Multi-spectral camera radiance responsivity calibration device and method

Abstract

The invention provides a multispectral camera radiance responsivity calibration device and method. The device comprises an integrating sphere, a multispectral camera, a three-dimensional adjusting mechanism, a control system and a data acquisition system. The integrating sphere is used for generating uniform spectral radiation brightness; the multispectral camera is used for image acquisition; the three-dimensional adjusting mechanism is used for moving the multispectral camera; the control system is used for controlling the spectrum output brightness of the integrating sphere and controlling the operation of the three-dimensional adjusting mechanism; and the data acquisition system is used for acquiring and storing image data of each channel of the multispectral camera. According to the invention, the multi-spectral camera pixel-by-pixel radiance responsivity calibration method which can be used under all exposure time and gain coefficient conditions is obtained from the measurement signal and exposure time of the multi-spectral camera and the emergent radiance of the integrating sphere; on the premise of ensuring the accuracy of the radiance responsivity calibration of the multispectral camera, the workload can be effectively reduced, and the calibration working efficiency is improved.

Description

Technical Field

[0001] This invention belongs to the field of radiation calibration technology, specifically relating to a multispectral camera radiance responsivity calibration device and a multispectral camera radiance responsivity calibration method. Background Technology

[0002] Visible light (400-700nm) and near-infrared light (700-2500nm) are the core operating bands of multispectral cameras. Through spectral splitting technology, they acquire image data of targets in multiple discrete spectral bands, forming unique spectral reflectance curves. This technical characteristic makes them crucial in the field of camouflaged target reconnaissance: multispectral cameras can not only capture the reflectance characteristics of targets in specific bands, but also reveal the spectral differentiation between camouflaged objects and the real environment by comparing the differences in spectral responses between targets and the background in the visible / near-infrared bands. This breaks through the visual camouflage limitations of traditional visible light reconnaissance, enabling accurate identification and differentiation of concealed targets.

[0003] Radiometric calibration of multispectral cameras is of great significance for reconnaissance of camouflaged targets. Radiometric calibration establishes an accurate relationship between the camera's output signal and the amount of radiation. By obtaining precise radiation intensity values ​​for each band, we can more accurately identify camouflaged targets, analyze their characteristics, and track and monitor them, providing a more accurate, reliable, and effective means for reconnaissance of camouflaged targets.

[0004] Traditional multispectral camera radiometric calibration involves using filters with known transmittance to disperse the light, calculating the energy of each band using the digital number (DN) output value from the integrating sphere, and then establishing a relationship between the camera's measurements and the integrating sphere's output value based on the actual measurements of each band. This traditional method requires collecting data under various camera operating conditions, resulting in a large workload and low efficiency. Furthermore, because it doesn't consider the relationship between exposure time and radiance, using the integrating sphere as a standard light source and establishing a linear relationship between camera output and radiance through a set of fixed exposure times, the radiometric calibration accuracy is low. Summary of the Invention

[0005] The purpose of this invention is to solve the problems of large workload, low efficiency and low accuracy of existing multispectral camera radiometric calibration methods, and to provide a multispectral camera radiance responsivity calibration device and a multispectral camera radiance responsivity calibration method.

[0006] To achieve the above objectives, the technical solution provided by this invention is:

[0007] A radiance responsivity calibration device for a multispectral camera is provided, comprising an integrating sphere, a multispectral camera, a three-dimensional adjustment mechanism, a control system, and a data acquisition system. The integrating sphere is used to generate uniform spectral radiance. The multispectral camera includes five independent spectral channels, each corresponding to a preset wavelength band, for band-specific image acquisition. The three-dimensional adjustment mechanism is used to move the multispectral camera, changing the relative position of each channel in the multispectral camera to the center of the integrating sphere's output port. The control system is used to control the operation of the three-dimensional adjustment mechanism and the output brightness of the integrating sphere under each channel of the multispectral camera. The data acquisition system is used to acquire and store image data from each channel of the multispectral camera, and to process the image data acquired under different output brightness levels for each channel to obtain the radiance responsivity calibration coefficients for each channel.

[0008] Furthermore, each of the five channels of the multispectral camera is equipped with a telescope, filter, optical system and CMOS, and the five channels correspond to the ultraviolet band, blue band, green band, red band and near-infrared band respectively.

[0009] Furthermore, the wavelength range of the ultraviolet band is 350nm~400nm, the wavelength range of the blue band is 410nm~490nm, the wavelength range of the green band is 525nm~575nm, the wavelength range of the red band is 640nm~750nm, and the wavelength range of the near-infrared band is 770nm~900nm.

[0010] A method for calibrating the radiance responsivity of a multispectral camera is also provided, which is used to calibrate the radiance responsivity of a multispectral camera using the aforementioned radiance responsivity calibration device. The method includes the following steps:

[0011] Step 1: Debug the radiance responsivity calibration device, use the control system to control the operation of the three-dimensional adjustment mechanism, and sequentially align the five channels of the multispectral camera with the center of the integrating sphere output end to ensure that the light output from the integrating sphere can fill the camera's field of view. Use the control system to adjust the integrating sphere output to multiple different output brightnesses under the current channel of the multispectral camera.

[0012] Step 2: For the current channel, at each output brightness, use a multispectral camera to acquire multiple sets of image data for the current channel, and take the average gray value of the multiple sets of image data as the original image. Perform non-uniformity correction on the original image to obtain the corrected image.

[0013] Step 3: Based on the corrected image obtained in Step 2, calculate the radiance responsivity calibration coefficient for the current channel, including the following sub-steps:

[0014] Step 3.1: For each output brightness, calculate the radiance value in the current channel band based on the corrected image, collect measurement data for multiple exposure times and corresponding background noise data, and calculate multiple sets of radiant flux respectively. Based on the radiant flux and radiance value, establish the response relationship of the current output brightness by linear fitting.

[0015] Step 3.2: Fit the radiant flux and radiant values ​​obtained under multiple output brightness conditions using a quadratic function to obtain the curve showing the relationship between radiant flux and output brightness;

[0016] Step 3.3: Based on the multispectral camera response value, gain value, and radiant flux, obtain the radiant responsivity calibration coefficient for the current channel;

[0017] Step 4: Repeat steps 2-3 until the radiance responsivity calibration of all five channels is completed.

[0018] Furthermore, in step 1, the control system adjusts the output brightness of the integrating sphere by controlling the aperture of the lamp inside the integrating sphere, so that the image acquired by the multispectral camera channel has a brightness close to the saturation threshold but not saturated.

[0019] Furthermore, in step 2, the non-uniformity correction of the original image is achieved through the following process: the original image is divided into regions, the mean pixel value of each sub-region is calculated, the ratio of the mean pixel value of the central sub-region to the mean pixel value of other sub-regions is calculated, the non-uniformity correction coefficient of the multispectral camera is obtained, and the original image is corrected using the non-uniformity correction coefficient.

[0020] The advantages of this invention are:

[0021] 1. The radiance responsivity calibration device designed in this invention can calibrate the radiance responsivity of a multispectral camera by collecting the emitted radiance of the integrating sphere through the five channels of the multispectral camera; at the same time, it eliminates the need to collect data under various operating conditions, reducing workload and improving work efficiency.

[0022] 2. The radiance responsivity calibration method provided by this invention starts with the measurement signal and exposure time of the multispectral camera and the emitted radiance of the integrating sphere. It obtains a pixel-by-pixel radiance responsivity calibration method for multispectral cameras that can be used under all exposure times and gain coefficient conditions. Non-uniformity correction is performed on the original image, which improves the stability of the data and reduces the error of radiance responsivity calibration. At the same time, by obtaining the relationship curves between radiant flux and output brightness under multiple output brightness conditions, the workload is effectively reduced and the calibration efficiency is improved while ensuring the accuracy of multispectral camera radiance responsivity calibration. Attached Figure Description

[0023] The features and advantages of the invention will become more readily understood from the following description with reference to the accompanying drawings, in which:

[0024] Figure 1 This is a schematic diagram of the structure of the multispectral camera radiance responsivity calibration device of the present invention;

[0025] Figure 2 This is a flowchart of the multispectral camera radiance responsivity calibration method of the present invention.

[0026] In the diagram: 1-Integrating sphere; 2-Multispectral camera; 3-Three-dimensional adjustment mechanism; 4-Control system; 5-Data acquisition system. Detailed Implementation

[0027] The present invention will now be described in detail with reference to the accompanying drawings and exemplary embodiments thereof. It should be noted that the following detailed description of the present invention is for illustrative purposes only and is not intended to limit the scope of the invention.

[0028] like Figure 1 As shown, the present invention provides a radiance response calibration device for a multispectral camera 2, including an integrating sphere 1, a multispectral camera 2, a three-dimensional adjustment mechanism 3, a control system 4, and a data acquisition system 5.

[0029] Integrating sphere 1 is used to generate uniform spectral radiance. In this embodiment, the integrating sphere 1 is a USLR-A12F-XAN2 type from Labsphere, USA. The integrating sphere 1 has an inner diameter of 30.48 cm and an exit diameter of 10.16 cm. The integrating sphere 1 is equipped with two light sources: a halogen lamp and a xenon lamp. The output dynamic range is adjusted by controlling the size of the apertures of the two light sources. The integrating sphere 1 can directly output the radiance value at the corresponding wavelength point in the 350 nm to 1000 nm band, which can meet the radiance response calibration requirements of the multispectral camera 2. With both the halogen lamp and the xenon lamp fully on, the maximum output illuminance is 117,800 lux, exceeding the spatial luminance uniformity of ±1% at the exit end (f / 4).

[0030] like Figure 1As shown, the multispectral camera 2 includes five independent spectral channels, each corresponding to a preset wavelength band for band-specific image acquisition. Each channel is equipped with a telescope, filter, optical system, and CMOS sensor. The five channels correspond to the ultraviolet, blue, green, red, and near-infrared bands, respectively. The wavelength range of the ultraviolet band is 350nm~400nm, the blue band is 410nm~490nm, the green band is 525nm~575nm, the red band is 640nm~750nm, and the near-infrared band is 770nm~900nm. The detector (the photosensitive chip in the multispectral camera 2) is an HT160A CMOS sensor from Hefei Haitu Microelectronics Co., Ltd., with a resolution of [resolution missing]. The pixel size is .

[0031] The three-dimensional adjustment mechanism 3 is used to move the multispectral camera 2 and change the relative position of each channel in the multispectral camera 2 with the center of the output port of the integrating sphere 1. In this embodiment, the three-dimensional adjustment mechanism 3 is a combination of the lifting platform PT-GD425 and the two-dimensional translation stage PT-XY100, with a positioning accuracy of 3µm.

[0032] The computer is equipped with data acquisition software, 3D stage control software, and integrating sphere 1 control software. The control system 4 includes the 3D stage control software and integrating sphere 1 control software, used to control the operation of the 3D adjustment mechanism 3 and the output brightness of the integrating sphere 1 under each channel of the multispectral camera 2. The data acquisition system 5 is data acquisition software used to acquire and store image data from each channel of the multispectral camera 2, and to process the image data acquired under different output brightness levels for each channel to obtain the radiance responsivity calibration coefficients for each channel.

[0033] The radiance responsivity calibration device provided in this embodiment can calibrate the radiance responsivity of the multispectral camera 2 by collecting the emitted radiance of the integrating sphere 1 through the five channels of the multispectral camera 2; at the same time, it does not require the collection of data under various operating conditions, reducing workload and improving work efficiency.

[0034] The present invention also provides a method for calibrating the radiance responsivity of a multispectral camera, used to calibrate the radiance responsivity of a multispectral camera 2 using the aforementioned radiance responsivity calibration device. The method includes the following steps:

[0035] Step 1: Debug the radiance responsivity calibration device. Use the control system 4 to control the operation of the three-dimensional adjustment mechanism 3. Align the five channels of the multispectral camera 2 with the center of the output end of the integrating sphere 1 in sequence to ensure that the output light of the integrating sphere 1 can fill the camera's field of view. Use the control system 4 to adjust the output brightness of the integrating sphere 1 under the current channel of the multispectral camera 2 to have multiple different output brightness.

[0036] Step 2: For the current channel, at each output brightness, multiple sets of image data are acquired using the current channel of the multispectral camera 2, and the average pixel value of the image is taken as the original image. Non-uniformity correction is performed on the original image to obtain the corrected image.

[0037] Step 3: Based on the corrected image obtained in Step 2, calculate the radiance responsivity calibration coefficient for the current channel, including the following sub-steps:

[0038] Step 3.1: For each output brightness, calculate the radiance value in the current channel band based on the corrected image, collect measurement data for multiple exposure times and corresponding background noise data, and calculate multiple sets of radiant flux respectively. Based on the radiant flux and radiance value, establish the response relationship of the current output brightness by linear fitting.

[0039] Step 3.2: Fit the radiant flux and radiant values ​​obtained under multiple output brightness conditions using a quadratic function to obtain the curve showing the relationship between radiant flux and output brightness;

[0040] Step 3.3: Based on the response value, gain value, and radiant flux of multispectral camera 2, obtain the radiance responsivity calibration coefficient for the current channel;

[0041] Step 4: Repeat steps 2-3 until the radiance responsivity calibration of all five channels is completed.

[0042] The following is a description of each step:

[0043] In step 1, the computer control system 4 drives the three-dimensional adjustment mechanism 3 to precisely adjust its horizontal displacement axis (X / Y direction) and height displacement axis (Z direction) to ensure that the optical axis of the current channel of the multispectral camera 2 is precisely aligned with the mechanical axis of the output port of the integrating sphere 1, so that the standard radiation field output by the integrating sphere 1 completely covers the imaging field of view of the channel, and the radiation flux is uniformly distributed within the field of view.

[0044] The constant current power supply system of integrating sphere 1 is activated. After its internal state stabilizes, the incident luminous flux of the light source is controlled via the controller of integrating sphere 1. The output brightness of integrating sphere 1 is adjusted to the linear response region of the multispectral camera channel 2, ensuring that the detector imaging data is within a dynamic range of 60-90% of its full-well capacity. By controlling the aperture of the lamp inside integrating sphere 1, the output brightness of integrating sphere 1 is adjusted so that the image acquired by the multispectral camera channel 2 has a brightness close to the saturation threshold but not saturated.

[0045] In step 2, the non-uniformity correction of the original image is achieved through the following process: the original image is divided into regions, the mean pixel value of each sub-region is calculated, and the ratio of the mean pixel value of the central sub-region to the mean pixel value of other sub-regions is calculated to obtain the non-uniformity correction coefficient of the multispectral camera 2. The non-uniformity correction coefficient is then used to correct the original image. The light response of the chip center and edges of the multispectral camera 2 is non-uniform, with a large response at the center and a small response at the edges. Therefore, non-uniformity correction is first used for correction before subsequent calibration calculations are performed.

[0046] Under the current output brightness of integrating sphere 1, in this embodiment, the multispectral camera 2 acquires three sets of data in the current channel, and calculates the average gray value of the three sets of data as the original image A. According to the resolution of CMOS, the original image A is divided into n sub-regions, and the mean gray value of all pixels in each sub-region is calculated. Taking the mean gray value of all pixels in the central sub-region as the benchmark, the ratio of the mean gray value of all pixels in the central sub-region to the mean gray value of all pixels in the other sub-regions is calculated to form matrix B. According to the bilinear interpolation method, formulas (1) to (3), the ratio of the mean gray value of all pixels in the central sub-region to the mean gray value of all pixels in the other sub-regions is calculated. The results are reconstructed into a ratio matrix of the same size as the original image A, and the non-uniformity correction coefficient matrix B of CMOS per pixel is obtained. 1 The corrected image A is obtained. 1 =A*B 1 .

[0047] (1)

[0048] (2)

[0049] (3)

[0050] in , , , Representing the coordinates of the points respectively , , , The correction factor value at the location, Indicates in fixed At that time, along Intermediate correction coefficients obtained through directional interpolation. Indicates in fixed At that time, along Intermediate correction coefficients obtained through directional interpolation. This indicates the target pixel. The final correction coefficients (i.e., the interpolation results) at the point.

[0051] In step 3, the radiance value L within the band range of each channel is calculated, and measurement data C1, C2, and C3 are collected for the channel under the current conditions at three different exposure times; then, integrating sphere 1 is turned off, and background noise data C is collected under the same three exposure time conditions as above. 1 off C 2 off C 3 off Dividing by the corresponding exposure time tint, the radiative flux is expressed as: , tint represents the value of the measurement data minus the background noise data under the same exposure time, and tint represents the exposure time. Based on the three obtained radiation fluxes, a three-point fitting is performed to obtain the response curve of the multispectral camera 2 channel under the output brightness of the integrating sphere 1, as shown in formula (4).

[0052] (4)

[0053] in This represents the initial radiative flux of the channel. Indicates the exposure time. This represents the slope of the response curve, i.e., the rate of change of the initial radiative flux with exposure time. This represents the intercept of the response curve, which is the radiant flux value when the exposure time is 0.

[0054] Then, seven different output brightness values ​​are set for integrating sphere 1 to obtain eight corresponding radiative fluxes and the output brightness of integrating sphere 1. The eight sets of data are then subjected to a second fitting operation to obtain the corresponding curve of radiative flux and output brightness, as shown in formula (5).

[0055] (5)

[0056] in This represents the radiative flux of the channel. This represents the calibration coefficient for the 2-channel radiance responsivity of a multispectral camera.

[0057] Based on the corresponding functional relationship between the CMOS response value and gain of the multispectral camera 2, formulas (6) and (7) yield the final radiance responsivity calibration coefficients of channel 1 of the multispectral camera 2. The value;

[0058] (6)

[0059] (7)

[0060] In the formula, Indicates the current gain value. This represents the gain value of the measured data. This represents the radiative flux of the calibrated channel. This represents a response coefficient under the currently applied gain condition. This represents the response coefficient under the gain conditions in actual use.

[0061] In step 4, the horizontal and vertical displacements of the three-dimensional adjustment mechanism 3 are precisely controlled by the three-dimensional stage controller to align channels 2, 3, 4, and 5 of the multispectral camera 2 with the center of the output port of the integrating sphere 1, ensuring that the output light from the integrating sphere 1 fills the field of view of each channel of the multispectral camera 2. Steps 2-3 are repeated until the radiance responsivity calibration of the five channels is completed. The radiance responsivity calibration coefficients of the five channels of the multispectral camera 2 are obtained, thus completing the radiance responsivity calibration of the five channels of the multispectral camera 2.

[0062] The radiance responsivity calibration method provided in this embodiment starts with the measurement signal of the multispectral camera 2 and the exposure time and the emitted radiance of the integrating sphere 1, and obtains a pixel-by-pixel radiance responsivity calibration method for the multispectral camera 2 that can be used under all exposure times and gain coefficients. The traditional method requires collecting data under m exposure times and n gain conditions. This method only requires collecting data from multiple datasets. The calculated coefficients from the set of data are applicable to each exposure time and gain. While ensuring the accuracy of the 2-radius brightness responsivity calibration of the multispectral camera, this method can effectively reduce workload and improve calibration efficiency.

[0063] The invention will now be illustrated with a specific example.

[0064] The multispectral camera 2 includes ultraviolet (350nm~400nm), blue (410nm~490nm), green (525nm~575nm), red (640nm~750nm), and near-infrared (770nm~900nm) bands. The radiance responsivity of the five channels of the multispectral camera 2 was calibrated, and the calibrated radiance values ​​were compared with the radiance values ​​calculated by the integrating sphere 1 according to the bands. The results are shown in Table 1.

[0065] Table 1. Actual output brightness value of integrating sphere 1 and fitted value of radiant brightness after calibration (w / m²) 2 / sr)

[0066]

[0067] As can be seen from the table, the average errors of the five channels of the multispectral camera 2 in the channel radiometric calibration under the eight output brightness values ​​of the integrating sphere 1 are 2.49%, 3.32%, 2.51%, 1.81%, and 1.67%, respectively, indicating that the calibration accuracy of the device and method proposed in this invention is high.

[0068] Finally, it should be noted that the features mentioned and / or shown in the above description of exemplary embodiments of the present invention can be combined in the same or similar manner with one or more other embodiments, combined with features in other embodiments, or substituted for corresponding features in other embodiments. These combined or substituted technical solutions should also be considered to be included within the scope of protection of the present invention.

Claims

1. A radiance responsivity calibration device for a multispectral camera, characterized in that, include: Integrating sphere, used to generate uniform spectral radiance; The multispectral camera includes five independent spectral channels, each corresponding to a preset wavelength band, for band-specific image acquisition. A three-dimensional adjustment mechanism is used to move the multispectral camera and change the relative position of each channel in the multispectral camera with the center of the integrating sphere output port. The control system is used to control the operation of the three-dimensional adjustment mechanism and the output brightness of the integrating sphere in each channel of the multispectral camera. The system also includes a data acquisition system for acquiring and storing image data from each channel of the multispectral camera, and for processing the image data acquired under different output brightness levels in each channel to obtain the radiance response calibration coefficients for each channel.

2. The radiance responsivity calibration device according to claim 1, characterized in that, Each of the five channels of the multispectral camera is equipped with a telescope, filter, optical system and CMOS. The five channels correspond to the ultraviolet band, blue band, green band, red band and near-infrared band respectively.

3. The radiance responsivity calibration device according to claim 2, characterized in that, The wavelength range of the ultraviolet band is 350nm~400nm, the wavelength range of the blue band is 410nm~490nm, the wavelength range of the green band is 525nm~575nm, the wavelength range of the red band is 640nm~750nm, and the wavelength range of the near-infrared band is 770nm~900nm.

4. A method for calibrating the radiance responsivity of a multispectral camera, characterized in that, The method for calibrating the radiance responsivity of a multispectral camera using the radiance responsivity calibration device according to any one of claims 1 to 3 includes the following steps: Step 1: Debug the radiance responsivity calibration device, use the control system to control the operation of the three-dimensional adjustment mechanism, and sequentially align the five channels of the multispectral camera with the center of the integrating sphere output end to ensure that the light output from the integrating sphere can fill the camera's field of view. Use the control system to adjust the integrating sphere output to multiple different output brightnesses under the current channel of the multispectral camera. Step 2: For the current channel, at each output brightness, use a multispectral camera to acquire multiple sets of image data for the current channel, and take the average gray value of the multiple sets of image data as the original image. Perform non-uniformity correction on the original image to obtain the corrected image. Step 3: Based on the corrected image obtained in Step 2, calculate the radiance responsivity calibration coefficient for the current channel, including the following sub-steps: Step 3.1: For each output brightness, calculate the radiance value in the current channel band based on the corrected image, collect measurement data for multiple exposure times and corresponding background noise data, and calculate multiple sets of radiant flux respectively. Based on the radiant flux and radiance value, establish the response relationship of the current output brightness by linear fitting. Step 3.2: Fit the radiant flux and radiant values ​​obtained under multiple output brightness conditions using a quadratic function to obtain the curve showing the relationship between radiant flux and output brightness; Step 3.3: Based on the multispectral camera response value, gain value, and radiant flux, obtain the radiant responsivity calibration coefficient for the current channel; Step 4: Repeat steps 2-3 until the radiance responsivity calibration of all five channels is completed.

5. The radiance responsivity calibration method according to claim 4, characterized in that, In step 1, the control system adjusts the output brightness of the integrating sphere by controlling the aperture of the lamp inside the integrating sphere, so that the image acquired by the multispectral camera channel has a brightness close to the saturation threshold but not saturated.

6. The radiance responsivity calibration method according to claim 4, characterized in that, In step 2, the non-uniformity correction of the original image is achieved through the following process: the original image is divided into regions, the mean pixel value of each sub-region is calculated, the ratio of the mean pixel value of the central sub-region to the mean pixel value of other sub-regions is calculated, the non-uniformity correction coefficient of the multispectral camera is obtained, and the original image is corrected using the non-uniformity correction coefficient.

Images