible light polarization detectors: Among them, C1 to C4 correspond to the response values ​​of the four polarization directions respectively; τ1 to τ4 represent the relative transmittance of P1 to P4; θ k represents the polarization azimuth angle; α and R are the photoelectric conversion coefficient and the electronic gain respectively; ρ(x,y) is the detection efficiency non-uniformity at position (x,y), that is, the response difference at different positions or directions; γ is a nonlinear index, which indicates the nonlinear degree of the response; c is the dark current response, which indicates the signal generated by the detector due to noise and other reasons when there is no light intensity incident; S02. Calculation of polarization calibration accuracy model parameters: Experimentally measure the responses of light sources with different known polarization states in the visible light polarization detector, and use these measurement data to infer the parameters of the polarization calibration accuracy model; S03. Verify the polarization calibration accuracy model after the parameter calibration in step S02 to obtain the polarization error of the visible light polarization detector; S04, updating the polarization calibration accuracy model according to the polarization degree error obtained in step S03, thereby optimizing the performance of the visible light polarization detector.

2. The method for constructing a polarization calibration accuracy model for a focal plane visible light polarization detector according to claim 1, characterized in that: The construction process of the polarization calibration accuracy model in step S01 is as follows: Step 1: Construct the Mueller matrix of the visible light polarization detector Assuming that the incident light beam corresponding to a certain pixel can be decomposed into mutually orthogonal p light and s light, the surface composed of p and s is parallel to the CMOS photosensitive surface, the transmittance of the microlens to p light and s light is η1 and η2 respectively, and the linear bidirectional attenuation of the microlens is ε, then Thus, the Mueller matrix M of the microlens array in the visible light polarization detector is lens : Where: η1 and η2: are the transmittance coefficients of the microlens array for two orthogonal polarized lights, reflecting the intensity change of polarized light; ε: is the cross-coupling effect, i.e. the interaction between p-light and s-light; Used to ensure that the total intensity of light does not change with changes in polarization; and Determines the intensity attenuation of light and the propagation of polarized light; since η1 and η2 are the transmittances of p-light and s-light respectively, their square sum reflects the total transmittance; and It involves the cross-coupling effect between p-light and s-light, and ε is the difference in transmittance between polarized lights; To deal with the nonlinear attenuation of polarized light; Adjusted the propagation intensity of polarized light; Ignoring the polarization effect of CMOS, the Mueller matrix M of the visible light polarization detector includes the Mueller matrix of the polarizer and the Mueller matrix M of the microlens lens ; The micro-polarization array is approximately regarded as an ideal linear polarization device, and the Mueller matrix of the θ polarization azimuth In an optical system, when light passes from one element to the next, the Mueller matrix of the latter element will act on the polarization state of the light output by the previous element. The microlens array first acts on the incident light to change the propagation direction and polarization state of the light; then, the micropolarizer array will act on the polarization state of the light output by the microlens, further changing the polarization characteristics of the light; so the Mueller matrix M of the detector is: After the incident light passes through the microlens and micropolarizer array, the Stokes vector reaching the CMOS can be represented by the Mueller matrix M and the Stokes vector of the incident light, and the physical quantity that the CMOS can detect is the total light intensity in the Stokes vector; the first term of the light Stokes vector is the light intensity, so the light intensity detected by the CMOS is I CMOS It is expressed as: After passing through the microlens and micropolarizer array, the incident light finally reaches the CMOS detector; Step 2: Construct a visible light polarization detector response model The response count value C of the visible light polarization detector and the incident light intensity I CMOS The relationship is: Among them: α is the photoelectric conversion coefficient, which indicates the efficiency of converting photons into electronic signals in the detector; R is the electronic gain, which indicates the electronic amplification factor of the detector; ρ(x,y) is the detection efficiency non-uniformity at position (x,y), that is, the response difference at different positions or directions; γ is a nonlinear index, which indicates the nonlinear degree of the response; c is the dark current response, which indicates the signal generated by the detector due to noise and other reasons when there is no light intensity incident; Substituting the light intensity of the visible light polarization detector into the response model of the visible light polarization detector, we finally get the expression of the response model of the visible light polarization detector: After passing through the microlens and micropolarizer array, the incident light finally reaches the CMOS detector. Here, the Stokes vector of the light will be affected by the polarization filter array. Specifically, the CMOS detector can detect the total light intensity part of the modulated Stokes vector. The specific form provided in the formula is: The response value of each channel can be expressed by the following formula: Among them, τ k is the transmittance of each channel, θ k is the polarization azimuth corresponding to each channel; According to the formula: Q=DoLP·I·cos(2AoP) U=DoLP·I·sin(2AoP) Substituting Q and U into the equation: The response models C(x, y) of the four polarization channels are combined into one matrix to obtain the polarization calibration accuracy model of the visible light polarization detector.

3. The method for constructing a polarization calibration accuracy model for a focal plane visible light polarization detector according to claim 1, characterized in that: The specific process in step S02 is: first obtain the relevant parameters of the relative transmittance, polarization azimuth, photoelectric conversion coefficient, electronic gain, detection efficiency non-uniformity and nonlinear index of the visible light polarization detector, and then substitute the above 7 data into the polarization calibration accuracy model to obtain the calibrated polarization calibration accuracy model.

4. The method for constructing a polarization calibration accuracy model for a focal plane visible light polarization detector according to claim 1, characterized in that: The verification method of the polarization calibration accuracy model in step S03 is: Start the visible light polarization detector, adjust the adjustable polarization light source to output 8 states, with polarization degrees of 0, 5%, 10%, 15%, 20%, 25%, 30% and 45% respectively; each polarization channel of the visible light polarization detector measures the 8 states 10 times respectively; Calculate the reference value of the measured polarization degree under different bands and different polarization conditions and the detection unit measurement value Statistical deviations; The measured value of the visible light polarization detector is inverted by the following formula. The deviation between the measured value of the visible light polarization detector and the theoretical reference value of the adjustable polarization light source VPOLS-II is compared by the following formula, and the statistical deviation is calculated. Calculate E according to the universal comparison standard for the qualification assessment of measurement uncertainty k ; in Calculated at 0.5%, Calculated at 2%.

5. The method for constructing a polarization calibration accuracy model for a focal plane visible light polarization detector according to any one of claims 1 to 4, characterized in that: The polarization calibration accuracy model error correction in step S04 includes: Step 1, comparing the measured polarization degree with the theoretical polarization degree and calculating the error; Step 2: Based on the characteristics of the error, use the error model to describe the error and find the appropriate error correction parameters; Step 3, using the error model to adjust the polarization calibration accuracy model of the visible light polarization detector; Step 4, re-measure the polarized light and calculate whether the corrected polarization calibration accuracy model can accurately restore the known polarization degree; if the error is reduced, it indicates that the correction is effective; if the error is still large, re-check the error model or correction method.

6. The method for constructing a polarization calibration accuracy model for a focal plane visible light polarization detector according to claim 5, characterized in that: The error model is the least squares method.

7. Polarization calibration accuracy model of focal plane visible light polarization detector, characterized in that: The polarization calibration accuracy model expression is as follows:

8. The polarization calibration accuracy model of the focal plane visible light polarization detector according to claim 7, characterized in that: The method for constructing the polarization calibration accuracy model includes: Step 1: Construct the Mueller matrix of the visible light polarization detector Assuming that the incident light beam corresponding to a certain pixel can be decomposed into mutually orthogonal p light and s light, the surface composed of p and s is parallel to the CMOS photosensitive surface, the transmittance of the microlens to p light and s light is η1 and η2 respectively, and the linear bidirectional attenuation of the microlens is ε, then Thus, the Mueller matrix M of the microlens array in the visible light polarization detector is lens : Among them: η1 and η2: are the transmittance coefficients of the microlens array for two orthogonal polarized lights, reflecting the intensity change of polarized light; ε: is the cross-coupling effect, that is, the interaction between p-light and s-light; Used to ensure that the total intensity of light does not change with changes in polarization; and Determines the intensity attenuation of light and the propagation of polarized light; since η1 and η2 are the transmittances of p-light and s-light respectively, their square sum reflects the total transmittance; and It involves the cross-coupling effect between p-light and s-light, and ε is the difference in transmittance between polarized lights; To deal with the nonlinear attenuation of polarized light; Adjusted the propagation intensity of polarized light; Ignoring the polarization effect of CMOS, the Mueller matrix M of the visible light polarization detector includes the Mueller matrix of the polarizer and the Mueller matrix M of the microlens lens ; The micro-polarization array is approximately regarded as an ideal linear polarization device, and the Mueller matrix of the θ polarization azimuth In an optical system, when light passes from one element to the next, the Mueller matrix of the latter element will act on the polarization state of the light output by the previous element. The microlens array first acts on the incident light to change the propagation direction and polarization state of the light; then, the micropolarizer array will act on the polarization state of the light output by the microlens, further changing the polarization characteristics of the light; so the Mueller matrix M of the detector is: After the incident light passes through the microlens and micropolarizer array, the Stokes vector reaching the CMOS can be represented by the Mueller matrix M and the Stokes vector of the incident light, and the physical quantity that the CMOS can detect is the total light intensity in the Stokes vector; the first term of the light Stokes vector is the light intensity, so the light intensity detected by the CMOS is I CMOS It is expressed as: After passing through the microlens and micropolarizer array, the incident light finally reaches the CMOS detector; Step 2: Construct a visible light polarization detector response model The response count value C of the visible light polarization detector and the incident light intensity I CMOS The relationship is: Among them: α is the photoelectric conversion coefficient, which indicates the efficiency of converting photons into electronic signals in the detector; R is the electronic gain, which indicates the electronic amplification factor of the detector; ρ(x,y) is the detection efficiency non-uniformity at position (x,y), that is, the response difference at different positions or directions; γ is a nonlinear index, which indicates the nonlinear degree of the response; c is the dark current response, which indicates the signal generated by the detector due to noise and other reasons when there is no light intensity incident; Substituting the light intensity of the visible light polarization detector into the response model of the visible light polarization detector, we finally get the expression of the response model of the visible light polarization detector: After passing through the microlens and micropolarizer array, the incident light finally reaches the CMOS detector. Here, the Stokes vector of the light will be affected by the polarization filter array. Specifically, the CMOS detector can detect the total light intensity part of the modulated Stokes vector. The specific form provided in the formula is: Polarization detectors generally use multiple polarization directions (such as P1 to P4) to measure polarization information at different angles. The response value of each channel can be expressed by the following formula: Among them, τ k is the transmittance of each channel, θ k is the polarization azimuth corresponding to each channel; According to the formula: Q=DoLP·I·cos(2AoP) U=DoLP·I·sin(2AoP) Substituting Q and U into the equation: The response models of the four polarization channels are combined into a matrix form, thus obtaining the polarization calibration accuracy model of the visible light polarization detector.

9. The polarization calibration accuracy model of the focal plane visible light polarization detector according to claim 7, characterized in that: The specific process of parameter calibration of the polarization calibration accuracy model is as follows: firstly, the relative transmittance, polarization azimuth, photoelectric conversion coefficient, electronic gain, detection efficiency non-uniformity and nonlinear index related parameters of the visible light polarization detector are obtained, and then the above 7 data are substituted into the polarization calibration accuracy model to obtain the calibrated polarization calibration accuracy model.

10. Using the polarization calibration accuracy model described in any one of claims 7 to 9 to calculate the degree of polarization DoLP of a visible light polarization detector, the calculation formula is:

Images