A method of imaging detector pixel internal quantum efficiency calibration

Abstract

This invention discloses a method for calibrating the quantum efficiency within a pixel of an imaging detector. First, a relationship is established between the grayscale value I and the number of photons S of the imaging detector. Each pixel on the surface of the imaging detector is uniformly divided into k×k ordered micropixels, and the cumulative number of photons received during the corresponding exposure time is denoted as S(i,j). A coefficient matrix equation reflecting the quantum efficiency characteristics of each region on the pixel (m,n) of the imaging detector is established, and the quantum efficiency distribution of the pixel (m,n) is obtained by solving this coefficient matrix equation. During the relative movement of the imaging detector and the target, k images are captured. 2 The images are arranged such that the target light field S(i,j) appears sequentially in k. 2 On each pixel, establish k 2 A system of equations reflecting the relationship between gray value I and photon number S(i,j); based on the quantum efficiency distribution and the established system of equations, solve for k. 2 The super-resolution image [S(i,j)] is calculated based on the intra-pixel quantum efficiency distribution, thus avoiding spurious information, being unrestricted by the dataset, and improving calibration efficiency.

Description

Technical Field

[0001] This invention relates to the field of optical sensor technology, and in particular to a method for calibrating the quantum efficiency within a pixel of an imaging detector. Background Technology

[0002] Currently, acquiring large amounts of image information is inseparable from optical imaging detectors. Electronic information products such as mobile phones, satellites, cameras, and surveillance systems all incorporate optical imaging detectors. However, due to limitations in manufacturing technology, the pixel array response of optical sensors is not uniform. Furthermore, limitations in existing photosensitive materials result in a non-linear optical response. The pixel size is also limited by factors such as series current, leading to limitations in the quality and resolution of the acquired images.

[0003] Existing technologies for improving the imaging quality of optical sensors mainly include:

[0004] 1. Flat field correction method: The sensor is illuminated with an approximately uniform light field, and the response difference of the sensor with uniform light intensity is obtained, so as to calibrate the sensor.

[0005] 2. Utilizing a large dataset, starting from the image level, a neural network is built through deep learning to map image quality, thereby restoring high-quality images from low-quality images.

[0006] The aforementioned existing technology has the following drawbacks:

[0007] (1) For flat field calibration, the uniformity of the light field used is not ideal, and only a few light field intensities are considered for independent calibration. The nonlinear differences between pixels are not considered, making it difficult to return to a unified response curve and without improving the physical resolution.

[0008] (2) For deep learning, a large amount of datasets need to be built. Real datasets are not easy to obtain. In principle, deep learning methods are based on image reconstruction methods, which generate image-to-image mappings rather than physical reconstructions of detectors. False information is likely to appear in the reconstructed images. Summary of the Invention

[0009] The purpose of this invention is to provide a method for calibrating the intra-pixel quantum efficiency of an imaging detector. This method is based on solving the intra-pixel quantum efficiency distribution to calibrate the super-resolution image, thus avoiding false information, not being limited by the dataset, and improving calibration efficiency.

[0010] The objective of this invention is achieved through the following technical solution:

[0011] A method for calibrating the quantum efficiency within a pixel of an imaging detector, the method comprising:

[0012] Step 1: Establish the relationship between the gray value I of the imaging detector and the number of photons S;

[0013] Step 2: Divide each pixel on the surface of the imaging detector into k×k ordered micropixels. The cumulative number of photons received during the corresponding exposure time is denoted as S(i,j); where k is a positive integer set as needed.

[0014] Step 3: Establish the coefficient matrix equation reflecting the quantum efficiency characteristics of each region on the pixel (m, n) of the imaging detector, and solve the coefficient matrix equation to obtain the quantum efficiency distribution of the pixel (m, n).

[0015] Step 4: Take a picture of k during the relative movement of the imaging detector and the target. 2 The images are arranged such that the target light field S(i,j) appears sequentially in k. 2 Establish k on each pixel 2 A set of equations reflecting the relationship between gray value I and photon number S(i,j);

[0016] Step 5: Based on the quantum efficiency distribution obtained in Step 3 and the system of equations established in Step 4, solve for k. 2 Super-resolution image [S(i,j)].

[0017] As can be seen from the technical solution provided by the present invention, the above method is based on solving the super-resolution image calibration based on the quantum efficiency distribution within pixels, thus avoiding false information, not being limited by the dataset, and improving calibration efficiency. Attached Figure Description

[0018] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the following description of the embodiments will be briefly introduced. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0019] Figure 1 This is a schematic flowchart of a method for calibrating the intra-pixel quantum efficiency of an imaging detector according to an embodiment of the present invention;

[0020] Figure 2 This is a schematic diagram of the surface division model of the imaging detector according to an embodiment of the present invention. Detailed Implementation

[0021] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments, and do not constitute a limitation of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the protection scope of the present invention.

[0022] like Figure 1 The diagram shows a flowchart of a method for calibrating the intra-pixel quantum efficiency of an imaging detector according to an embodiment of the present invention. The method includes:

[0023] Step 1: Establish the relationship between the gray value I of the imaging detector and the number of photons S;

[0024] In this step, firstly, according to camera industry standards, the relationship between the grayscale value I and the photon number S in the linear mathematical model of an ideal camera is as follows:

[0025] I = I d +GN e =I d +GηS (1)

[0026] Where G is the system gain; N e N represents the number of photogenerated electrons generated in the study area on the surface of the imaging detector after being illuminated; d I represents the number of electrons generated in the study area on the surface of the imaging detector when there is no illumination. d This represents the grayscale value of the area when there is no light. This is called quantum efficiency;

[0027] Based on equation (1) and combined with the basic definitions and content in camera industry standards, the relationship between the gray value I of the imaging detector and the number of photons S can be expressed using polynomial expansion as follows:

[0028] I = a + bS + cS 2 +dS 3 +…(2)

[0029] a, b, c, and d are coefficient matrices reflecting quantum efficiency characteristics;

[0030] Since the imaging detector response is nearly linear within the illumination intensity range, the polynomial in equation (2) is omitted to the second degree term, and expressed as:

[0031] I = a + bS + cS 2 (3).

[0032] Step 2: Divide each pixel on the surface of the imaging detector into k×k ordered micropixels. The cumulative number of photons received during the corresponding exposure time is denoted as S(i,j); where k is a positive integer set as needed.

[0033] In this step, the relationship between the gray value I and the number of photons S(i,j) is expressed as:

[0034]

[0035] Where i and j are the coordinates of the micro-pixel to be studied within the pixel or the row and column number corresponding to its position. i and j are both positive integers and are both less than or equal to k.

[0036] like Figure 2 The diagram shown is a schematic representation of the surface division model of the imaging detector according to an embodiment of the present invention. In the grayscale image, one grayscale value I corresponds to k 2 If there are ordered labeled photon numbers S, then the image [S] composed of all photon numbers S on the imaging detector is k. 2 Super-resolution image.

[0037] Step 3: Establish the coefficient matrix equation reflecting the quantum efficiency characteristics of each region on the pixel (m, n) of the imaging detector, and solve the coefficient matrix equation to obtain the quantum efficiency distribution of the pixel (m, n).

[0038] In this step, the coefficient matrix equation reflecting the quantum efficiency characteristics of each region on the imaging detector pixel (m, n) is expressed as:

[0039] R i，j =[a m，n (i, j), b m，n (i, j), c m，n (i, j)] (5)

[0040] By solving R i，j The quantum efficiency distribution of pixels (m, n) is obtained, and the specific solution process is as follows:

[0041] The R value for each small region is solved using the following system of equations. i，j :

[0042]

[0043] Wherein, the ηth image grayscale value of pixel (m, n) is denoted as I. η，m，n Solving the system of equations (6) requires at least 3k 2 Group of corresponding captured images [I η ] and the corresponding light field distribution [S η ];

[0044] In order to find [S]η The basic output pattern of the imaging detector is obtained by generating an easily fitable interference pattern on the surface of the imaging detector using a laser interference fringe system. The pattern and the parameters set by the interference system are input into a computer system. An algorithm based on pre-written fitting equation parameters is used to iteratively fit the parameters in both time and space dimensions to determine the interference parameters that meet the accuracy requirements. This allows the determination of the light field distribution at any point on the detector surface. η ];

[0045] like Figure 2 As shown, a Cartesian coordinate system is established on the surface of the imaging detector CCD. The instantaneous photon number density distribution s(x, y, t) arriving at the CCD surface at time t is:

[0046]

[0047] Where x and y are the corresponding coordinates in the Cartesian coordinate system established on the imaging detector CCD; fx and fy are the spatial frequencies of the interference fringes in the x and y directions; A is the average value of the light field intensity; and B is the amplitude of the interference light field intensity. It is the frequency in the time domain; These are the initial phases, and all are constants.

[0048] t η The exposure time corresponds to the ηth frame, where η = 1, 2, 3, ..., the exposure time is denoted as τ, the pixel side length is L, and the number of photons received by the small unit (i, j) on pixel (m, n) after exposure is S. η，m，n (i, j) is represented as:

[0049]

[0050] in:

[0051]

[0052] F y =Lf y F x =Lf x ,

[0053] A S B S F x F y φ0 is a constant to be solved or fitted;

[0054] At 1x resolution:

[0055]

[0056] In this case:

[0057]

[0058] Therefore, we only need to fit A. S B S F x F y ,φ0, With a total of six constants, the standard continuous two-dimensional sinusoidal function interferogram corresponding to each frame can be drawn in space, where the gray-level matrix of the captured image of the standard sinusoidal function [I] m，n [S] is a noisy [S] m，n ], represented as:

[0059] [I t，m，n ] = [S t，m，n ]×[QE m，n (10)

[0060] Since the gain G does not affect the relative distribution of the optical field, we take G = 1 here, and the formula no longer contains G;

[0061] In order to obtain the grayscale matrix [I] m，n [S] m，n ], using multiple frames of captured images [I m，n (t η [S] replaces [S] m，n (t η Substituting into formula (9), the least squares method is used for iterative fitting in time and space, and noise is removed by mutual correction between multiple pixels and between multiple frames until the set tolerance of 10 is met. -5 , and output the results of the six constants in formula (8);

[0062] The noise-free photon count S corresponding to the required resolution is obtained by formula (8).

[0063] Step 4: Take a picture of k during the relative movement of the imaging detector and the target. 2 The images are arranged such that the target light field S(i,j) appears sequentially in k. 2 Establish k on each pixel 2 A set of equations reflecting the relationship between gray value I and photon number S(i,j);

[0064] In this step, k is established. 2 A set of equations reflecting the relationship between gray value I and photon number S(i,j) is expressed as follows:

[0065]

[0066] Step 5: Based on the quantum efficiency distribution obtained in Step 3 and the system of equations established in Step 4, solve for k.2 Super-resolution image [S(i,j)].

[0067] In this step, specifically, the k obtained in step 3 is... 2 Substituting the coefficient matrices a, b, c, and gray value I, which reflect the quantum efficiency characteristics, into the equation system 11 established in step 4, we can solve the equation system to obtain k. 2 Super-resolution image [S(i,j)].

[0068] The above method achieves physical super-resolution images based on the distribution of quantum efficiency within pixels, and is not limited by the dataset.

[0069] It is worth noting that the contents not described in detail in the embodiments of the present invention belong to the prior art known to those skilled in the art.

[0070] In summary, the method described in this embodiment of the invention retains the quadratic term of the detector response, and the obtained quantum efficiency can better match the actual detector response; at the same time, the method of this application is based on solving the calibration super-resolution image based on the intra-pixel quantum efficiency, so there will be no false information and it is not limited by the dataset.

[0071] Furthermore, those skilled in the art will understand that all or part of the steps in the methods of the above embodiments can be implemented by a program instructing related hardware, and the corresponding program can be stored in a computer-readable storage medium, such as a read-only memory, a disk, or an optical disk.

[0072] The above description is merely a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims. The information disclosed in the background section is intended only to enhance the understanding of the overall background technology of the present invention and should not be construed as an admission or implication in any way that such information constitutes prior art known to those skilled in the art.

Claims

1. A method for calibrating the quantum efficiency within a pixel of an imaging detector, characterized in that, The method includes: Step 1: Establish the relationship between the gray value I of the imaging detector and the number of photons S; Step 2: Divide each pixel on the surface of the imaging detector into k×k ordered micropixels. The cumulative number of photons received during the corresponding exposure time is denoted as S(i,j); where k is a positive integer set as needed; i and j are the coordinates of the micropixel to be studied within the pixel or the row and column number corresponding to its position. i and j are both positive integers and are both less than or equal to k. Step 3: Establish the coefficient matrix equation reflecting the quantum efficiency characteristics of each region on the pixel (m, n) of the imaging detector, and solve the coefficient matrix equation to obtain the quantum efficiency distribution of the pixel (m, n). Step 4: Take a picture of k during the relative movement of the imaging detector and the target. 2 The images are arranged such that the target light field S(i,j) appears sequentially in k. 2 Establish k on each pixel 2 A set of equations reflecting the relationship between gray value I and photon number S(i,j); Step 5: Based on the quantum efficiency distribution obtained in Step 3 and the system of equations established in Step 4, solve for k. 2 Super-resolution image [S(i,j)].

2. The method for calibrating the quantum efficiency within a pixel of an imaging detector according to claim 1, characterized in that, The process of step 1 is as follows: First, according to camera industry standards, the relationship between the grayscale value I and the photon number S in the linear mathematical model of an ideal camera is as follows: I=I d +GN e =I d +GηS (1) Where G is the system gain; N e N represents the number of photogenerated electrons generated in the study area on the surface of the imaging detector after being illuminated; d I represents the number of electrons generated in the study area on the surface of the imaging detector when there is no illumination. d This represents the grayscale value of the area when there is no light. This is called quantum efficiency; The relationship between the gray value I of the imaging detector and the number of photons S can be expressed using polynomial expansion as follows: I=a+bS+cS 2 +dS 3 +…(2) a, b, c, and d are coefficient matrices reflecting quantum efficiency characteristics; Since the imaging detector response is nearly linear within the illumination intensity range, the polynomial in equation (2) is omitted to the second degree term, and expressed as: I=a+bS+cS 2 (3)。 3. The method for calibrating the intra-pixel quantum efficiency of an imaging detector according to claim 2, characterized in that, In step 2, the relationship between the gray value I and the number of photons S(i,j) is expressed as follows: Where i and j are the coordinates of the micro-pixel under study within the pixel or the row and column number corresponding to its position, i and j are both positive integers and are both less than or equal to k; a, b, and c are coefficient matrices reflecting quantum efficiency characteristics; From equation (3), it can be seen that: a gray value I in a grayscale image corresponds to k 2 If there are ordered labeled photon numbers S, then the image [S] composed of all photon numbers S on the imaging detector is k. 2 Super-resolution image.

4. The method for calibrating the quantum efficiency within a pixel of an imaging detector according to claim 1, characterized in that, In step 3, the coefficient matrix equation reflecting the quantum efficiency characteristics of each region on the imaging detector pixel (m, n) is expressed as: R i，j =[a m，n (i,j),b m，n (i,j),c m，n (i,j)] (5) By solving R i，j The quantum efficiency distribution of pixels (m, n) is obtained, and the specific solution process is as follows: The R value for each small region is solved using the following system of equations. i，j : Wherein, the ηth image grayscale value of pixel (m, n) is denoted as I. η，m，n Solving the system of equations (6) requires at least 3k 2 Group of corresponding captured images [I η ] and the corresponding light field distribution [S η ]; t η The exposure time corresponds to the ηth frame, where η = 1, 2, 3, ..., the exposure time is denoted as τ, the pixel side length is L, and the number of photons received in the small region (i, j) of pixel (m, n) after the exposure is completed is S. η,m,n (i,j); Establish a Cartesian coordinate system on the surface of the imaging detector CCD. The instantaneous photon number density distribution arriving at the CCD surface at time t is: Where x and y are the corresponding coordinates in the Cartesian coordinate system established on the imaging detector CCD; f x f y A is the spatial frequency of the interference fringes in the x and y directions; B is the average value of the light field intensity; and C is the amplitude of the interference light field intensity. It is the frequency in the time domain; This is the initial phase; S η,m,n (i,j) is represented as: in: A S B S F x F y φ0 is a constant to be solved or fitted; At 1x resolution: In this case: Therefore, we only need to fit A. S B S F x F y ,φ0, With a total of six constants, the standard continuous two-dimensional sinusoidal function interferogram corresponding to each frame can be drawn in space, where the gray-level matrix of the captured image of the standard sinusoidal function [I] m，n [S] is the [S] that includes noise. m，n ], represented as: [I t，m，n ]=[S t，m，n ]×[QE m，n ] (10) Since the gain G does not affect the relative distribution of the optical field, we take G = 1 here, and the formula no longer contains G; In order to obtain the grayscale matrix [I] m，n [S] m，n ], using multiple frames of captured images [I m，n (t η [S] replaces [S] m，n (t η Substituting into formula (9), the least squares method is used for iterative fitting in time and space, and noise is removed by mutual correction between multiple pixels and between multiple frames until the set tolerance of 10 is met. -5 , and output the results of the six constants in formula (8); The noise-free photon count S corresponding to the required resolution is obtained by formula (8).

5. The method for calibrating the intra-pixel quantum efficiency of an imaging detector according to claim 4, characterized in that, In step 4, establish k 2 A set of equations reflecting the relationship between gray value I and photon number S(i,j) is expressed as follows:

Images