Method for automatically measuring the thickness of a multi-alkali photocathode film and storage medium

Abstract

The application discloses a kind of automatic measurement multi-base photocathode film thickness method and storage medium, the method is realized using algorithm and software, through the form of artificial interaction, adaptive calculation model multiple parameters, and according to model parameters and the image collected by high-definition camera to the multi-base photocathode film thickness is calculated. Including: 1) setting the rectangular area of light reflection color model calculation;2) input product line as sample;3) input film thickness empirical value and error range;4) whether sample point number reaches requirement is judged, if not reached, return 2);5) the photocathode film thickness is calculated.This application can guarantee that the working environment of sample point is consistent with the working environment of test product, greatly enhances the robustness and universality in actual production process.

Description

Technical Field

[0001] This invention relates to the field of multi-alkali photocathode film thickness measurement, specifically to a method and storage medium for automatically measuring the thickness of multi-alkali photocathode films, and more particularly to a method for automatically measuring the thickness of photocathode films based on a multi-parameter adaptive model. Background Technology

[0002] Low-light image intensifiers are currently the company's mainstream products. As the core component of low-light image intensifiers, the performance of multi-alkali photocathodes directly affects the overall performance of the image intensifier.

[0003] In the fabrication of multi-alkali photocathodes, especially in processes such as alternating reverse-pull switching, it is necessary to determine the film thickness. Under current process standards, obtaining cathode film thickness information through visual observation is a mature method. However, since this method relies on subjective judgment, misjudgments or errors are inevitable. Furthermore, some cathode components have unique structures that prevent obtaining film thickness information, requiring only rough judgments based on experience, leading to fluctuations in device performance and affecting product yield. In addition, observing color requires manual drying in an oven, which is a high-temperature environment, posing a risk of burns to operators and potential safety accidents. Moreover, during the oven's raising and lowering process, vibrations in the worktable frequently cause fluctuations in photocurrent, ultimately leading to incorrect exhaust vehicle judgments, Sb flushing affecting cathode performance, and even necessitating tube removal.

[0004] In current practical applications, automatically calculating the cathode film thickness based on film color has become an urgent problem to be solved in the preparation of multi-alkali photocathodes. Current film thickness testing methods are broadly divided into two types: contact measurement and non-contact measurement. Contact measurement methods mainly include profilometers and elliptic polarization methods. These methods can damage the optical film during application and are not conducive to automated production, severely restricting the improvement of production efficiency. Non-contact methods, such as the method and system for measuring optical film thickness based on reflectance spectroscopy disclosed in CN106441126A, are based on the principle of reflectance spectroscopy measurement. Using multiple sets of standard samples with known optical film thicknesses, the method obtains the correspondence between the optical film thickness and the color space coordinates of the optical film reflectance spectra in the standard samples. By measuring the reflectance spectrum of the sample to be tested and obtaining its corresponding color space coordinates, the optical film thickness of the sample to be tested is obtained according to the correspondence. This method has outstanding advantages such as no damage to the optical film and rapid measurement.

[0005] In practical applications, the aforementioned non-contact method and system for measuring optical thin film thickness based on reflectivity spectroscopy has the following shortcomings:

[0006] (1) This method measures the thickness of an optical thin film by measuring and processing the spectrum of reflected light. However, a dedicated spectral acquisition device is required for this measurement process. Under current process conditions, such a dedicated spectral acquisition device is usually not configured.

[0007] (2) This method requires multiple sets of standard samples with known optical film thicknesses, and based on the relationship between the standard sample spectrum and the optical film thickness, it can be extended to the actual application scenarios of optical films. However, the research and testing environment of the standard samples is inconsistent with the actual measurement of the optical film thickness. In the actual working environment, the reflected light is often affected by factors such as end face reflection, cathode activation lamp illumination, and alkali source energization. The actual spectrum differs from the measured spectrum of the standard sample. Therefore, the universality and robustness of this method need to be improved.

[0008] (3) If the method cannot achieve the expected measurement effect in actual application, the algorithm and system need to be improved and perfected multiple times based on the collected data. Data collection and system algorithm improvement are usually carried out by different personnel. Data collection is usually provided by production line technicians, while system algorithm is completed by R&D personnel. This process of multiple interactive R&Ds is carried out to improve the applicability of the system and algorithm, which will seriously affect the production progress of multi-alkali photocathode. Summary of the Invention

[0009] This invention is based on currently used equipment and mature processes, and proposes a series of improvements to address the shortcomings of non-contact methods, as follows:

[0010] (1) In the existing non-contact measurement scheme, in order to measure the thickness of the optical film, it is necessary to use a special spectral device to measure and analyze the spectrum of the reflected light. However, such equipment is not configured in the current production process. Considering that the current process uses the human eye to observe the film color to determine the film thickness, this invention uses the high-definition camera attached to the current production process to collect the color of the emitted light. It mainly uses a software algorithm scheme to be compatible with the current equipment configuration and process flow.

[0011] (2) In existing non-contact measurement schemes, especially in the method of estimating the thickness of the cathode film using spectroscopy, it is necessary to provide standard samples for research in order to establish the correspondence between the sample reflected light spectrum and the thickness of the cathode film. However, the adaptive multi-parameter model proposed in this invention does not require standard samples for testing. Instead, it uses the products on the production line as parameter calculation samples and performs several manual interactions to adaptively obtain the corresponding model parameters. If the difference in the parameters of the produced products is small, automatic detection can be achieved.

[0012] (3) In the non-contact measurement scheme for estimating the thickness of the cathode film using spectroscopy, there are differences between the actual measurement environment and the measurement environment of the standard sample. In addition, the reflected light is affected by factors such as end face reflection, cathode activation lamp illumination, and alkali source energization. Many factors will cause the test results to fall short of expectations. At this time, the production line technicians and system algorithm technicians need to repeatedly correct the data, which will seriously affect the production progress of the multi-alkali photocathode. However, the model proposed in this invention, as described in (2), does not require the provision of standard samples. All parameters are provided by the online product. If the calculation results are not ideal, the production line technicians only need to update the parameters on-site. The parameter update is fast and convenient and does not affect the production progress.

[0013] This invention utilizes existing equipment as a foundation, is compatible with current multi-alkali photocathode production processes, and is mainly implemented using algorithms and software. Through human interaction, it adaptively calculates multiple parameters in the model and calculates the thickness of the multi-alkali photocathode film based on the model parameters and images acquired through a high-definition camera.

[0014] (1) Set the rectangular region for calculating the light reflection color model. Since a high-definition color camera is used, its color channel number is 3. The dimension of the selected region's pixel points is H×W, where H represents the number of pixels in the height direction of the selected rectangular region, and W represents the number of pixels in the width direction of the selected rectangular region. The acquired color image is denoted by C, and its corresponding dimension is H×W×3. Therefore, the matrix element corresponding to each channel is C. (k) (k = 0, 1, 2), where k represents the channel number, i.e., the red, green, and blue primary color channels. The number of elements in each channel is H × W, and the primary color parameters in each channel are respectively... As shown in equation (1),

[0015]

[0016] (2) During the application of the model, technicians on the production line input the corresponding empirical color and empirical photocathode film thickness values ​​through experience and manual methods, and assume that the empirical color value output for the nth time is r. (n) The experience thickness is d (n) Design a corresponding multi-parameter model A, whose dimension is the same as that of the acquired light reflection image, i.e., H×W×3, and the parameter A corresponding to each channel. (k) (k = 0, 1, 2), each model parameter element is... As shown in equation (2),

[0017]

[0018] The model A has multiple parameters, h = 0 to H-1, w = 0 to W-1, and the number of parameters is H×W×3. In practical applications, the values ​​of H and W can be adjusted according to the actual situation. Within the acceptable range of the thickness calculation results, the corresponding model parameters can be quickly calculated and the thickness of the optical thin film can be calculated based on the parameters.

[0019] (3) Based on the input experience color value r (n) (n = 0, 1, ..., N-1) (N represents the number of input sample points), calculate multiple parameters of channel color mapping. By using the formula shown in equation (3), the error between the obtained color and the actual empirical color can be minimized. Through this method, multiple parameters of the color mapping in the model can be obtained.

[0020]

[0021] (4) Based on the input empirical thickness value d (n) (m=0,1,...,M-1), calculate the multi-parameter b of the photocathode thin film thickness mapping. (m) (n=0,1,...,N-1)(M represents the number of parameters that satisfy the empirical input thickness error). By using the formula shown in equation (4), the error between the obtained thickness and the empirical input thickness is minimized. Through this method, multiple parameters in the photocathode thin film thickness mapping in the model can be obtained.

[0022]

[0023] (5) Color mapping parameters in multi-model and thickness mapping parameter b (m) Calculated through the above steps, in practical applications, the light reflection image C acquired this time can be used... (k) (k = 0, 1, 2), the thickness of the photocathode film is calculated. As shown in equation (5),

[0024]

[0025] (6) If the film thickness calculated by the model of formula (5) differs greatly from the thickness obtained by manual detection, a new light reflection image can be collected again, and steps (2)-(5) can be repeated until the calculated film thickness is within an acceptable range.

[0026] Preferably, the values ​​of H and W are adjusted according to the actual situation. For example, in order to speed up the test, H=2 and W=2 are taken.

[0027] Preferably, M is an integer in the range of 2-10.

[0028] A computer-readable storage medium having a computer program stored thereon, characterized in that the program is executed by a processor to implement the steps of the method for automatically measuring the thickness of a multi-alkali photocathode film according to the present invention.

[0029] The data acquisition equipment used in the above process is based on the equipment on the current production line and is compatible with the current production process. The products on the production line are used as sample points for model parameter calculation. There is no need to provide additional standard sample points for processing. The data interaction input can be completed by the technical personnel on the production line, and the model parameter update is convenient and fast.

[0030] The beneficial effects of this invention are:

[0031] The proposed method for automatically measuring cathode film thickness based on a multi-parameter adaptive model utilizes equipment compatible with current production line equipment, eliminating the need for specialized equipment and offering excellent cost-effectiveness. The method primarily employs software and algorithms to achieve real-time calculation of multi-cathode film thickness on the production line. The sample points used for model parameter calculation are provided by products on the production line, eliminating the need for standard samples for calibration. This ensures that the working environment of the sample points matches that of the tested products, significantly enhancing the robustness and universality of the proposed method in actual production. When the production environment changes and the measured cathode film thickness differs significantly from the expected results, production line technicians can provide products as samples, and the system automatically updates the model parameters on-site. This eliminates the need to submit data to system and algorithm developers for refinement, effectively shortening data update time, providing convenience, and not impacting production progress. Attached Figure Description

[0032] Figure 1 This is the overall flowchart of the present invention.

[0033] Figure 2 This is a data processing flowchart of the present invention.

[0034] Figure 3 This is the error diagram of the multi-parameter model of the present invention. Detailed Implementation

[0035] The present invention will be further described below with reference to the accompanying drawings and embodiments.

[0036] like Figure 1 As shown, the method of the present invention includes three basic units: a system initialization unit, a model parameter adaptive calculation unit, and a cathode film thickness automatic calculation unit.

[0037] The system initialization unit initializes the working equipment by setting parameters and assigning initial values ​​to the model parameters. The model parameter adaptive calculation unit uses the products on the production line as samples through human-computer interaction and inputs color and thickness values ​​derived from human experience. The algorithm automatically calculates the corresponding model parameters based on the input samples and experience values. The cathode film thickness automatic calculation unit collects the color of the reflected light from the products on the production line in real time and automatically calculates the thickness of the multi-alkali photocathode film using the model parameters.

[0038] Example 1

[0039] like Figure 2 As shown, the implementation process of the present invention includes the following steps:

[0040] (1) Set the rectangular region for calculating the light reflection color model. Since a high-definition color camera is used, its color channel number is 3. The dimension of the selected region's pixel points is H×W, where H represents the number of pixels in the height direction of the selected rectangular region, and W represents the number of pixels in the width direction of the selected rectangular region. The acquired color image is denoted by C, and its corresponding dimension is H×W×3. Therefore, the matrix element corresponding to each channel is C. (k) (k = 0, 1, 2), where k represents the channel number, i.e., the red, green, and blue primary color channels. The number of elements in each channel is H × W, and the primary color parameters in each channel are respectively... As shown in equation (1),

[0041]

[0042] (2) During the application of the model, technicians on the production line input the corresponding empirical color and empirical photocathode film thickness values ​​through experience and manual methods, and assume that the empirical color value output for the nth time is r. (n) The experience thickness is d (n) Design a corresponding multi-parameter model A, whose dimension is the same as that of the acquired light reflection image, i.e., H×W×3, and the parameter A corresponding to each channel. (k) (k = 0, 1, 2), each model parameter element is... As shown in equation (2),

[0043]

[0044] As a model with multiple parameters, the number of parameters is H×W×3. In practical applications, the values ​​of H and W can be adjusted according to the actual situation. Within the acceptable range of the thickness calculation results, the corresponding model parameters can be quickly calculated and the thickness of the optical thin film can be calculated based on the parameters.

[0045] (3) Based on the input experience color value r (n)(n = 0, 1, ..., N-1), calculate multiple parameters of channel color mapping By using the formula shown in equation (3), the error between the obtained color and the actual empirical color can be minimized. Through this method, multiple parameters of the color mapping in the model can be obtained.

[0046]

[0047] (4) Based on the input empirical thickness value d (n) (n = 0, 1, ..., N-1), calculate the multi-parameter b of the photocathode thin film thickness mapping. (m) (m=0,1,…,M-1), by using the formula shown in equation (4), the error between the obtained thickness and the empirical input thickness is minimized. Through this method, multiple parameters in the photocathode thin film thickness mapping in the model can be obtained.

[0048]

[0049] (5) Color mapping parameters in multi-model and thickness mapping parameter b (m) Calculated through the above steps, in practical applications, the light reflection image C acquired this time can be used... (k) (k = 0, 1, 2), the thickness of the photocathode film is calculated. As shown in equation (5),

[0050]

[0051] (6) If the film thickness calculated by the model of formula (5) differs greatly from the thickness obtained by manual detection, a new light reflection image can be collected again, and steps (2)-(5) can be repeated until the calculated film thickness is within an acceptable range.

[0052] Example 2

[0053] The specific steps of the above algorithm are explained and its performance is analyzed with the help of field examples, and the effect of multi-cathode thin film thickness detection is tested for a certain model.

[0054] To expedite testing, the parameter matrix A was designed to be 2×2×3, i.e., H=2 and W=2. Twelve (H×W×3) online products were randomly selected from the production line as parameter samples. A human-computer interaction input method was used, allowing users to input the corresponding empirical color and empirical photocathode film thickness values. Calculations then yielded the corresponding model parameter matrix A and the corresponding mapping parameter b. (m) (m=0,1,...,M-1)(M=3), based on the input sample and empirical thickness, the specific parameters are as follows:

[0055]

[0056]

[0057]

[0058] b (0) =1.021×10 -8 ,b (1) =0.972×10 -8 ,b (2) =0.963×10 -8

[0059] Using the above parameters, the thickness of the multi-alkali cathode film in the same batch on the production line was tested, and 100 sample points were randomly selected for manual thickness inspection. The results were then analyzed and compared with those obtained from the multi-parameter model. In the figure, the error Δd is defined as: the film thickness d calculated using the model parameters. mod The film thickness d determined by human experience exp The difference between them, i.e.

[0060] Δd=d mod -d exp

[0061] The acceptable range of the calculated film thickness is ±0.25 μm.

[0062] 100 measured error graphs as shown Figure 3 As shown in the figure, the method for automatically measuring cathode film thickness using the multi-parameter adaptive model proposed in this invention yields multi-cathode film thicknesses within an acceptable range. This method can replace current manual experience methods and be applied to real-time detection of film thickness on the production line.

Claims

1. A method for automatically measuring the thickness of a multi-alkali photocathode film, characterized in that, Includes the following steps: (1) A high-definition color camera is used to capture light reflection color images. The image has 3 color channels. The captured color image is represented by C, and the matrix element corresponding to each channel is C. (k) for: Where: H represents the number of pixels in the height direction of the selected rectangular area, W represents the number of pixels in the width direction of the selected rectangular area, the pixel dimension of the selected area is H×W, and its corresponding dimension is H×W×3; k represents the channel number, i.e., the red, green, and blue primary color channels, k=0,1,2; the number of elements in each channel is H×W. This represents the primary color parameter on each channel; (2) On the production line, technicians input the corresponding empirical color and empirical photocathode film thickness values ​​manually based on their experience, and assume that the empirical color value output for the nth time is r. (n) The experience thickness is d (n) Design a corresponding multi-parameter model A, whose dimension is the same as that of the acquired light reflection image, and the parameters A corresponding to each channel. (k) for: Where: H represents the number of pixels in the height direction of the selected rectangular area, W represents the number of pixels in the width direction of the selected rectangular area, the pixel dimension of the selected area is H×W, and its corresponding dimension is H×W×3; k represents the channel number, i.e., the red, green and blue primary color channels, k=0,1,2; the number of elements in each channel is H×W; This represents the multiple parameters of model A, where h = 0 to H-1, w = 0 to W-1; (3) Based on the input experience color value r (n) Calculate multiple parameters of channel color mapping Through formula To minimize the error between the obtained color and the actual empirical color, this method can be used to obtain multiple parameters of the color mapping in the model, where: m = 0, 1, ..., M-1, and N represents the number of input sample points; (4) Based on the input empirical thickness value d (n) Calculate the multi-parameter b mapping of photocathode thin film thickness (m) Through formula To minimize the error between the obtained thickness and the empirical input thickness, multiple parameters in the photocathode film thickness mapping in the model can be obtained through this method, where: n=0,1,...,N-1, and M represents the number of parameters that satisfy the error with the empirical input thickness; (5) The color mapping parameters and thickness mapping parameter b (m) The light reflection image C obtained through steps (3) and (4) is calculated. (k) The thickness of the photocathode film was calculated. (6) If the film thickness calculated in step (5) differs significantly from the thickness obtained by manual detection, a new light reflection image is acquired again, and steps (2)-(5) are repeated until the calculated film thickness is within an acceptable range.

2. The method according to claim 1, characterized in that: The values ​​of H and W are adjusted according to the actual situation.

3. The method according to claim 1, characterized in that: H=2, W=2.

4. The method according to claim 1, characterized in that: The value of M is 2 to 10.

5. The method according to any one of claims 1-4, characterized in that: In step (6), the error Δd is defined as the photocathode film thickness d calculated in step (4). mod The film thickness d determined by human experience exp The difference between them, that is: Δd=d mod -d exp If Δd is within the range of ±0.25 μm, it represents the calculated photocathode film thickness d. mod Within acceptable limits.

6. A computer-readable storage medium having a computer program stored thereon, characterized in that, The program is executed by a processor to implement the steps of an automatic method for measuring the thickness of a multi-alkali photocathode film as described in any one of claims 1 to 5.

Images