Multi-light fusion high-precision imaging performance testing device and use method thereof

Abstract

The invention discloses a multi-light fusion high-precision imaging performance testing device and a use method thereof, is applied to the field of optical detection, and aims to solve the problems that a single-spectrum testing method needs to repeat a testing process for multiple times, time is consumed, the testing process is tedious and the testing efficiency is low in the existing multi-spectrum testing. According to the invention, the light beam combiner is combined with information of different spectrum sections (visible light and infrared light) to realize fusion between spectrums, so that functional requirements of a test system are analyzed, a test framework is determined, emissivity of two blackbody sources and power of two integrating spheres are regulated and controlled, and simulation of a target object and a background environment is realized. The target feature, the background feature and the environment feature are controlled at the same time, and multi-dimensional complex working conditions are simulated, so that imaging performance test data with higher precision are obtained.

Description

Technical Field

[0001] The invention belongs to the field of optical detection, and in particular relates to a testing technology for multi-spectral integrated equipment. Background Art

[0002] In recent decades, optoelectronic technology has experienced significant progress and is widely used in many fields such as medical, security and defense. Looking to the future, the optoelectronic industry is rapidly moving towards portability, user-friendliness and high integration, which will promote the deeper penetration of optoelectronic technology into all walks of life. At present, most optoelectronic devices are mainly concentrated in the visible and infrared spectrum, among which visible light cameras and infrared thermal imagers are particularly prominent.

[0003] Modern conflicts, including the Gulf War, have clearly demonstrated the indispensable value of thermal imaging technology in night operations. For the production and inspection of optoelectronic devices, performance indicators are the key factors to determine whether they meet the standards, and how to effectively test them is directly related to whether the equipment can operate normally. As the two most common optoelectronic tools on the market, the performance improvement of visible light cameras and infrared thermal imagers depends on advanced testing technology to ensure their reliability. With the continuous advancement of optoelectronic technology, it has also driven the innovation of testing technology, which in turn has promoted the rise of integrated products of multi-functional spectral equipment.

[0004] At present, there are still some problems in the testing of multi-spectral integrated equipment: the test efficiency is low, because multi-spectral equipment integrates the functions of multiple spectral ranges, and the testing method using a single spectrum requires repeated testing processes, which is time-consuming and cumbersome; the redundancy of the test system increases, and the coexistence of multiple sets of test systems with different spectra not only takes up more space resources, but may also complicate data management due to incompatibility or repeated settings between systems; usually only fixed target features can be provided, which means that imaging performance can only be evaluated under limited conditions. For complex application scenarios, such test conditions are too simple and cannot fully reflect the performance of imaging equipment in the real world. Therefore, a new technology is needed to solve existing problems. Summary of the invention

[0005] In order to solve the above technical problems, the present invention proposes a high-precision imaging performance testing device with multi-light fusion and a method of using the same. The integrating sphere light source and the surface source blackbody in the device are used as the visible light source and the infrared light source respectively, and a beam synthesizer is used to realize the transmission of the visible light band and the reflection of the infrared band, thereby achieving the purpose of light source fusion.

[0006] One of the technical solutions adopted by the present invention is: a high-precision imaging performance testing device with multi-light fusion, comprising: a target light source simulation system 1, a background light source simulation system 2, an optical collimation system 3, an imaging acquisition system 4 and an imaging performance testing system 5; the incident target light generated by the target light source simulation system 1 is received by the optical collimation system 3, the incident background light generated by the background light source simulation system 2 is received by the optical collimation system 3, the optical collimation system 3 collimates the received incident target light and incident background light and transmits them to the imaging acquisition system 4, the imaging acquisition system 4 receives the output signal from the optical collimation system 3, and converts it into a digital signal and transmits it to the imaging performance testing system 5.

[0007] The target light source simulation system 1 includes: a target integrating sphere light source 11, a target black body unit 12, and a target beam synthesizer unit 13; the target integrating sphere light source 11 is located in the transmission direction of the target beam synthesizer unit 13, and is used to generate incident target visible light; the target black body unit 12 is located in the reflection direction of the target beam synthesizer unit 13, and is used to generate incident target infrared light; the target beam synthesizer unit 13 synthesizes the incident target visible light and the incident target infrared light into one beam as the incident target light output.

[0008] The background light source simulation system 2 includes: a background integrating sphere light source 21, a background black body unit 22, and a background beam synthesizer unit 23; the background integrating sphere light source 21 is located in the transmission direction of the background beam synthesizer unit 23, and is used to generate incident background visible light; the background black body unit 22 is located in the reflection direction of the background beam synthesizer unit 23, and is used to generate incident background infrared light; the background beam synthesizer unit 23 synthesizes the incident background visible light and the incident background infrared light into one beam as the incident background light output.

[0009] The second technical solution adopted by the present invention is: a testing method for a high-precision imaging performance testing device of multi-light fusion, the testing method specifically comprising the following steps:

[0010] S1. Turn on the target light source, background light source, visible light camera, thermal imager and industrial computer;

[0011] S2, fuse the visible light signal and the infrared signal through the beam fuser and transmit them to the target unit;

[0012] S3, adjusting the plane reflector unit and the off-axis parabolic mirror unit so that the main light passing through the target unit is incident on the visible light camera unit and the thermal imager unit in parallel, and ensuring that the incident angles of the main light of the central field of view on the visible light camera unit and the thermal imager unit are both 0°;

[0013] S4. Using the light source control unit, according to the use scenario of the camera to be tested, set the temperature of the background black body unit, keep the temperature of the background black body unit unchanged, and set the starting temperature difference between the target black body unit and the background black body unit to 1°C;

[0014] S5. According to the usage scenario of the camera to be tested, set the brightness of the background integrating sphere light source and keep it unchanged. By adjusting the power of the target integrating sphere light source, change the brightness contrast between the target and the background. The initial brightness contrast is set to 0%;

[0015] S6. Use a visible light camera and a thermal imager to respectively obtain the optical signals of the visible light part and the infrared part passing through the target, and transmit them to the industrial computer unit. The industrial computer unit converts the optical signals into digital signals for storage and sorting, thereby completing the target image acquisition; the target image includes a visible light image and an infrared image;

[0016] S7, the image registration and fusion unit performs image registration processing on the visible light image and the infrared image, adjusts the field of view and focal length of the visible light camera and the infrared thermal imager, makes the field of view of the two detectors consistent, and ensures that the test target image is located at the center of the two detectors and has the same size and clarity;

[0017] S8, the image registration and fusion unit fuses the visible light image and the infrared image based on the Laplace pyramid method;

[0018] S9, after each image fusion is completed, the resolvable spatial frequency test unit determines whether the target of the current spatial frequency can be distinguished based on the BP neural network. If it can be distinguished, the instruction is passed to the target control unit to switch to a target with a higher spatial frequency, and return to step S6 until the resolution limit is reached, and the spatial frequency corresponding to the current target is recorded;

[0019] S10, using the light source control unit, keeping the temperature difference between the target black body unit and the background black body unit unchanged, changing the target integrating sphere light source power, so that the brightness contrast between the target and the background changes from 0% to 100%, with a step amplitude of 5%, and returning to step S6 each time the brightness contrast changes, and recording the numerical change of the spatial frequency corresponding to the brightness contrast change under the current temperature difference;

[0020] S11, using the light source control unit, gradually reducing the temperature difference between the target black body unit and the background black body unit by 0.05°C, and repeating step S10 each time the temperature difference is changed until the temperature difference changes from 1°C to -1°C, and obtaining a spatial frequency data matrix;

[0021] S12, the minimum resolvable contrast response spatial frequency changes with the contrast between target brightness and background brightness, and the minimum resolvable temperature difference response spatial frequency changes with the temperature difference between target and background; therefore, the common dependent variable spatial frequency is set as the z-axis, the temperature difference between target and background is set as the x-axis, and the contrast between target brightness and background brightness is set as the y-axis;

[0022] S13. Using a data fitting unit, the spatial frequency data matrix is ​​stored in a coordinate space, and a relationship surface in which the spatial frequency changes simultaneously with the minimum resolvable contrast and the minimum resolvable temperature difference is fitted; and based on the relationship surface, the detector is divided into a recognizable area and an unrecognizable area.

[0023] Beneficial effects of the invention: The high-precision imaging performance test device of the invention with multi-light fusion is built based on a visible light camera and an infrared imager. It solves the problem that the traditional test system needs to install and align multiple optoelectronic devices multiple times, and simplifies the overall structure by sharing a set of optical systems for visible light and infrared light paths, while improving the versatility of the test system.

[0024] The dual light source system of the present invention can simulate more realistic usage conditions during testing, making the background environment controllable, thereby improving the accuracy of detection. In addition, the dual light source system can provide information in multiple dimensions at the same time, making the performance evaluation of the detector more comprehensive and accurate, and can better simulate various situations in actual application scenarios, thereby improving the reliability and repeatability of the test results. It has significant advantages such as a wide test wavelength range, strong temperature adaptability, high efficiency of multi-index synchronous testing, and accurate multi-modal imaging performance evaluation;

[0025] The present invention can judge the resolution limit based on the fused image and record the corresponding spatial frequency value, and then make a high-precision fitting of the mapping relationship between the spatial frequency and the minimum resolvable contrast and the minimum resolvable temperature difference, thereby realizing the accurate division of the detector's identifiable area and unrecognizable area, and achieving the purpose of high-quality integrated testing and evaluation of multiple optoelectronic devices. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] Figure 1 It is a flow chart of the testing method of the present invention;

[0027] Figure 2 It is a structural diagram of the system of the present invention;

[0028] Figure 3 This is a flow chart for resolvable spatial frequency testing.

[0029] Among them, 1-target light source simulation system, 11-target integrating sphere light source, 12-target blackbody unit, 13-target beam synthesizer unit, 2-background light source simulation system, 21-background integrating sphere light source, 22-background blackbody unit, 23-background beam synthesizer unit, 3-optical collimation system, 31-target unit, 32-plane reflector unit, 33-off-axis parabolic mirror unit, 4-imaging acquisition system, 41-visible light camera unit, 42-thermal imager unit, 43-industrial computer unit, 5-imaging performance test system, 51-image registration and fusion unit, 52-resolvable spatial frequency test unit, 53-light source control unit, 54-target control unit, 55-data fitting unit. DETAILED DESCRIPTION

[0030] To facilitate those skilled in the art to understand the technical content of the present invention, the present invention is further explained below with reference to the accompanying drawings.

[0031] like Figure 1 As shown, a high-precision imaging performance test device for multi-light fusion, the high-precision imaging performance test device includes: a target light source simulation system 1, a background light source simulation system 2, an optical collimation system 3, an imaging acquisition system 4 and an imaging performance test system 5; the incident target light generated by the target light source simulation system 1 is received by the optical collimation system 3, the incident background light generated by the background light source simulation system 2 is received by the optical collimation system 3, the optical collimation system 3 receives the output signals of the target light source simulation system 1 and the background light source simulation system 2, and passes them to the imaging acquisition system 4 after collimation, the imaging acquisition system 4 receives the output signal from the optical collimation system 3, and converts it into a digital signal and passes it to the imaging performance test system 5.

[0032] The target light source simulation system 1 realizes the transmission of the visible light band and the reflection of the infrared band of the target light source through light beam synthesis, thereby achieving the purpose of merging and regulating the target light source.

[0033] The background light source simulation system 2 realizes the transmission of the visible light band and the reflection of the infrared band of the background light source through light beam synthesis, thereby achieving the purpose of integrating and regulating the background light source.

[0034] The optical collimation system 3 is used to collimate the target light source and the background light source passing through the system;

[0035] The imaging acquisition system 4 is used to acquire visible light images and infrared images.

[0036] The imaging performance testing system 5 is used to perform high-precision testing on imaging performance.

[0037] like Figure 2As shown, the target light source simulation system 1 includes a target integrating sphere light source 11, a target black body unit 12, and a target beam synthesizer unit 13; the target integrating sphere light source 11 is located in the transmission direction of the target beam synthesizer unit 13, the target black body unit 12 is located in the reflection direction of the target beam synthesizer unit 13, and the target beam synthesizer unit 13 receives output signals from the target integrating sphere light source 11 and the target black body unit 12.

[0038] The output signal of the target beam synthesizer unit 13 is transmitted to the optical collimation system 3 .

[0039] like Figure 2 As shown, the background light source simulation system 2 includes a background integrating sphere light source 21, a background black body unit 22, and a background beam synthesizer unit 23; the background integrating sphere light source 21 is located in the transmission direction of the background beam synthesizer unit 23, the background black body unit 22 is located in the reflection direction of the background beam synthesizer unit 23, and the background beam synthesizer unit 23 receives output signals from the background integrating sphere light source 21 and the background black body unit 22.

[0040] The output signal of the background beam combiner unit 23 is transmitted to the optical collimation system 3 .

[0041] like Figure 2 As shown, the optical collimation system 3 includes a target unit 31, a plane reflector unit 32, and an off-axis parabolic mirror unit 33; the light from the target light source simulation system 1 and the background light source simulation system 2 passes through the target unit 31, is reflected and collimated by the plane reflector unit 32 and the off-axis parabolic mirror unit 33, and then is transmitted to the imaging acquisition system 4.

[0042] like Figure 2 As shown, the imaging acquisition system 4 includes a visible light camera unit 41, a thermal imager unit 42, and an industrial computer unit 43; the visible light camera unit 41 and the thermal imager unit 42 respectively obtain the output light signals of the visible light part and the infrared part, and pass them to the industrial computer unit 43, and the industrial computer unit 43 converts the output light signal into a digital signal, stores and organizes it, and then passes it to the imaging performance test system 5.

[0043] The imaging performance testing system 5 includes an image registration and fusion unit 51 , a resolvable spatial frequency testing unit 52 , a light source control unit 53 , a target control unit 54 , and a data fitting unit 55 .

[0044] The image registration and fusion unit 51 is connected to the industrial computer unit 43 , acquires digital signals of the visible light image and the infrared image in real time, performs registration and fusion, and then transmits the fused image to the resolvable spatial frequency testing unit 52 .

[0045] The light source control unit 53 is connected to the resolvable spatial frequency test unit 52, and after completing a test, automatically adjusts the target light source simulation system 1 and the background light source simulation system 2 to change the brightness contrast and temperature difference between the target and the background.

[0046] The target control unit 54 is connected to the resolvable spatial frequency test unit 52 and switches the target according to the test requirements.

[0047] The data fitting unit 55 is connected to the resolvable spatial frequency testing unit 52 to record the test data in real time and fit it into a curved surface.

[0048] like Figure 3 As shown, a high-precision imaging performance test method for multi-light fusion, the specific test method includes the following steps:

[0049] Step 1: Boot up the system

[0050] The target light source, background light source, visible light camera, thermal imager and industrial computer are started. The light sources include integrating sphere light source and black body light source, which provide optical signals in visible light band and infrared band respectively.

[0051] Step 2: Multi-beam fusion

[0052] The visible light signal and the infrared signal are fused through a beam fuser, and the fused light signal is transmitted to the target unit. The beam fuser adopts a multi-layer film optimization design method of dielectric material-metal material-dielectric material.

[0053] Step 3: Optical alignment system debugging

[0054] By adjusting the center distance, tilt, and pitch of the plane reflector unit and the off-axis parabolic mirror unit, the optical axis consistency, light parallelism, and the central field angle of the main light of the optical collimation system are adjusted. Make the main light passing through the target unit parallel to the visible light camera unit and the thermal imager unit, and ensure that the incident angle of the main light of the central field of view is 0° on the visible light camera unit and the thermal imager unit;

[0055] Step 4: Blackbody unit temperature difference setting

[0056] Use the light source control unit to set the background blackbody unit temperature according to the usage scenario of the camera to be tested (make the background blackbody unit temperature close to the camera usage scenario, and simulate normal ambient temperature scenarios, high temperature scenarios, low temperature scenarios, etc. as needed), and keep the background blackbody unit temperature unchanged, and set the starting temperature difference between the target blackbody unit and the background blackbody unit to 1°C.

[0057] Step 5: Brightness and contrast setting of integrating sphere light source

[0058] Use the light source control unit to set the background integrating sphere light source brightness according to the usage scenario of the camera to be tested (make the background integrating sphere light source brightness close to the camera usage scenario, and simulate normal weather scenes, strong background light scenes, low light scenes, etc. as needed), and keep the background integrating sphere light source brightness unchanged. By adjusting the target integrating sphere light source power, change the brightness contrast between the target and the background, and set the initial brightness contrast to 0%.

[0059] Step 6: Target Image Acquisition

[0060] The target unit uses a four-bar target as the simulation target. Different spatial frequencies can be switched according to test requirements. The spatial frequencies of the selected four-bar targets are as follows from low to high: 0.2, 0.5, 1.0, 1.2, 1.5, 1.7, 2.0, 2.5, 3.0, 3.5, 4.0, 4.5, 5.0, 5.5, 6.0, 6.5, 7.0, 7.5, 8.0, 8.5, 9.0, 9.5, 10.0, 12.0, 14.0, 18.0, 20.0, 22.0, 24.0, 26.0, 28.0, 30.0, 32.0, 34.0, 36.0, 38.0, 40.0 unit lp / mm; accuracy: ±0.005mm. Use a visible light camera and a thermal imager to respectively acquire the visible light and infrared light signals passing through the target, and convert them into digital signals and transmit them to the industrial computer unit. The industrial computer unit stores and organizes the digital signals to complete the target image acquisition.

[0061] Step 7: Image Registration

[0062] The image registration and fusion unit is used to register the visible light image and the infrared image, and the field of view and focal length of the visible light camera and the infrared thermal imager are adjusted to make the field of view of the two detectors consistent, and to ensure that the test target image is located at the center of the two detectors and has the same size and clarity;

[0063] Step 8: Image Fusion

[0064] This embodiment takes the traditional Laplacian pyramid method as an example to illustrate the image fusion process:

[0065] The visible light image and infrared image are fused based on the Laplacian pyramid method. The method first sets the number of pyramid layers, which depends on the application requirements and the size of the original image, and then constructs the pyramid. The image of the first layer is initialized as the input image. Each subsequent layer reduces the image of the previous layer by half and uses bicubic interpolation to generate a new image. For each layer, the Laplacian representation of the layer is obtained by subtracting the image of the next layer magnified twice from the current layer. Finally, the Laplacian pyramid layers of the corresponding infrared and visible light images are averaged to achieve image fusion.

[0066] Step 9: Resolvable Spatial Frequency Test

[0067] After each image fusion is completed, the resolvable spatial frequency test unit is used to determine whether the target of the current spatial frequency can be distinguished based on the BP neural network (BP neural network, full name Back Propagation Neural Network). If it can be distinguished, the target control unit transmits the target switching instruction to the target unit, switches to a target with a higher spatial frequency, and returns to step 6 until the resolution limit is reached, and the spatial frequency corresponding to the current target is recorded. The process is as follows:

[0068] (1) Interference from the background area of ​​the image will affect the resolution. Therefore, after reaching the detector resolution limit, the resolvable spatial frequency test unit will first select the approximate target area in the center of the field of view;

[0069] (2) The resolvable spatial frequency test unit automatically calculates the grayscale threshold α based on the grayscale value of the pixel in the framed area using the maximum inter-class variance method (OSTU), and then performs binary segmentation processing on the framed area based on the grayscale threshold α, that is, the grayscale value of the pixel point greater than the grayscale threshold α is set to 1, and the grayscale value of the pixel point less than the grayscale threshold α is set to 0;

[0070] (3) Detect the distribution of pixels with grayscale values ​​of 1 and 0 in the area. If the number of pixels with grayscale value of 1 is greater than half of the total number of pixels, invert the black and white distribution of the image.

[0071] (4) Extract the coordinates of the pixels with a gray value of 0 from the binary image, and extract the pixels with the corresponding coordinates in the original image to complete the target area extraction;

[0072] (5) Use the weighted moving average function to preprocess the target area image. The expression of this process is:

[0073] A j =λi j +λ(1-λ)i j-1 +λ(1-λ) 2 i j-2 ……+λ(1-λ) j-1 i1+λ(1-λ) j i0 (1)

[0074] Among them, i j is any column of a row of data perpendicular to the direction of the four-bar target, i is the row number corresponding to the pixel point in the image digital matrix, j is the column number corresponding to the pixel point in the image digital matrix, λ is the smoothing coefficient and λ=1 / j, A j is the arithmetic mean of the grayscale values ​​of the pixels;

[0075] (6) After completing the weighted moving average function preprocessing, find the maximum pixel point and the minimum pixel point in the target area image. For any pixel point in the image, take 7 neighborhoods in the j direction. If the gray value of the pixel point is greater than the neighboring pixel point, it is marked as the maximum pixel point. If the gray value of the pixel point is less than the neighboring pixel point, it is marked as the minimum pixel point.

[0076] (7) After extracting the maximum pixel and the minimum pixel, calculate the grayscale mean A of the maximum pixel respectively. h and the minimum pixel grayscale mean A l , and calculate the image background gray mean A b ;

[0077] (8) Calculate the stripe discernibility and mean discernibility of the four-bar target image using the following formula:

[0078]

[0079] Among them, δ1 is the fringe discernibility and δ2 is the mean discernibility.

[0080] (9) After the calculation of the four-bar target image discernibility is completed, the BP neural network is used to train the discernibility features. The input layer of the BP neural network is the spatial frequency, stripe discernibility and mean discernibility of the four-bar target corresponding to the resolution limit. The hidden layer uses the ReLU activation function to make the network more sparse and provide better training data. The output layer uses the Sigmoid activation function to output whether the target of the current spatial frequency can be distinguished. The accuracy of the network is observed after each training. CC , (Before training the neural network, the data set is divided into a training set and a test set. The training set is used to train the neural network, and the test set is used to evaluate the performance of the neural network, providing an evaluation index. The evaluation index of the present invention is accuracy, which is defined as the ratio of all correctly predicted samples to the total number of samples. The calculation formula is as follows:)

[0081]

[0082] Among them, T c represents the number of correctly identified distinguishable d represents the number of correctly identified indistinguishable numbers, F c represents the number of incorrectly identified distinguishable d It indicates the number of incorrectly identified as indistinguishable. CCWhen it is close to 100% and reaches stability, the training is stopped, and the trained BP neural network is used to determine whether the subsequent test results have reached the resolution limit. If the resolution limit is reached, it is resolvable. This method eliminates the need for the operator to use the human eye to judge whether the resolution limit has been reached after each image acquisition, thereby avoiding errors caused by repeatability and subjective judgment, significantly reducing test costs and improving test efficiency.

[0083] For details about BP neural network, please refer to the paper "Learning representations by back-propagating errors" published by David Rumelhart, Geoffrey Hinton, Ronald Williams and others in 1986.

[0084] Step 10: Set Brightness Contrast Changes

[0085] Using the light source control unit, keep the temperature difference between the target black body unit and the background black body unit unchanged, change the target integrating sphere light source power, and make the brightness contrast between the target and the background change from 0% to 100%, with a step amplitude of 5%. After each brightness contrast change, return to step 6 and record the numerical change of the spatial frequency corresponding to the brightness contrast change under the current temperature difference.

[0086] Step 11: Setting the Temperature Difference

[0087] Use the light source control unit to gradually reduce the temperature difference between the target black body unit and the background black body unit by 0.05°C. Repeat step 10 each time the temperature difference is changed until the temperature difference changes from 1°C to -1°C.

[0088] Step 12: Create a coordinate space

[0089] The minimum resolvable contrast response spatial frequency changes with the contrast between target brightness and background brightness, and the minimum resolvable temperature difference response spatial frequency changes with the temperature difference between target and background. Therefore, the common dependent variable spatial frequency is set as the z-axis, the temperature difference between target and background is set as the x-axis, and the contrast between target brightness and background brightness is set as the y-axis;

[0090] Step 13: Data Fitting

[0091] The test data is stored in the coordinate space using the data fitting unit, and the relationship surface of the spatial frequency changing with the minimum resolvable contrast and the minimum resolvable temperature difference is fitted. The fitting surface indicates that the fused image can be recognized at the corresponding temperature difference and contrast, so the curve divides the entire image into two parts: the unrecognizable area below the surface, that is, the fused image cannot be clearly recognized under the temperature difference and contrast less than the calibration of the fitting surface; the recognizable area above the surface, that is, the fused image can be clearly recognized under the temperature difference and contrast greater than the calibration of the fitting surface.

[0092] Those skilled in the art will appreciate that the embodiments described herein are intended to help readers understand the principles of the present invention, and should be understood that the scope of protection of the present invention is not limited to such specific statements and embodiments. For those skilled in the art, the present invention may have various changes and variations. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of the claims of the present invention.

Claims

1. A high-precision imaging performance test device for multi-light fusion, characterized in that: include: A target light source simulation system (1), a background light source simulation system (2), an optical collimation system (3), an imaging acquisition system (4) and an imaging performance test system (5); the incident target light generated by the target light source simulation system (1) is received by the optical collimation system (3), the incident background light generated by the background light source simulation system (2) is received by the optical collimation system (3), the optical collimation system (3) collimates the received incident target light and incident background light and transmits them to the imaging acquisition system (4), the imaging acquisition system (4) receives the output signal from the optical collimation system (3), converts it into a digital signal and transmits it to the imaging performance test system (5).

2. The high-precision imaging performance testing device for multi-light fusion according to claim 1, characterized in that: The target light source simulation system (1) comprises: a target integrating sphere light source (11), a target black body unit (12), and a target light beam synthesizer unit (13); the target visible light generated by the target integrating sphere light source (11) is transmitted through the target light beam synthesizer unit (13); the target infrared light generated by the target black body unit (12) is reflected through the target light beam synthesizer unit (13); the target visible light transmitted through the target light beam synthesizer unit (13) and the target infrared light reflected through the target light beam synthesizer unit (13) are synthesized into a beam of light as an incident target light output.

3. The high-precision imaging performance testing device for multi-light fusion according to claim 2, characterized in that: The background light source simulation system (2) comprises: a background integrating sphere light source (21), a background black body unit (22), and a background light beam synthesizer unit (23); the background visible light generated by the background integrating sphere light source (21) is transmitted through the background light beam synthesizer unit (23); the background infrared light generated by the background black body unit (22) is reflected through the background light beam synthesizer unit (23); the background visible light transmitted through the background light beam synthesizer unit (23) and the background infrared light reflected through the background light beam synthesizer unit (23) are synthesized into a beam of light as the incident background light output.

4. The high-precision imaging performance testing device for multi-light fusion according to claim 3, characterized in that: The optical collimation system (3) comprises: a target unit (31), a plane reflector unit (32), and an off-axis parabolic mirror unit (33); the target unit (31), the plane reflector unit (32), and the off-axis parabolic mirror unit (33) are located on the same optical path, and incident target light and incident background light passing through the target unit (31) are collimated by the plane reflector unit (32) and the off-axis parabolic mirror unit (33) and then incident on the imaging acquisition system (4) in parallel.

5. The high-precision imaging performance testing device for multi-light fusion according to claim 4, characterized in that: The imaging acquisition system (4) comprises a visible light camera unit (41), a thermal imager unit (42), and an industrial computer unit (43); the visible light camera unit (41) and the thermal imager unit (42) respectively transmit incident target light and incident background light obtained after being collimated by the optical collimation system (3) to the industrial computer unit (43), and the industrial computer unit (43) converts the input light signal into a digital signal.

6. The high-precision imaging performance testing device for multi-light fusion according to claim 5, characterized in that: The target unit uses a four-rod target as the simulation target.

7. The high-precision imaging performance testing device for multi-light fusion according to claim 6, characterized in that: The imaging performance test system (5) comprises: an image registration and fusion unit (51), a resolvable spatial frequency test unit (52), a light source control unit (53), a target control unit (54), and a data fitting unit (55); The image registration and fusion unit (51) is connected to the industrial control computer unit (43), and the image registration and fusion unit (51) acquires digital signals of visible light and infrared light in real time, performs registration and fusion, and transmits the fused image to the resolvable spatial frequency test unit (52) to obtain test data; The resolvable spatial frequency test unit (52) is connected to a light source control unit (53), and the light source control unit (53) is used to automatically adjust the target light source simulation system (1) and the background light source simulation system (2) after completing a test, so as to change the brightness contrast and temperature difference between the target and the background; The resolvable spatial frequency test unit (52) is also connected to a target control unit (54), and the target control unit (54) switches the target according to the test requirements; The resolvable spatial frequency test unit (52) is also connected to a data fitting unit (55), and the data fitting unit (55) records the test data in real time and fits it into a curved surface.

8. A testing method for a high-precision imaging performance testing device for multi-light fusion, characterized in that: A high-precision imaging performance testing device for multi-light fusion according to any one of claims 5 to 7, wherein the testing method specifically comprises the following steps: S1, turning on the target integrating sphere light source (11), the target black body unit (12), the background integrating sphere light source (21), the background black body unit (22), the visible light camera unit (41), the thermal imager unit (42), and the industrial computer unit (43); S2, the target beam synthesizer unit (13) fuses the target visible light signal and the target infrared signal to obtain incident target light; A background visible light signal and a background infrared signal are fused by a background beam synthesizer unit (23) to obtain incident background light; The incident target light and the incident background light are transmitted to the target unit; S3, adjusting the plane reflector unit (32) and the off-axis parabolic mirror unit (32) so that the main light passing through the target unit (31) is incident on the visible light camera unit (41) and the thermal imager unit (42) in parallel, and ensuring that the incident angle of the main light of the central field of view on the visible light camera unit (41) and the thermal imager unit (42) is 0°; S4, using the light source control unit (53), according to the use scenario of the camera to be tested, setting the temperature of the background black body unit (22), keeping the temperature of the background black body unit (22) unchanged, and setting the initial temperature difference between the target black body unit (12) and the background black body unit (22) to 1°C; S5. According to the usage scenario of the camera to be tested, the brightness of the background integrating sphere light source (21) is set, and the brightness of the background integrating sphere light source (21) is kept unchanged. By adjusting the power of the target integrating sphere light source (11), the brightness contrast between the target and the background is changed, and the initial brightness contrast is set to 0%; S6, using the visible light camera unit (41) and the thermal imager unit (42) to respectively obtain the optical signals of the visible light part and the infrared part passing through the target, and transmit them to the industrial control computer unit (43), and the industrial control computer unit (43) converts the optical signals into digital signals for storage and arrangement, thereby completing the target image acquisition; the target image includes the visible light image and the infrared image; S7, the image registration and fusion unit (51) performs image registration processing on the visible light image and the infrared image, adjusts the field of view and focal length of the visible light camera unit (41) and the thermal imager unit (42), makes the field of view of the two detectors consistent, and ensures that the test target image is located at the center of the two detectors and has the same size and definition; S8, the image registration and fusion unit (51) fuses the visible light image and the infrared image after image registration processing based on the Laplace pyramid method; S9, after each image fusion is completed, the resolvable spatial frequency test unit (52) determines whether the target of the current spatial frequency can be distinguished based on the BP neural network. If it can be distinguished, the instruction is transmitted to the target control unit (54), and the target with a higher spatial frequency is switched, and the process returns to step S6 until the resolution limit is reached, and the spatial frequency corresponding to the current target is recorded; then step S10 is executed; S10, using the light source control unit (53), keeping the temperature difference between the target black body unit (12) and the background black body unit (22) unchanged, changing the target integrating sphere light source power, so that the brightness contrast between the target and the background changes from 0% to 100%, with a step amplitude of 5%, and returning to step S6 each time the brightness contrast changes, and recording that the brightness contrast no longer changes under the current temperature difference, then executing step S11; the numerical change of the spatial frequency corresponding to the brightness contrast change; S11, using the light source control unit (53), gradually reducing the temperature difference between the target black body unit (12) and the background black body unit (22) by 0.05°C, and repeating step S10 each time the temperature difference is changed until the temperature difference changes from 1°C to -1°C, thereby obtaining a spatial frequency data matrix; S12, the minimum resolvable contrast response spatial frequency changes with the contrast between target brightness and background brightness, and the minimum resolvable temperature difference response spatial frequency changes with the temperature difference between target and background; therefore, the common dependent variable spatial frequency is set as the z-axis, the temperature difference between target and background is set as the x-axis, and the contrast between target brightness and background brightness is set as the y-axis; S13. Using a data fitting unit, the spatial frequency data matrix is ​​stored in a coordinate space, and a relationship surface in which the spatial frequency changes simultaneously with the minimum resolvable contrast and the minimum resolvable temperature difference is fitted; and based on the relationship surface, the detector is divided into a recognizable area and an unrecognizable area.

9. The testing method of a high-precision imaging performance testing device for multi-light fusion according to claim 8, characterized in that: The implementation process of step S9 is: S91, after reaching the resolution limit, the resolvable spatial frequency testing unit (52) selects a first area corresponding to the target at the center of the field of view; S92, the resolvable spatial frequency testing unit (52) performs binary segmentation processing on the first area based on the grayscale threshold according to the grayscale values ​​of the pixels in the first area selected by the frame; S93, detecting the distribution of pixels with grayscale values ​​of 1 and 0 in the first area, and if the number of pixels with grayscale values ​​of 1 is greater than half of the total number of pixels, inverting the black and white distribution of the image; S94, extracting the coordinates of the pixels with a gray value of 0 from the binary image obtained in step S93, and extracting the pixels with corresponding coordinates in the image fused in step S8, thereby completing the target area extraction; S95, preprocessing the target area image using a weighted moving average function; S96, after completing the weighted moving average function preprocessing, find the maximum pixel point and the minimum pixel point in the first region image, take 7 neighborhoods in the j direction for any pixel point in the image, if the gray value of the pixel point is greater than the neighboring pixel point, mark it as the maximum pixel point, if the gray value of the pixel point is less than the neighboring pixel point, mark it as the minimum pixel point; S97, after extracting the maximum pixel point and the minimum pixel point, calculate the grayscale mean A of the maximum pixel point respectively h and the minimum pixel grayscale mean A l , and calculate the image background grayscale mean A b ; S98, according to A h , A l , A b Calculate the fringe discernibility and mean discernibility of the four-bar target image; S99. After the calculation of the four-bar target image discernibility is completed, the BP neural network is used to train the discernibility features. The input layer of the BP neural network is the spatial frequency, stripe discernibility and mean discernibility of the four-bar target corresponding to the resolution limit. The hidden layer uses the ReLU activation function, and the output layer uses the Sigmoid activation function to output whether the target of the current spatial frequency can be distinguished. The accuracy of the network is observed after each training. CC , when the accuracy A CC When it reaches 100%, the training is stopped and this is used as the training result to determine whether the subsequent test results have reached the resolution limit.

10. The testing method of a high-precision imaging performance testing device for multi-light fusion according to claim 9, characterized in that: According to the grayscale values ​​of the pixels in the first region, the grayscale threshold is calculated using the maximum inter-class variance method.

Images