A method for determining a correction coefficient, a method and device for correcting stripe noise, and an apparatus

Abstract

This application provides a method for determining correction coefficients, a method for correcting stripe noise, an apparatus, and a device, relating to the field of infrared detection technology. The method for determining correction coefficients includes: acquiring the response output of each pixel in a thermal imaging device to a uniform blackbody at a first transmittance and a second transmittance, respectively, for a first temperature and a second temperature; and calculating a first gain matrix and a second gain matrix based on the response outputs acquired at the first transmittance and the second transmittance, respectively, using a two-point correction method; wherein the first temperature is greater than the second temperature and the second transmittance is less than the first transmittance; extracting high-frequency information from the first gain matrix and the second gain matrix in a first direction to obtain first high-frequency information and second high-frequency information; and replacing the second high-frequency information contained in the second gain matrix with the first high-frequency information to obtain a third gain matrix for correcting stripe noise in a second direction. This improves the applicability of the determined correction coefficients.

Description

Technical Field

[0001] This application relates to the field of infrared detection technology, and in particular to a method for determining correction coefficients, a method for correcting stripe noise, an apparatus, and equipment. Background Technology

[0002] Infrared detectors in thermal imaging equipment convert the infrared radiation emitted by an object into electrical signals to obtain an infrared image. Infrared detectors typically use an infrared focal plane array to receive infrared radiation, combined with an integrating readout circuit and an ADC (Analog-to-Digital Converter) to convert the infrared radiation into electrical signals, obtaining the response output of each pixel in the infrared focal plane array to obtain an infrared image. The integrating readout circuit and the ADC can be called signal conversion circuits. Pixels in the same row and / or column share a single signal conversion circuit; that is, the signal conversion circuit is shared at the row and / or column level. Differences often exist between signal conversion circuits in different rows, resulting in inconsistent response outputs from pixels in different rows. That is, pixels in different rows will produce different response outputs to the same infrared radiation, resulting in horizontal lines in the infrared image. Similarly, inconsistent signal conversion circuits in different columns result in vertical lines in the infrared image. Both horizontal and vertical lines can be called stripe noise, and the presence of stripe noise leads to poor image quality in infrared images.

[0003] For example, fringe noise can be corrected using a two-point correction method. Based on the response output of each pixel in the thermal imaging device to two uniform blackbodies at relatively different temperatures, a gain coefficient for compensating for gain-related fringe noise and a offset coefficient for compensating for offset-related fringe noise can be calibrated. Both the gain coefficient and the offset coefficient can be called correction coefficients. Gain-related fringe noise is affected by the intensity of infrared radiation emitted by the object being detected. When the object is at an extremely high temperature, the intensity of the emitted infrared radiation is high, and the intensity of gain-related fringe noise is also high. However, since the temperature of the uniform blackbodies used in calibration cannot reach extremely high temperatures, the calibrated gain coefficients cannot effectively correct for gain-related fringe noise, thus limiting their applicability.

[0004] Therefore, improving the applicability of the determined correction coefficients to effectively correct stripe noise and improve the image quality of infrared images has become an urgent problem to be solved. Summary of the Invention

[0005] The purpose of this application is to provide a method for determining correction coefficients, a method for correcting stripe noise, an apparatus, and a device to improve the applicability of the determined correction coefficients, effectively correct stripe noise, and improve the image quality of infrared images. The specific technical solution is as follows:

[0006] A first aspect of this application provides a method for determining a correction coefficient, the method comprising:

[0007] The response output of each pixel in the thermal imaging device to a uniform blackbody at a first temperature and a second temperature is obtained at a first transmittance. Based on the obtained response output, the gain coefficient of each pixel is calculated according to the two-point correction method to obtain the first gain matrix; wherein the first temperature is greater than the second temperature.

[0008] The response output of each pixel to a uniform blackbody at the first and second temperatures is obtained at the second transmittance, and the gain coefficient of each pixel is calculated according to the two-point correction method based on the obtained response output to obtain the second gain matrix; wherein the second transmittance is less than the first transmittance.

[0009] High-frequency information of the first gain matrix and the second gain matrix are extracted in the first direction to obtain first high-frequency information and second high-frequency information; wherein, the first direction is one of the horizontal and column directions;

[0010] The second high-frequency information contained in the second gain matrix is ​​replaced with the first high-frequency information to obtain a third gain matrix for correcting stripe noise in the second direction; wherein the second direction is another direction other than the first direction in the horizontal and column directions.

[0011] Optionally, the step of extracting high-frequency information from the first gain matrix and the second gain matrix in the first direction to obtain first high-frequency information and second high-frequency information includes:

[0012] The low-frequency information of the first gain matrix and the second gain matrix in the second direction are extracted respectively to obtain the first low-frequency information and the second low-frequency information;

[0013] The first low-frequency information and the second low-frequency information are extracted respectively in the first direction to obtain the third low-frequency information and the fourth low-frequency information;

[0014] Remove the third low-frequency information from the first low-frequency information to obtain the high-frequency information of the first gain matrix in the first direction, which is used as the first high-frequency information; and remove the fourth low-frequency information from the second low-frequency information to obtain the high-frequency information of the second gain matrix in the first direction, which is used as the second high-frequency information.

[0015] Optionally, the step of extracting the low-frequency information of the first gain matrix and the second gain matrix in the second direction to obtain the first low-frequency information and the second low-frequency information includes:

[0016] Calculate the mean value of the gain coefficients in each row of the second direction in the first gain matrix to obtain the elements in the first low-frequency information; calculate the mean value of the gain coefficients in each row of the second direction in the second gain matrix to obtain the elements in the second low-frequency information.

[0017] The step of removing the third low-frequency information from the first low-frequency information to obtain the high-frequency information of the first gain matrix in the first direction, as the first high-frequency information, and removing the fourth low-frequency information from the second low-frequency information to obtain the high-frequency information of the second gain matrix in the first direction, as the second high-frequency information, includes:

[0018] The third low-frequency information in the first low-frequency information is removed, and the removal result is expanded in the second direction to obtain the high-frequency information of the first gain matrix in the first direction, which is used as the first high-frequency information. The fourth low-frequency information in the second low-frequency information is removed, and the removal result is expanded in the second direction to obtain the high-frequency information of the second gain matrix in the first direction, which is used as the second high-frequency information.

[0019] Optionally, the step of extracting the low-frequency information of the first low-frequency information and the second low-frequency information in the first direction to obtain the third low-frequency information and the fourth low-frequency information includes:

[0020] The elements in the first low-frequency information are filtered according to the moving average filtering algorithm to obtain the third low-frequency information.

[0021] The elements in the second low-frequency information are filtered using the moving average filtering algorithm to obtain the fourth low-frequency information.

[0022] Optionally, the gain coefficient of each pixel in the third gain matrix represents: the difference between the gain coefficient of the pixel in the second gain matrix and the corresponding element in the second high-frequency information, and the sum of the gain coefficient of the pixel and the corresponding element in the first high-frequency information.

[0023] or,

[0024] The gain coefficient of each pixel in the third gain matrix represents the product of the ratio of the gain coefficient of that pixel in the second gain matrix to the corresponding element in the second high-frequency information and the corresponding element of that pixel in the first high-frequency information.

[0025] Optionally, after replacing the second high-frequency information contained in the second gain matrix with the first high-frequency information to obtain a third gain matrix for correcting stripe noise in the second direction, the method further includes:

[0026] Using the third gain matrix and the two-point correction method, the response output of each pixel to a uniform blackbody at the first temperature and the second temperature is calculated at the first transmittance. Based on the calculated response output, the gain coefficient of each pixel is calculated according to the two-point correction method to obtain the fourth gain matrix.

[0027] Using the third gain matrix and the two-point correction method, the response output of each pixel to a uniform blackbody at the first temperature and the second temperature is calculated at the second transmittance. Based on the calculated response output, the gain coefficient of each pixel is calculated according to the two-point correction method to obtain the fifth gain matrix.

[0028] High-frequency information in the second direction is extracted from the fourth gain matrix and the fifth gain matrix respectively to obtain the third high-frequency information and the fourth high-frequency information;

[0029] The fourth high-frequency information contained in the fifth gain matrix is ​​replaced with the third high-frequency information to obtain a sixth gain matrix for correcting stripe noise in the first direction.

[0030] Optionally, the first transmittance characterizes the overall transmittance inside the thermal imaging device when the thermal imaging device is not equipped with a lens, or the overall transmittance inside the thermal imaging device when the first lens is equipped; the second transmittance characterizes the overall transmittance inside the thermal imaging device when the second lens is equipped.

[0031] A second aspect of this application also provides a stripe noise correction method, the method comprising:

[0032] Obtain a gain matrix for correcting stripe noise; wherein the obtained gain matrix is ​​obtained based on any of the correction coefficient determination methods described above;

[0033] Using the obtained gain matrix and the two-point correction method, the initial response output detected by each pixel in the thermal imaging device is corrected to obtain the corrected target response output;

[0034] Infrared images are generated using the obtained target response output.

[0035] Optionally, the step of using the acquired gain matrix and the two-point correction method to correct the initial response output detected by each pixel in the thermal imaging device to obtain the corrected target response output includes:

[0036] Using the obtained third gain matrix for correcting stripe noise in the second direction and the two-point correction method, the initial response output detected by each pixel in the thermal imaging device is corrected to obtain the corrected target response output.

[0037] or,

[0038] Using the obtained third gain matrix for correcting stripe noise in the second direction and the two-point correction method, the initial response output detected by each pixel in the thermal imaging device is corrected to obtain the response output after one correction; using the obtained sixth gain matrix for correcting stripe noise in the first direction and the two-point correction method, the response output after one correction is corrected to obtain the corrected target response output.

[0039] A third aspect of this application also provides a correction coefficient determination apparatus, the apparatus comprising:

[0040] The first gain matrix calculation module is used to obtain the response output of each pixel in the thermal imaging device to a uniform blackbody at a first temperature and a second temperature under a first transmittance, and to calculate the gain coefficient of each pixel according to the two-point correction method based on the obtained response output to obtain the first gain matrix; wherein, the first temperature is greater than the second temperature.

[0041] The second gain matrix calculation module is used to obtain the response output of each pixel to a uniform blackbody at the first temperature and the second temperature respectively under the second transmittance, and calculate the gain coefficient of each pixel according to the two-point correction method based on the obtained response output to obtain the second gain matrix; wherein, the second transmittance is less than the first transmittance.

[0042] A high-frequency information extraction module is used to extract high-frequency information from the first gain matrix and the second gain matrix in a first direction, respectively, to obtain first high-frequency information and second high-frequency information; wherein, the first direction is one of the horizontal and column directions;

[0043] The third gain matrix determination module is used to replace the second high-frequency information contained in the second gain matrix with the first high-frequency information to obtain a third gain matrix for correcting stripe noise in the second direction; wherein, the second direction is another direction other than the first direction in the horizontal and column directions.

[0044] Optionally, the high-frequency information extraction module includes:

[0045] The first extraction submodule is used to extract the low-frequency information of the first gain matrix and the second gain matrix in the second direction, respectively, to obtain the first low-frequency information and the second low-frequency information.

[0046] The second extraction submodule is used to extract the low-frequency information of the first low-frequency information and the second low-frequency information in the first direction, respectively, to obtain the third low-frequency information and the fourth low-frequency information.

[0047] The high-frequency information determination submodule is used to remove the third low-frequency information from the first low-frequency information to obtain the high-frequency information of the first gain matrix in the first direction, which is used as the first high-frequency information, and to remove the fourth low-frequency information from the second low-frequency information to obtain the high-frequency information of the second gain matrix in the first direction, which is used as the second high-frequency information.

[0048] Optionally, the first extraction submodule is specifically used to calculate the mean of the gain coefficients of each row in the second direction of the first gain matrix to obtain the elements in the first low-frequency information; and to calculate the mean of the gain coefficients of each row in the second direction of the second gain matrix to obtain the elements in the second low-frequency information.

[0049] The high-frequency information determination submodule is specifically used to remove the third low-frequency information from the first low-frequency information and expand the removal result in the second direction to obtain the high-frequency information of the first gain matrix in the first direction, which is used as the first high-frequency information; and to remove the fourth low-frequency information from the second low-frequency information and expand the removal result in the second direction to obtain the high-frequency information of the second gain matrix in the first direction, which is used as the second high-frequency information.

[0050] Optionally, the second extraction submodule is specifically used to filter the elements in the first low-frequency information according to the moving average filtering algorithm to obtain the third low-frequency information; and to filter the elements in the second low-frequency information according to the moving average filtering algorithm to obtain the fourth low-frequency information.

[0051] Optionally, the gain coefficient of each pixel in the third gain matrix represents: the difference between the gain coefficient of the pixel in the second gain matrix and the corresponding element in the second high-frequency information, and the sum of the difference ...

[0052] Optionally, the device further includes:

[0053] The fourth gain matrix calculation module is used to calculate the response output of each pixel to a uniform blackbody at the first temperature and the second temperature respectively under the first transmittance after replacing the second high-frequency information contained in the second gain matrix with the first high-frequency information to obtain a third gain matrix for correcting stripe noise in the second direction, using the third gain matrix and the two-point correction method, and based on the calculated response output, calculate the gain coefficient of each pixel according to the two-point correction method to obtain the fourth gain matrix.

[0054] The fifth gain matrix calculation module is used to calculate the response output of each pixel to a uniform blackbody at the first temperature and the second temperature under the second transmittance using the third gain matrix and the two-point correction method, and to calculate the gain coefficient of each pixel according to the two-point correction method based on the calculated response output to obtain the fifth gain matrix.

[0055] The high-frequency information determination module is used to extract the high-frequency information in the second direction from the fourth gain matrix and the fifth gain matrix respectively, to obtain the third high-frequency information and the fourth high-frequency information;

[0056] The sixth gain matrix determination module is used to replace the fourth high-frequency information contained in the fifth gain matrix with the third high-frequency information to obtain a sixth gain matrix for correcting stripe noise in the first direction.

[0057] Optionally, the first transmittance characterizes the overall transmittance inside the thermal imaging device when the thermal imaging device is not equipped with a lens, or the overall transmittance inside the thermal imaging device when the first lens is equipped; the second transmittance characterizes the overall transmittance inside the thermal imaging device when the second lens is equipped.

[0058] A fourth aspect of this application also provides a stripe noise correction device, the device comprising:

[0059] A gain matrix acquisition module is used to acquire a gain matrix for correcting stripe noise; wherein the acquired gain matrix is ​​obtained based on any of the correction coefficient determination methods described above.

[0060] The response output correction module is used to correct the initial response output detected by each pixel in the thermal imaging device using the acquired gain matrix and the two-point correction method, so as to obtain the corrected target response output.

[0061] The infrared image generation module is used to generate an infrared image using the obtained target response output.

[0062] Optionally, the response output correction module is specifically used to correct the initial response output detected by each pixel in the thermal imaging device using the obtained third gain matrix for correcting stripe noise in the second direction and the two-point correction method, so as to obtain the corrected target response output.

[0063] or,

[0064] Using the obtained third gain matrix for correcting stripe noise in the second direction and the two-point correction method, the initial response output detected by each pixel in the thermal imaging device is corrected to obtain the response output after one correction; using the obtained sixth gain matrix for correcting stripe noise in the first direction and the two-point correction method, the response output after one correction is corrected to obtain the corrected target response output.

[0065] Another aspect of this application provides an electronic device, comprising:

[0066] Memory, used to store computer programs;

[0067] When a processor executes a program stored in a memory, it implements either the correction coefficient determination method or the stripe noise correction method described above.

[0068] This application also provides a computer-readable storage medium storing a computer program, which, when executed by a processor, implements any of the above-described methods for determining correction coefficients or for correcting stripe noise.

[0069] This application also provides a computer program product containing instructions that, when run on a computer, cause the computer to execute any of the above-described correction coefficient determination methods or any of the above-described stripe noise correction methods.

[0070] Beneficial effects of the embodiments in this application:

[0071] This application provides a method for determining correction coefficients, which involves acquiring the response output of each pixel in a thermal imaging device to a uniform blackbody at a first temperature and a second temperature, respectively, under a first transmittance. Based on the acquired response output, the gain coefficient of each pixel is calculated using a two-point correction method to obtain a first gain matrix; wherein the first temperature is greater than the second temperature. The method also involves acquiring the response output of each pixel to a uniform blackbody at a second transmittance, respectively, under a first temperature and a second temperature, and calculating the gain coefficient of each pixel using a two-point correction method to obtain a second gain matrix; wherein the second transmittance is less than the first transmittance. High-frequency information is extracted from the first and second gain matrices in a first direction to obtain first high-frequency information and second high-frequency information; wherein the first direction is one of the transverse and column directions. The second high-frequency information contained in the second gain matrix is ​​replaced with the first high-frequency information to obtain a third gain matrix for correcting stripe noise in the second direction; wherein the second direction is another direction besides the first direction among the transverse and column directions.

[0072] Based on the above processing, for the same temperature, the higher the transmittance of the thermal imaging device, the greater the intensity of infrared radiation received by each pixel. Correspondingly, by detecting uniform blackbodies at first and second temperatures at a higher transmittance (i.e., the first transmittance), the detection of ultra-high temperature objects at a lower transmittance (i.e., the second transmittance) can be simulated. Consequently, the response output of each pixel at the first transmittance can effectively represent the response output when detecting ultra-high temperature objects, and thus effectively reflect the fringe noise in the gain aspect, which is greatly affected by infrared radiation intensity. In this case, the high-frequency information (i.e., the first high-frequency information) in the first direction of the calculated gain matrix (i.e., the first gain matrix) can effectively compensate for the fringe noise in the second direction when detecting ultra-high temperature objects.

[0073] Because ultra-high temperature objects emit high levels of infrared radiation, the transmittance of thermal imaging devices is often low in practical applications to avoid infrared radiation saturation in each pixel. Correspondingly, the response output of each pixel in the thermal imaging device at a second transmittance effectively reflects the impact of lower transmittance on the response output of each pixel. In this case, the low-frequency information in the calculated gain matrix (i.e., the second gain matrix) can effectively compensate for low-frequency noise when detecting ultra-high temperature objects.

[0074] Therefore, the second high-frequency information in the second gain matrix can be replaced with the first high-frequency information to obtain the third gain matrix. Subsequently, using the third gain matrix, it is possible to effectively compensate for stripe noise in the second direction while effectively detecting ultra-high temperature objects, and also effectively compensate for the impact of low transmittance on the response output of each pixel. That is, the third gain matrix is ​​applicable to correcting stripe noise with high intensity in the gain aspect. This improves the applicability of the determined correction coefficients, effectively correcting stripe noise and improving the quality of the infrared image.

[0075] Of course, implementing any product or method of this application does not necessarily require achieving all of the advantages described above at the same time. Attached Figure Description

[0076] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other embodiments can be obtained based on these drawings.

[0077] Figure 1 This is a schematic diagram of a first flowchart of the method for determining correction coefficients provided in an embodiment of this application;

[0078] Figure 2 This is a second flowchart illustrating the method for determining correction coefficients provided in an embodiment of this application.

[0079] Figure 3 A schematic flowchart illustrating a stripe noise correction method provided in an embodiment of this application;

[0080] Figure 4 A schematic flowchart illustrating another stripe noise correction method provided in an embodiment of this application;

[0081] Figure 5a A schematic diagram of an infrared image generated using the initial response output detected by each pixel, provided in an embodiment of this application;

[0082] Figure 5b A schematic diagram of an infrared image generated using the corrected target response output, provided as an embodiment of this application;

[0083] Figure 6 This is a schematic diagram of a correction coefficient determination device provided in an embodiment of this application;

[0084] Figure 7 This is a schematic diagram of the structure of a stripe noise correction device provided in an embodiment of this application;

[0085] Figure 8 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. Detailed Implementation

[0086] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art based on this application are within the scope of protection of this application.

[0087] Objects with temperatures above absolute zero (-273°C) emit infrared radiation. Infrared detectors in thermal imaging devices convert this radiation into electrical signals to obtain an infrared image. The infrared radiation emitted by an object passes through the lens of the thermal imaging device and is received by the infrared focal plane array of the infrared detector. Combined with a signal conversion circuit, the infrared radiation is converted into electrical signals, resulting in the response output of each pixel in the infrared focal plane array, thus producing an infrared image.

[0088] To address the issue of stripe noise and low image quality in infrared images obtained from thermal imaging equipment, a two-point correction method can be used to correct the stripe noise. However, the calibrated gain coefficient cannot effectively correct stripe noise with high intensity, thus limiting its applicability.

[0089] To improve the applicability of the determined correction coefficients, effectively correct stripe noise, and improve the image quality of infrared images, embodiments of this application provide a method for determining correction coefficients. This method can be applied to electronic devices. For example, the electronic device can be a processor in a thermal imaging device, which can execute steps S101-S104 to obtain a third gain matrix for correcting stripe noise in a second direction. See also... Figure 1 , Figure 1 This is a schematic flowchart of a first method for determining correction coefficients provided in an embodiment of this application. The method may include the following steps:

[0090] Step S101: Obtain the response output of each pixel in the thermal imaging device for a uniform blackbody at the first and second temperatures under the first transmittance, and calculate the gain coefficient of each pixel according to the two-point correction method based on the obtained response output to obtain the first gain matrix.

[0091] The first temperature is greater than the second temperature.

[0092] Step S102: Obtain the response output of each pixel to a uniform blackbody at the first and second temperatures under the second transmittance, and calculate the gain coefficient of each pixel according to the two-point correction method based on the obtained response output to obtain the second gain matrix.

[0093] The second transmittance is less than the first transmittance.

[0094] Step S103: Extract the high-frequency information of the first gain matrix and the second gain matrix in the first direction respectively to obtain the first high-frequency information and the second high-frequency information.

[0095] The first direction is either the horizontal or the vertical direction.

[0096] Step S104: Replace the second high-frequency information contained in the second gain matrix with the first high-frequency information to obtain a third gain matrix for correcting stripe noise in the second direction.

[0097] The second direction is the other direction besides the first direction in the horizontal and column directions.

[0098] Based on the above processing, for the same temperature, the higher the transmittance of the thermal imaging device, the greater the intensity of infrared radiation received by each pixel. Correspondingly, by detecting uniform blackbodies at first and second temperatures at a higher transmittance (i.e., the first transmittance), the detection of ultra-high temperature objects at a lower transmittance (i.e., the second transmittance) can be simulated. Consequently, the response output of each pixel at the first transmittance can effectively represent the response output when detecting ultra-high temperature objects, and thus effectively reflect the fringe noise in the gain aspect, which is greatly affected by infrared radiation intensity. In this case, the high-frequency information (i.e., the first high-frequency information) in the first direction of the calculated gain matrix (i.e., the first gain matrix) can effectively compensate for the fringe noise in the second direction when detecting ultra-high temperature objects.

[0099] Because ultra-high temperature objects emit high levels of infrared radiation, the transmittance of thermal imaging devices is often low in practical applications to avoid infrared radiation saturation in each pixel. Correspondingly, the response output of each pixel in the thermal imaging device at a second transmittance effectively reflects the impact of lower transmittance on the response output of each pixel. In this case, the low-frequency information in the calculated gain matrix (i.e., the second gain matrix) can effectively compensate for low-frequency noise when detecting ultra-high temperature objects.

[0100] Therefore, the second high-frequency information in the second gain matrix can be replaced with the first high-frequency information to obtain the third gain matrix. Subsequently, using the third gain matrix, it is possible to effectively compensate for stripe noise in the second direction while effectively detecting ultra-high temperature objects, and also effectively compensate for the impact of low transmittance on the response output of each pixel. That is, the third gain matrix is ​​applicable to correcting stripe noise with high intensity in the gain aspect. This improves the applicability of the determined correction coefficients, effectively correcting stripe noise and improving the quality of the infrared image.

[0101] For steps S101 and S102, two uniform blackbodies at different temperatures can be preset, namely, a first temperature and a second temperature. The first temperature is higher than the second temperature; for example, the first temperature can be 50°C or 60°C, and the second temperature can be 10°C or 15°C. The transmittance of the thermal imaging device can be set to a first transmittance and a second transmittance, respectively. The second transmittance is lower than the first transmittance.

[0102] For the same temperature, the higher the transmittance of a thermal imaging device, the greater the intensity of infrared radiation received by each pixel. Correspondingly, by detecting a uniform blackbody at a first and second temperature under higher transmittance (i.e., the first transmittance), the detection of an ultra-high temperature object under lower transmittance (i.e., the second transmittance) can be simulated. The response output of each pixel at the first transmittance can effectively represent the response output when detecting an ultra-high temperature object, and thus effectively reflect the stripe noise in the gain aspect, which is greatly affected by the intensity of infrared radiation. Stripe noise is a high-frequency noise.

[0103] Because ultra-high temperature objects emit high levels of infrared radiation, the transmittance of thermal imaging devices is often low in practical applications when measuring ultra-high temperature objects to avoid saturation of infrared radiation received by each pixel. Correspondingly, the response output of each pixel in the thermal imaging device at the second transmittance effectively reflects the impact of lower transmittance on the response output of each pixel; that is, it effectively reflects low-frequency noise, but cannot effectively reflect stripe noise in the gain aspect, which is significantly affected by infrared radiation intensity.

[0104] The first and second transmittance values ​​characterize the overall transmittance within a thermal imaging device. For example, the overall transmittance within the thermal imaging device can be adjusted by modifying the lens and / or the optical parameters within the device itself. These optical parameters may include aperture and exposure time; a larger aperture and a longer exposure time result in higher overall transmittance.

[0105] In one implementation, the first transmittance characterizes the overall transmittance inside the thermal imaging device when the thermal imaging device is not equipped with a lens, or the overall transmittance inside the thermal imaging device when the first lens is equipped; the second transmittance characterizes the overall transmittance inside the thermal imaging device when the second lens is equipped.

[0106] In this implementation, the thermal imaging device may not be equipped with a lens, or a first lens may be provided to achieve a first transmittance level for the overall internal transmittance of the thermal imaging device. Alternatively, a second lens may be provided to achieve a second transmittance level for the overall internal transmittance of the thermal imaging device. The first and second lenses can be made of different materials, or they can be made of the same material. The aperture of the first lens is larger than that of the second lens. The first lens can be called a high transmittance lens, and the second lens can be called a low transmittance lens or an ultra-high temperature lens. For example, the aperture of the first lens can be F1.0, and the aperture of the second lens can be F4.0.

[0107] In this way, the transmittance of the thermal imaging device can be set, and the first transmittance can be guaranteed to be greater than the second transmittance. This ensures that the response output of each pixel in the thermal imaging device under different transmittances can be obtained to calculate the gain matrix.

[0108] At a first transmittance, the thermal imaging device detects uniform blackbodies at a first temperature and a second temperature, respectively. Correspondingly, the response output of each pixel in the thermal imaging device at the first transmittance, targeting the uniform blackbodies at the first and second temperatures, can be acquired. Based on the acquired response outputs, the gain coefficient of each pixel can be calculated using the two-point correction method, resulting in a first gain matrix. The coordinates of any pixel in the infrared focal plane array correspond to the position of its gain coefficient in the gain matrix. For example, for a pixel with coordinates (i,j) in the infrared focal plane array—that is, the pixel in the i-th row and j-th column of the infrared focal plane array—the gain coefficient of that pixel can be located in the i-th row and j-th column of the gain matrix.

[0109] The two-point correction response model of the infrared detector is as follows:

[0110] Y i,j =G i,j ·K i,j ·X i,j +C i,j

[0111] Among them, Y i,j G represents the response output of the pixel with coordinates (i,j) in the infrared focal plane array after correction using the two-point correction method. i,j G represents the initial gain coefficient of the pixel with coordinates (i,j). i,j =1,K i,j X represents the non-uniform correction coefficient for the pixel with coordinates (i,j), i.e., the gain coefficient of the pixel with coordinates (i,j) in the gain matrix (which can be represented as K). i,j C represents the initial response output detected by the pixel at coordinates (i,j). i,j is the offset coefficient of the pixel with coordinates (i,j).

[0112] The gain coefficient of each pixel can be calculated using the two-point correction method according to the following formula:

[0113]

[0114] Among them, K i,j H represents the gain coefficient of the pixel with coordinates (i,j) in the gain matrix. i,j L represents the response output of the pixel at coordinates (i,j) to a uniform blackbody at the first temperature. i,jThis represents the response output of the pixel at coordinates (i,j) to a uniform blackbody at the second temperature, where I and J represent the number of rows and columns of pixels in the infrared focal plane array, respectively.

[0115] Based on the above method, the first gain matrix can be calculated. Similarly, the second gain matrix can be calculated under the second transmittance. The execution order of steps S101 and S102 can be set as needed and is not limited.

[0116] For steps S103 and S104, the first direction can be horizontal or column-oriented, and the second direction can be either horizontal or column-oriented, excluding the first direction. For example, the first direction can be horizontal, and the second direction can be column-oriented; or, the first direction can be column-oriented, and the second direction can be horizontal. High-frequency information from the first gain matrix and the second gain matrix in the first direction can be extracted to obtain first high-frequency information and second high-frequency information. The high-frequency information can also be in matrix form. The extraction method for high-frequency information can be found in the descriptions of Method 1 and Method 2 in subsequent embodiments.

[0117] The high-frequency information (first high-frequency information) in the first direction of the gain matrix (i.e., the gain matrix calculated under high transmittance conditions) can effectively compensate for fringe noise in the second direction when detecting ultra-high temperature objects. The low-frequency information in the gain matrix (i.e., the second gain matrix) calculated under low transmittance conditions can effectively compensate for low-frequency noise when detecting ultra-high temperature objects. However, since the temperature of the uniform blackbody used to calibrate the second gain matrix cannot reach ultra-high temperatures, the response output of each pixel in the thermal imaging device at low transmittance (i.e., the second transmittance) cannot effectively represent the response output when detecting ultra-high temperature objects. In this case, the high-frequency information (i.e., the second high-frequency information) in the calculated second gain matrix cannot effectively compensate for fringe noise in the second direction when detecting ultra-high temperature objects.

[0118] Therefore, the second high-frequency information contained in the second gain matrix can be replaced with the first high-frequency information to obtain a third gain matrix for correcting the fringe noise in the second direction. Subsequently, by using the third gain matrix, it is possible to effectively compensate for the fringe noise in the second direction while effectively detecting ultra-high temperature objects, and also effectively compensate for the impact of low transmittance on the response output of each pixel.

[0119] In one embodiment, the gain coefficient of each pixel in the third gain matrix represents: the difference between the gain coefficient of the pixel in the second gain matrix and the corresponding element in the second high-frequency information, and the sum of the gain coefficient of the pixel and the corresponding element in the first high-frequency information; or, the gain coefficient of each pixel in the third gain matrix represents: the ratio of the gain coefficient of the pixel in the second gain matrix to the corresponding element in the second high-frequency information, and the product of the gain coefficient of the pixel and the corresponding element in the first high-frequency information.

[0120] In this embodiment, the second high-frequency information contained in the second gain matrix can be replaced with the first high-frequency information by addition, subtraction, multiplication, or division to obtain the third gain matrix. This can also be described as obtaining the third gain matrix through a masking scheme.

[0121] When using addition and subtraction operations, for each pixel, the difference between the pixel's gain coefficient in the second gain matrix and its corresponding element in the second high-frequency information can be calculated. The sum of this difference and the corresponding element in the first high-frequency information can then be calculated to obtain the pixel's gain coefficient in the third gain matrix. Alternatively, the sum of the pixel's gain coefficient in the second gain matrix and its corresponding element in the first high-frequency information can be calculated first, and the difference between this sum and the corresponding element in the second high-frequency information can then be calculated to obtain the pixel's gain coefficient in the third gain matrix. The order of addition and subtraction is not restricted, as long as the obtained gain coefficient in the third gain matrix can represent the pixel's gain coefficient in the third gain matrix as the sum of the difference between the pixel's gain coefficient in the second gain matrix and its corresponding element in the second high-frequency information, and the sum of the difference ...

[0122] When using multiplication and division operations, for each pixel, the ratio of its gain coefficient in the second gain matrix to its corresponding element in the second high-frequency information can be calculated. The product of this ratio and the corresponding element in the first high-frequency information is then calculated to obtain the pixel's gain coefficient in the third gain matrix. Alternatively, the product of the pixel's gain coefficient in the second gain matrix and its corresponding element in the first high-frequency information can be calculated first, and the ratio of this product to the corresponding element in the second high-frequency information is then calculated to obtain the pixel's gain coefficient in the third gain matrix. The order of multiplication and division is not restricted, as long as the obtained gain coefficient in the third gain matrix can be represented as the product of the ratio of the pixel's gain coefficient in the second gain matrix to its corresponding element in the second high-frequency information and the corresponding element in the first high-frequency information.

[0123] Based on the above processing, the second high-frequency information contained in the second gain matrix can be replaced with the first high-frequency information to obtain a third gain matrix for correcting fringe noise in the second direction. Subsequently, using the third gain matrix, it is possible to effectively correct fringe noise in the second direction while effectively detecting ultra-high temperature objects, and simultaneously compensate for the impact of low transmittance on the response output of each pixel. In other words, the third gain matrix is ​​suitable for correcting fringe noise with high intensity in the gain direction.

[0124] In addition, compared to addition and subtraction, multiplication and division amplify the noise information in the first and second high-frequency information. Therefore, the first and second high-frequency information obtained by addition and subtraction are less affected by noise information, which can further improve the effectiveness of the third gain matrix obtained by replacing the second high-frequency information in the second gain matrix with the first high-frequency information. This further ensures that the applicability of the determined correction coefficients can be improved, so as to effectively correct stripe noise and improve the quality of infrared images.

[0125] In this application, the high-frequency information contained in the gain matrix can be extracted in the following ways:

[0126] In Method 1, low-frequency information from the first and second gain matrices in the first direction can be extracted separately to obtain fifth and sixth low-frequency information. For example, for any gain matrix in the first and second gain matrices, filtering can be performed in the first direction using a first filtering method to obtain the low-frequency information of that gain matrix in the first direction. For example, the first filtering method can be mean filtering, moving average filtering, or Gaussian filtering. The fifth low-frequency information in the first gain matrix can be removed to obtain the high-frequency information of the first gain matrix in the first direction, i.e., the first high-frequency information. Similarly, the sixth low-frequency information in the second gain matrix can be removed to obtain the high-frequency information of the second gain matrix in the first direction, i.e., the second high-frequency information. For example, the difference between the first gain matrix and the fifth low-frequency information can be calculated to obtain the first high-frequency information, and the difference between the second gain matrix and the sixth low-frequency information can be calculated to obtain the second high-frequency information.

[0127] In method two, low-frequency information from the first and second gain matrices in the second direction can be extracted. This extracted low-frequency information is then processed to obtain high-frequency information from the first and second gain matrices in the first direction. (See also...) Figure 2 , Figure 2 A second flowchart illustrating the correction coefficient determination method provided in this application embodiment. Step S103 includes:

[0128] Step S1031: Extract the low-frequency information of the first gain matrix and the second gain matrix in the second direction respectively to obtain the first low-frequency information and the second low-frequency information.

[0129] Step S1032: Extract the low-frequency information of the first low-frequency information and the second low-frequency information in the first direction respectively to obtain the third low-frequency information and the fourth low-frequency information.

[0130] Step S1033: Remove the third low-frequency information from the first low-frequency information to obtain the high-frequency information of the first gain matrix in the first direction, which is used as the first high-frequency information; and remove the fourth low-frequency information from the second low-frequency information to obtain the high-frequency information of the second gain matrix in the first direction, which is used as the second high-frequency information.

[0131] In this embodiment, low-frequency information of the first gain matrix and the second gain matrix in the second direction can be extracted to obtain first low-frequency information and second low-frequency information. For example, for any gain matrix in the first gain matrix and the second gain matrix, filtering processing in the second direction can be performed on the gain matrix according to the second filtering method to obtain the low-frequency information of the gain matrix in the second direction. For example, the second filtering method can be mean filtering, moving average filtering, or Gaussian filtering. Furthermore, low-frequency information of the first low-frequency information and the second low-frequency information in the first direction can be extracted to obtain third low-frequency information and fourth low-frequency information. For example, for the first low-frequency information or the second low-frequency information, filtering processing in the first direction can be performed on the low-frequency information according to the third filtering method to obtain the low-frequency information of the low-frequency information in the first direction. For example, the third filtering method can be mean filtering, moving average filtering, or Gaussian filtering. The first filtering method, the second filtering method, and the third filtering method can be the same or different, and can be selected as needed without specific limitations. Removing the third low-frequency information from the first low-frequency information yields the high-frequency information of the first gain matrix in the first direction, which is used as the first high-frequency information. The fourth low-frequency information is removed from the second low-frequency information to obtain the high-frequency information of the second gain matrix in the first direction, which is used as the second high-frequency information.

[0132] Based on the above processing, by extracting low-frequency information from the first and second gain matrices in the second direction, and then processing the extracted low-frequency information, high-frequency information from the first and second gain matrices in the first direction is obtained. This reduces the influence of high-frequency noise in the gain matrices, further ensuring that the obtained high-frequency information can effectively compensate for stripe noise in the second direction when detecting ultra-high temperature objects, thus improving the accuracy of the obtained high-frequency information. Consequently, this further ensures the accuracy of the subsequently obtained third gain matrix, further improving the applicability of the determined correction coefficients, effectively correcting stripe noise, and improving the quality of the infrared image.

[0133] In one embodiment, step S1031 may include: calculating the mean of the gain coefficients of each row in the second direction of the first gain matrix to obtain the elements in the first low-frequency information; and calculating the mean of the gain coefficients of each row in the second direction of the second gain matrix to obtain the elements in the second low-frequency information.

[0134] Step S1033 may include: removing the third low-frequency information from the first low-frequency information and expanding the removal result in the second direction to obtain the high-frequency information of the first gain matrix in the first direction, which is used as the first high-frequency information; and removing the fourth low-frequency information from the second low-frequency information and expanding the removal result in the second direction to obtain the high-frequency information of the second gain matrix in the first direction, which is used as the second high-frequency information.

[0135] In this embodiment, the second filtering method can be mean filtering. For each row in the second direction of the first gain matrix, the mean value of the gain coefficients in that row can be calculated and used as an element of that row in the first low-frequency information. For each row in the second direction of the second gain matrix, the mean value of the gain coefficients in that row can be calculated and used as an element of that row in the second low-frequency information. When the second direction is horizontal, each row in the second direction is a row. Accordingly, the first low-frequency information and the second low-frequency information can be column matrices, and the number of elements contained in the first low-frequency information and the second low-frequency information is the same as the number of rows in the second gain matrix. When the second direction is column-oriented, each row in the second direction is a column. Accordingly, the first low-frequency information and the second low-frequency information can be row matrices, and the number of elements contained in the first low-frequency information and the second low-frequency information is the same as the number of columns in the second gain matrix.

[0136] Taking the second direction as the column direction as an example, the first gain matrix can be represented as K1, and the second gain matrix can be represented as K2. The first and second low-frequency information can be extracted using the following formula:

[0137]

[0138] Among them, K1L j K1 represents the element in the j-th column of the first low-frequency information. i,j This represents the element in the i-th row and j-th column of the first gain matrix; I and J represent the row and column numbers of the pixels in the infrared focal plane array, respectively. K2L j This represents the element in the j-th column of the second low-frequency information; K2 i,j This represents the element in the i-th row and j-th column of the second gain matrix. Correspondingly, K1L represents the first low-frequency information, and K2L represents the second low-frequency information. The first and second low-frequency information can also be referred to as column-oriented low-frequency matrices.

[0139] In this case, in the first and second low-frequency information, each row in the second direction contains only one element. Similarly, in the third and fourth low-frequency information obtained from the first and second low-frequency information, each row in the second direction also contains only one element. The removal result obtained by removing the third low-frequency information from the first low-frequency information can be called the first removal result, and the removal result obtained by removing the fourth low-frequency information from the second low-frequency information can be called the second removal result. In both the first and second removal results, each row in the second direction contains only one element.

[0140] To ensure that the second high-frequency information contained in the second gain matrix can be replaced with the first high-frequency information, the dimensions of the first and second high-frequency information and the second gain matrix must be consistent. That is, the number of elements in each row of the first and second high-frequency information along the second direction must be the same as the number of gain coefficients in each row of the second direction in the second gain matrix. Therefore, the first and second removal results need to be expanded in the second direction to ensure that the expanded first and second high-frequency information have the same dimensions as the second gain matrix.

[0141] For example, the elements in each row of the first high-frequency information and the second high-frequency information obtained by expansion can be the same in the first direction. If the second direction is horizontal, the first removal result and the second removal result can be column matrices, and the first removal result and the second removal result can be expanded in the column direction respectively, so that the elements in each row of the expanded first high-frequency information and the second high-frequency information are the same. If the second direction is column-oriented, the first removal result and the second removal result can be row matrices, and the first removal result and the second removal result can be expanded in the horizontal direction respectively, so that the elements in each column of the expanded first high-frequency information and the second high-frequency information are the same. For example, if the number of rows of each pixel is I and the number of columns is J, then the second gain matrix is ​​a matrix of I rows and J columns, and the first removal result and the second removal result are row matrices of J columns. The first removal result and the second removal result can be expanded in the horizontal direction according to the following formula:

[0142] K1LH′ i,j =K1LH j i = 1, 2, 3...I

[0143] K2LH' i,j =K2LH j i = 1, 2, 3...I

[0144] Where K1LH and K2LH represent the first removal result and the second removal result, respectively, and K1LH' i,j and K2LH' i,j K1LH represents the element in the i-th row and j-th column of the first high-frequency information and the second high-frequency information, respectively.j This represents the element in the j-th column of the first removal result; K2LH j This represents the element in the j-th column of the second removal result.

[0145] Based on the above processing, mean filtering can be used to extract the first and second low-frequency information. This ensures that by extracting the low-frequency information from the first and second gain matrices in the second direction, and then processing this extracted low-frequency information, the high-frequency information from the first and second gain matrices in the first direction can be obtained. This reduces the influence of high-frequency noise in the gain matrix, further ensuring that the obtained high-frequency information can effectively compensate for stripe noise in the second direction when detecting ultra-high temperature objects, thus improving the accuracy of the obtained high-frequency information. Consequently, this further ensures the accuracy of the subsequently obtained third gain matrix, further improving the applicability of the determined correction coefficients to effectively correct stripe noise and improve the quality of the infrared image.

[0146] In one embodiment, step S1032 may include: filtering the elements in the first low-frequency information according to the moving average filtering algorithm to obtain the third low-frequency information; and filtering the elements in the second low-frequency information according to the moving average filtering algorithm to obtain the fourth low-frequency information.

[0147] In this embodiment, the third filtering method can be moving average filtering. The size of the moving window can be set as needed, without specific limitations. For example, the size of the moving window can be 8 or 10. For each element in the first low-frequency information, according to the size of the moving window, the moving window corresponding to that element in the first direction can be determined, and the mean of each element contained in the moving window corresponding to that element can be calculated to obtain the moving average corresponding to that element, which is used as the element at the position of that element in the third low-frequency information. Taking the first direction as horizontal as an example, the first gain matrix can be represented as K1, and the second gain matrix can be represented as K2. The third low-frequency information and the fourth low-frequency information can be extracted according to the following formula:

[0148]

[0149] Among them, K1LL j This represents the element in the j-th column of the third low-frequency information; K1L j+m This represents the element in the (j+m)th column of the first low-frequency information; n represents the sliding window size. K2LL j This represents the element in the j-th column of the fourth low-frequency information; K2L j+m This represents the element in the (j+m)th column of the second low-frequency information. Correspondingly, K1LL represents the third low-frequency information, and K2LL represents the fourth low-frequency information. The third and fourth low-frequency information can also be referred to as the horizontal low-frequency matrix.

[0150] The first and second removal results can be obtained using the following formula:

[0151] K1LH j =K1L j -K1LL j

[0152] K2LH j =K2L j -K2LL j

[0153] Among them, K1LH j This represents the element in the j-th column of the first removal result; K1L j This represents the element in the j-th column of the first low-frequency information; K1LL j This represents the element in the j-th column of the third low-frequency information. K2LH j This represents the element in the j-th column of the second removal result; K2L j This represents the element in the j-th column of the second low-frequency information; K2LL j This represents the element in the j-th column of the fourth low-frequency information.

[0154] Accordingly, the third gain matrix can be calculated using any of the following formulas:

[0155] Knew i,j =K2 i,j -K2LH′ i,j +K1LH' i,j

[0156] Knew i,j =K2 i,j / K2LH' i,j ×K1LH' i,j

[0157] Among them, Knew i,j K2 represents the gain coefficient in the i-th row and j-th column of the third gain matrix. i,j K1LH' represents the gain coefficient in the i-th row and j-th column of the second gain matrix. i,j and K2LH' i,j These represent the elements in the i-th row and j-th column of the first and second high-frequency information, respectively.

[0158] Based on the above processing, the third and fourth low-frequency information can be extracted using moving average filtering. This ensures that by extracting the low-frequency information from the first and second gain matrices in the second direction, and then processing this extracted low-frequency information, the high-frequency information from the first and second gain matrices in the first direction can be obtained. This reduces the influence of high-frequency noise in the gain matrices, further ensuring that the obtained high-frequency information can effectively compensate for stripe noise in the second direction when detecting ultra-high temperature objects, thus improving the accuracy of the obtained high-frequency information. Consequently, this further ensures the accuracy of the subsequently obtained third gain matrix, further improving the applicability of the determined correction coefficients to effectively correct stripe noise and improve the quality of the infrared image.

[0159] In one embodiment, correction coefficients (i.e., a sixth gain matrix) may also be determined for correcting fringe noise in the first direction. Following step S104, the correction coefficient determination method further includes:

[0160] Step 1: Using the third gain matrix and the two-point correction method, calculate the response output of each pixel to a uniform blackbody at the first and second temperatures respectively under the first transmittance. Based on the calculated response output, calculate the gain coefficient of each pixel according to the two-point correction method to obtain the fourth gain matrix.

[0161] Step 2: Using the third gain matrix and the two-point correction method, calculate the response output of each pixel to a uniform blackbody at the first and second temperatures at the second transmittance. Based on the calculated response output, calculate the gain coefficient of each pixel according to the two-point correction method to obtain the fifth gain matrix.

[0162] Step 3: Extract the high-frequency information in the second direction from the fourth and fifth gain matrices respectively to obtain the third and fourth high-frequency information.

[0163] Step 4: Replace the fourth high-frequency information contained in the fifth gain matrix with the third high-frequency information to obtain the sixth gain matrix used to correct the stripe noise in the first direction.

[0164] In this embodiment, the response output calculated in steps 1 and 2 is the corrected response output obtained by correcting the response output of each pixel at the first transmittance for a uniform blackbody at the first and second temperatures using the third gain matrix according to the two-point correction method. The process of calculating the gain matrix based on the calculated response output in steps 1 and 2 using the two-point correction method is similar to the process of calculating the gain matrix in steps S101 and S102 of the above embodiment; please refer to the relevant descriptions of steps S101-S102 in the above embodiment for details. Steps 3 and 4 are similar to steps S103 and S104 of the above embodiment; please refer to the relevant descriptions of steps S103 and S104 in the above embodiment for details.

[0165] Based on the above processing, a sixth gain matrix can be further determined for correcting stripe noise in the first direction, building upon the third gain matrix used to correct stripe noise in the second direction. Subsequently, the third and sixth gain matrices can be used to correct stripe noise in both directions using a two-point correction method. This further ensures effective stripe noise correction, thereby improving the quality of the infrared image.

[0166] Based on the same inventive concept, embodiments of this application also provide a stripe noise correction method, which can be applied to electronic devices. For example, the electronic device can be a processor in a thermal imaging device. The processor can execute steps S301-S303 to correct the response output detected by each pixel and generate an infrared image based on the corrected response output. Subsequently, the processor can also send the obtained infrared image to a display device for displaying the image, so that the obtained infrared image can be displayed by the display device. Alternatively, the thermal imaging device may include a display module for displaying the image, and the processor can send the obtained infrared image to the display module for display by the display module. See also Figure 3 , Figure 3 This application provides a flowchart illustrating a stripe noise correction method, which includes:

[0167] Step S301: Obtain the gain matrix used to correct stripe noise.

[0168] The obtained gain matrix is ​​derived from any of the correction coefficient determination methods described above.

[0169] Step S302: Using the acquired gain matrix and the two-point correction method, the initial response output detected by each pixel in the thermal imaging device is corrected to obtain the corrected target response output.

[0170] Step S303: Generate an infrared image using the obtained target response output.

[0171] In this embodiment, the gain matrix can be obtained in advance based on the correction coefficient determination method described in any of the above embodiments. The obtained gain matrix can be only the third gain matrix, or it can be both the third and sixth gain matrices. The obtained gain matrix and the two-point correction method can be used to correct the initial response output detected by each pixel in the thermal imaging device to obtain the corrected target response output. Furthermore, the obtained target response output can be used to generate an infrared image. For example, the obtained target response output can be mapped to the range of 0-255 to obtain the pixel values ​​in the infrared image, i.e., to obtain the infrared image.

[0172] Based on the above processing, a gain matrix with high applicability can be used to correct stripe noise using the two-point correction method. This effectively corrects stripe noise and improves the quality of infrared images.

[0173] In one embodiment, step S302 includes: using the acquired third gain matrix for correcting stripe noise in the second direction and the two-point correction method to correct the initial response output detected by each pixel in the thermal imaging device, and obtaining the corrected target response output.

[0174] or,

[0175] Using the obtained third gain matrix for correcting stripe noise in the second direction and the two-point correction method, the initial response output detected by each pixel in the thermal imaging device is corrected to obtain the response output after one correction; using the obtained sixth gain matrix for correcting stripe noise in the first direction and the two-point correction method, the response output after one correction is corrected to obtain the corrected target response output.

[0176] In this embodiment, the stripe noise in the second direction can be corrected using only the third gain matrix and the two-point correction method for the initial response output detected by each pixel, thus obtaining the corrected target response output. For example, according to the two-point correction response model in the above embodiment, the initial response output detected by each pixel in the thermal imaging device can be taken as X, and the third gain matrix can be taken as K to calculate the response output after correcting the stripe noise in the second direction, which is then used as the corrected target response output.

[0177] Alternatively, the third gain matrix and the two-point correction method can be used to correct the fringe noise in the second direction for the initial response output detected by each pixel, resulting in a first-corrected response output. Then, the sixth gain matrix and the two-point correction method are used to correct the fringe noise in the first direction for the first-corrected response output, resulting in the corrected target response output. For example, according to the two-point correction response model in the above embodiment, the initial response output detected by each pixel in the thermal imaging device can be taken as X, and the third gain matrix can be taken as K. The response output after correcting the fringe noise in the second direction can be calculated as the first-corrected response output. Then, according to the two-point correction response model in the above embodiment, the response output after the first correction can be taken as X, and the sixth gain matrix can be taken as K. The response output after further correcting the fringe noise in the first direction can be calculated, resulting in the corrected target response output.

[0178] This ensures that a highly applicable gain matrix can be used to correct stripe noise in both the first and second directions using a two-point correction method. This effectively corrects stripe noise and improves the quality of infrared images.

[0179] In one embodiment, see Figure 4 , Figure 4 A flowchart illustrating another stripe noise correction method provided in this application embodiment, the method comprising:

[0180] Step S401: Calibrate the high-gain lens K1 and the ultra-high temperature lens K2. The thermal imaging device can be equipped with a high-gain lens (i.e., the first lens in the above embodiment) to calibrate the first gain matrix K1 at the first transmittance, i.e., step S101 in the above embodiment. The thermal imaging device can also be equipped with an ultra-high temperature lens (i.e., the second lens in the above embodiment) to calibrate the second gain matrix K2 at the second transmittance, i.e., step S102 in the above embodiment.

[0181] Step S402: Perform lateral or column filtering on K1 and K2 respectively to obtain K1L and K2L. That is, extract the low-frequency information of the first gain matrix K1 and the second gain matrix K2 in the second direction to obtain the first low-frequency information K1L and the second low-frequency information K2L. This is the same as step S1031 in the above embodiment.

[0182] Step S403: Process K1L and K2L to extract high-frequency information from the high-gain and ultra-high-temperature lenses, respectively. That is, process the first low-frequency information K1L and the second low-frequency information K2L to obtain the first high-frequency information K1LH' and the second high-frequency information K2LH'. This corresponds to steps S1032 and S1033 in the above embodiment.

[0183] Step S404: Obtain the ultra-high temperature compensation Knew through a masking scheme. Knew is the gain matrix in the above embodiment. That is, step S104 in the above embodiment.

[0184] Step S405: Output image. This can be done according to Y... i,j =Knew i,j ·X i,j +C i,j The response output obtained from the detection of each pixel is corrected, where Y i,j Knew represents the corrected target response output for the pixel at coordinates (i,j) in the infrared focal plane array. i,j X represents the gain coefficient in the i-th row and j-th column of the gain matrix. i,j C represents the initial response output detected by the pixel at coordinates (i,j). i,j This is the offset coefficient for the pixel with coordinates (i,j). The corrected target response output can be used to generate an infrared image as the final output image. This corresponds to steps S302 and S303 in the above embodiment.

[0185] In one embodiment, see Figure 5a and Figure 5b , Figure 5a This is a schematic diagram illustrating an infrared image generated using the initial response output detected by each pixel, as provided in an embodiment of this application. In this case, the generated infrared image exhibits noticeable vertical lines, resulting in low image quality. The stripe noise correction method provided in this embodiment can be used to correct the initial response output detected by each pixel, yielding a corrected target response output. Figure 5b This is a schematic diagram illustrating an infrared image generated using the corrected target response output, as provided in an embodiment of this application. Using the target response output corrected by the stripe noise correction method provided in this embodiment, the vertical lines in the generated infrared image are significantly improved, resulting in higher image quality.

[0186] Based on the same inventive concept, this application also provides a correction coefficient determination device, see [link to relevant documentation]. Figure 6 , Figure 6 This is a schematic diagram of a correction coefficient determination device provided in an embodiment of this application. The device includes:

[0187] The first gain matrix calculation module 601 is used to acquire the response output of each pixel in the thermal imaging device to a uniform blackbody at a first temperature and a second temperature under a first transmittance, and to calculate the gain coefficient of each pixel according to the two-point correction method based on the acquired response output to obtain the first gain matrix; wherein the first temperature is greater than the second temperature.

[0188] The second gain matrix calculation module 602 is used to obtain the response output of each pixel to a uniform blackbody at the first temperature and the second temperature respectively under the second transmittance, and calculate the gain coefficient of each pixel according to the two-point correction method based on the obtained response output to obtain the second gain matrix; wherein, the second transmittance is less than the first transmittance.

[0189] The high-frequency information extraction module 603 is used to extract high-frequency information of the first gain matrix and the second gain matrix in a first direction to obtain first high-frequency information and second high-frequency information; wherein, the first direction is one of the horizontal and column directions;

[0190] The third gain matrix determination module 604 is used to replace the second high-frequency information contained in the second gain matrix with the first high-frequency information to obtain a third gain matrix for correcting stripe noise in the second direction; wherein, the second direction is another direction other than the first direction in the horizontal and column directions.

[0191] Based on the correction coefficient determination device provided in this application embodiment, for the same temperature, the higher the transmittance of the thermal imaging device, the greater the infrared radiation intensity received by each pixel. Correspondingly, by detecting a uniform blackbody at a first temperature and a second temperature at a higher transmittance (i.e., the first transmittance), the detection of an ultra-high temperature object at a lower transmittance (i.e., the second transmittance) can be simulated. Accordingly, the response output of each pixel at the first transmittance can effectively represent the response output when detecting an ultra-high temperature object, and thus can effectively reflect the stripe noise in the gain aspect, which is greatly affected by the infrared radiation intensity. In this case, the high-frequency information (i.e., the first high-frequency information) in the first direction of the calculated gain matrix (i.e., the first gain matrix) can effectively compensate for the stripe noise in the second direction when detecting an ultra-high temperature object.

[0192] Because ultra-high temperature objects emit high levels of infrared radiation, the transmittance of thermal imaging devices is often low in practical applications to avoid infrared radiation saturation in each pixel. Correspondingly, the response output of each pixel in the thermal imaging device at a second transmittance effectively reflects the impact of lower transmittance on the response output of each pixel. In this case, the low-frequency information in the calculated gain matrix (i.e., the second gain matrix) can effectively compensate for low-frequency noise when detecting ultra-high temperature objects.

[0193] Therefore, the second high-frequency information in the second gain matrix can be replaced with the first high-frequency information to obtain the third gain matrix. Subsequently, using the third gain matrix, it is possible to effectively compensate for stripe noise in the second direction while effectively detecting ultra-high temperature objects, and also effectively compensate for the impact of low transmittance on the response output of each pixel. That is, the third gain matrix is ​​applicable to correcting stripe noise with high intensity in the gain aspect. This improves the applicability of the determined correction coefficients, effectively correcting stripe noise and improving the quality of the infrared image.

[0194] In one embodiment, the high-frequency information extraction module 603 includes:

[0195] The first extraction submodule is used to extract the low-frequency information of the first gain matrix and the second gain matrix in the second direction, respectively, to obtain the first low-frequency information and the second low-frequency information.

[0196] The second extraction submodule is used to extract the low-frequency information of the first low-frequency information and the second low-frequency information in the first direction, respectively, to obtain the third low-frequency information and the fourth low-frequency information.

[0197] The high-frequency information determination submodule is used to remove the third low-frequency information from the first low-frequency information to obtain the high-frequency information of the first gain matrix in the first direction, which is used as the first high-frequency information, and to remove the fourth low-frequency information from the second low-frequency information to obtain the high-frequency information of the second gain matrix in the first direction, which is used as the second high-frequency information.

[0198] In one embodiment, the first extraction submodule is specifically used to calculate the mean of the gain coefficients of each row in the second direction of the first gain matrix to obtain the elements in the first low-frequency information; and to calculate the mean of the gain coefficients of each row in the second direction of the second gain matrix to obtain the elements in the second low-frequency information.

[0199] The high-frequency information determination submodule is specifically used to remove the third low-frequency information from the first low-frequency information and expand the removal result in the second direction to obtain the high-frequency information of the first gain matrix in the first direction, which is used as the first high-frequency information; and to remove the fourth low-frequency information from the second low-frequency information and expand the removal result in the second direction to obtain the high-frequency information of the second gain matrix in the first direction, which is used as the second high-frequency information.

[0200] In one embodiment, the second extraction submodule is specifically used to filter the elements in the first low-frequency information according to the moving average filtering algorithm to obtain the third low-frequency information; and to filter the elements in the second low-frequency information according to the moving average filtering algorithm to obtain the fourth low-frequency information.

[0201] In one embodiment, the gain coefficient of each pixel in the third gain matrix represents: the difference between the gain coefficient of the pixel in the second gain matrix and the corresponding element in the second high-frequency information, and the sum of the difference ...

[0202] In one embodiment, the apparatus further includes:

[0203] The fourth gain matrix calculation module is used to calculate the response output of each pixel to a uniform blackbody at the first temperature and the second temperature respectively under the first transmittance after replacing the second high-frequency information contained in the second gain matrix with the first high-frequency information to obtain a third gain matrix for correcting stripe noise in the second direction, using the third gain matrix and the two-point correction method, and based on the calculated response output, calculate the gain coefficient of each pixel according to the two-point correction method to obtain the fourth gain matrix.

[0204] The fifth gain matrix calculation module is used to calculate the response output of each pixel to a uniform blackbody at the first temperature and the second temperature under the second transmittance using the third gain matrix and the two-point correction method, and to calculate the gain coefficient of each pixel according to the two-point correction method based on the calculated response output to obtain the fifth gain matrix.

[0205] The high-frequency information determination module is used to extract the high-frequency information in the second direction from the fourth gain matrix and the fifth gain matrix respectively, to obtain the third high-frequency information and the fourth high-frequency information;

[0206] The sixth gain matrix determination module is used to replace the fourth high-frequency information contained in the fifth gain matrix with the third high-frequency information to obtain a sixth gain matrix for correcting stripe noise in the first direction.

[0207] In one embodiment, the first transmittance represents the overall transmittance inside the thermal imaging device when the thermal imaging device is not equipped with a lens, or the overall transmittance inside the thermal imaging device when the first lens is equipped; the second transmittance represents the overall transmittance inside the thermal imaging device when the second lens is equipped.

[0208] This application also provides a stripe noise correction device, see [link to relevant documentation]. Figure 7 , Figure 7 This is a schematic diagram of a stripe noise correction device provided in an embodiment of this application. The device includes:

[0209] The gain matrix acquisition module 701 is used to acquire a gain matrix for correcting stripe noise; wherein the acquired gain matrix is ​​obtained based on any of the correction coefficient determination methods described above.

[0210] The response output correction module 702 is used to correct the initial response output detected by each pixel in the thermal imaging device using the acquired gain matrix and the two-point correction method, so as to obtain the corrected target response output.

[0211] The infrared image generation module 703 is used to generate an infrared image using the obtained target response output.

[0212] The stripe noise correction device provided in this application embodiment can correct stripe noise using a gain matrix with high applicability and a two-point correction method. This effectively corrects stripe noise and improves the quality of infrared images.

[0213] In one embodiment, the response output correction module 702 is specifically used to correct the initial response output detected by each pixel in the thermal imaging device using the acquired third gain matrix for correcting stripe noise in the second direction and the two-point correction method, so as to obtain the corrected target response output.

[0214] or,

[0215] Using the obtained third gain matrix for correcting stripe noise in the second direction and the two-point correction method, the initial response output detected by each pixel in the thermal imaging device is corrected to obtain the response output after one correction; using the obtained sixth gain matrix for correcting stripe noise in the first direction and the two-point correction method, the response output after one correction is corrected to obtain the corrected target response output.

[0216] This application also provides an electronic device, such as... Figure 8 As shown, it includes:

[0217] Memory 801 is used to store computer programs;

[0218] When the processor 802 executes the program stored in the memory 801, it implements either the above-described method for determining the correction coefficient or either of the above-described methods for correcting stripe noise.

[0219] Furthermore, the aforementioned electronic device may also include a communication bus and / or a communication interface, with the processor 802, the communication interface, and the memory 801 communicating with each other via the communication bus.

[0220] The communication bus mentioned in the above electronic devices can be a Peripheral Component Interconnect (PCI) bus or an Extended Industry Standard Architecture (EISA) bus, etc. This communication bus can be divided into address bus, data bus, control bus, etc. For ease of illustration, only one thick line is used to represent it in the diagram, but this does not mean that there is only one bus or one type of bus.

[0221] The communication interface is used for communication between the aforementioned electronic devices and other devices.

[0222] The memory may include random access memory (RAM) or non-volatile memory (NVM), such as at least one disk storage device. Optionally, the memory may also be at least one storage device located remotely from the aforementioned processor.

[0223] The processors mentioned above can be general-purpose processors, including central processing units (CPUs), network processors (NPs), etc.; they can also be digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components.

[0224] In another embodiment provided in this application, a computer-readable storage medium is also provided, which stores a computer program that, when executed by a processor, implements the steps of any of the above-described methods for determining correction coefficients or any of the above-described methods for correcting stripe noise.

[0225] In another embodiment provided in this application, a computer program product containing instructions is also provided, which, when run on a computer, causes the computer to execute any of the correction coefficient determination methods or any of the stripe noise correction methods in the above embodiments.

[0226] In the above embodiments, implementation can be achieved entirely or partially through software, hardware, firmware, or any combination thereof. When implemented using software, it can be implemented entirely or partially in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, all or part of the processes or functions described in the embodiments of this application are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., coaxial cable, fiber optic, digital subscriber line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that integrates one or more available media. The available medium can be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a solid-state drive (SSD), etc.

[0227] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0228] The various embodiments in this specification are described in a related manner. Similar or identical parts between embodiments can be referred to mutually. Each embodiment focuses on describing the differences from other embodiments. In particular, the embodiments of apparatus, electronic devices, computer-readable storage media, and computer program products are basically similar to the method embodiments, and therefore the descriptions are relatively simple; relevant parts can be referred to the descriptions of the method embodiments.

[0229] The above description is merely a preferred embodiment of this application and is not intended to limit the scope of protection of this application. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application are included within the scope of protection of this application.

Claims

1. A method for determining correction coefficients, characterized in that, The method includes: The response output of each pixel in the thermal imaging device to a uniform blackbody at a first temperature and a second temperature is obtained at a first transmittance. Based on the obtained response output, the gain coefficient of each pixel is calculated according to the two-point correction method to obtain the first gain matrix; wherein the first temperature is greater than the second temperature. The response output of each pixel to a uniform blackbody at the first and second temperatures is obtained at the second transmittance, and the gain coefficient of each pixel is calculated according to the two-point correction method based on the obtained response output to obtain the second gain matrix; wherein the second transmittance is less than the first transmittance. High-frequency information of the first gain matrix and the second gain matrix are extracted in the first direction to obtain first high-frequency information and second high-frequency information; wherein, the first direction is one of the horizontal and column directions; The second high-frequency information contained in the second gain matrix is ​​replaced with the first high-frequency information to obtain a third gain matrix for correcting stripe noise in the second direction; wherein the second direction is another direction other than the first direction in the horizontal and column directions.

2. The method according to claim 1, characterized in that, The step of extracting high-frequency information from the first gain matrix and the second gain matrix in the first direction to obtain first high-frequency information and second high-frequency information includes: The low-frequency information of the first gain matrix and the second gain matrix in the second direction are extracted respectively to obtain the first low-frequency information and the second low-frequency information; The first low-frequency information and the second low-frequency information are extracted respectively in the first direction to obtain the third low-frequency information and the fourth low-frequency information; Remove the third low-frequency information from the first low-frequency information to obtain the high-frequency information of the first gain matrix in the first direction, which is used as the first high-frequency information; and remove the fourth low-frequency information from the second low-frequency information to obtain the high-frequency information of the second gain matrix in the first direction, which is used as the second high-frequency information.

3. The method according to claim 2, characterized in that, The step of extracting low-frequency information from the first gain matrix and the second gain matrix in the second direction to obtain first low-frequency information and second low-frequency information includes: Calculate the mean value of the gain coefficients in each row of the second direction in the first gain matrix to obtain the elements in the first low-frequency information; calculate the mean value of the gain coefficients in each row of the second direction in the second gain matrix to obtain the elements in the second low-frequency information. The step of removing the third low-frequency information from the first low-frequency information to obtain the high-frequency information of the first gain matrix in the first direction, as the first high-frequency information, and removing the fourth low-frequency information from the second low-frequency information to obtain the high-frequency information of the second gain matrix in the first direction, as the second high-frequency information, includes: The third low-frequency information in the first low-frequency information is removed, and the removal result is expanded in the second direction to obtain the high-frequency information of the first gain matrix in the first direction, which is used as the first high-frequency information. The fourth low-frequency information in the second low-frequency information is removed, and the removal result is expanded in the second direction to obtain the high-frequency information of the second gain matrix in the first direction, which is used as the second high-frequency information.

4. The method according to claim 2 or 3, characterized in that, The step of extracting low-frequency information from the first low-frequency information and the second low-frequency information in a first direction to obtain third low-frequency information and fourth low-frequency information includes: The elements in the first low-frequency information are filtered according to the moving average filtering algorithm to obtain the third low-frequency information. The elements in the second low-frequency information are filtered using the moving average filtering algorithm to obtain the fourth low-frequency information.

5. The method of claim 1, wherein, The gain coefficient of each pixel in the third gain matrix represents the difference between the gain coefficient of the pixel in the second gain matrix and the corresponding element in the second high-frequency information, and the sum of the gain coefficient of the pixel and the corresponding element in the first high-frequency information. or, The gain coefficient of each pixel in the third gain matrix represents the product of the ratio of the gain coefficient of that pixel in the second gain matrix to the corresponding element in the second high-frequency information and the corresponding element of that pixel in the first high-frequency information.

6. The method of claim 1, wherein, After replacing the second high-frequency information contained in the second gain matrix with the first high-frequency information to obtain a third gain matrix for correcting stripe noise in the second direction, the method further includes: Using the third gain matrix and the two-point correction method, the response output of each pixel to a uniform blackbody at the first temperature and the second temperature is calculated at the first transmittance. Based on the calculated response output, the gain coefficient of each pixel is calculated according to the two-point correction method to obtain the fourth gain matrix. Using the third gain matrix and the two-point correction method, the response output of each pixel to a uniform blackbody at the first temperature and the second temperature is calculated at the second transmittance. Based on the calculated response output, the gain coefficient of each pixel is calculated according to the two-point correction method to obtain the fifth gain matrix. High-frequency information in the second direction is extracted from the fourth gain matrix and the fifth gain matrix respectively to obtain the third high-frequency information and the fourth high-frequency information; The fourth high-frequency information contained in the fifth gain matrix is ​​replaced with the third high-frequency information to obtain a sixth gain matrix for correcting stripe noise in the first direction.

7. The method of claim 1, wherein, The first transmittance characterizes the overall transmittance inside the thermal imaging device when the thermal imaging device is not equipped with a lens, or the overall transmittance inside the thermal imaging device when the first lens is equipped; the second transmittance characterizes the overall transmittance inside the thermal imaging device when the second lens is equipped.

8. A method of correcting for stripe noise, characterized by, The method includes: Obtain a gain matrix for correcting stripe noise; wherein the obtained gain matrix is ​​obtained based on the method described in any one of claims 1-7; Using the obtained gain matrix and the two-point correction method, the initial response output detected by each pixel in the thermal imaging device is corrected to obtain the corrected target response output; Infrared images are generated using the obtained target response output.

9. The method of claim 8, wherein, The process of correcting the initial response output detected by each pixel in the thermal imaging device using the acquired gain matrix and the two-point correction method to obtain the corrected target response output includes: Using the obtained third gain matrix for correcting stripe noise in the second direction and the two-point correction method, the initial response output detected by each pixel in the thermal imaging device is corrected to obtain the corrected target response output. or, Using the obtained third gain matrix for correcting stripe noise in the second direction and the two-point correction method, the initial response output detected by each pixel in the thermal imaging device is corrected to obtain the response output after one correction; using the obtained sixth gain matrix for correcting stripe noise in the first direction and the two-point correction method, the response output after one correction is corrected to obtain the corrected target response output.

10. A correction coefficient determination apparatus characterized by comprising: The device includes: The first gain matrix calculation module is used to obtain the response output of each pixel in the thermal imaging device to a uniform blackbody at a first temperature and a second temperature under a first transmittance, and to calculate the gain coefficient of each pixel according to the two-point correction method based on the obtained response output to obtain the first gain matrix; wherein, the first temperature is greater than the second temperature. The second gain matrix calculation module is used to obtain the response output of each pixel to a uniform blackbody at the first temperature and the second temperature respectively under the second transmittance, and calculate the gain coefficient of each pixel according to the two-point correction method based on the obtained response output to obtain the second gain matrix; wherein, the second transmittance is less than the first transmittance. A high-frequency information extraction module is used to extract high-frequency information from the first gain matrix and the second gain matrix in a first direction, respectively, to obtain first high-frequency information and second high-frequency information; wherein, the first direction is one of the horizontal and column directions; The third gain matrix determination module is used to replace the second high-frequency information contained in the second gain matrix with the first high-frequency information to obtain a third gain matrix for correcting stripe noise in the second direction; wherein, the second direction is another direction other than the first direction in the horizontal and column directions.

11. The apparatus of claim 10, wherein, The high-frequency information extraction module includes: The first extraction submodule is used to extract the low-frequency information of the first gain matrix and the second gain matrix in the second direction, respectively, to obtain the first low-frequency information and the second low-frequency information. The second extraction submodule is used to extract the low-frequency information of the first low-frequency information and the second low-frequency information in the first direction, respectively, to obtain the third low-frequency information and the fourth low-frequency information. The high-frequency information determination submodule is used to remove the third low-frequency information from the first low-frequency information to obtain the high-frequency information of the first gain matrix in the first direction, which is used as the first high-frequency information; and to remove the fourth low-frequency information from the second low-frequency information to obtain the high-frequency information of the second gain matrix in the first direction, which is used as the second high-frequency information. And / or, The first extraction submodule is specifically used to calculate the mean of the gain coefficients of each row in the second direction of the first gain matrix to obtain the elements in the first low-frequency information; and to calculate the mean of the gain coefficients of each row in the second direction of the second gain matrix to obtain the elements in the second low-frequency information. The high-frequency information determination submodule is specifically used to remove the third low-frequency information from the first low-frequency information and expand the removal result in the second direction to obtain the high-frequency information of the first gain matrix in the first direction as the first high-frequency information; and to remove the fourth low-frequency information from the second low-frequency information and expand the removal result in the second direction to obtain the high-frequency information of the second gain matrix in the first direction as the second high-frequency information. And / or, The second extraction submodule is specifically used to filter the elements in the first low-frequency information according to the moving average filtering algorithm to obtain the third low-frequency information; and to filter the elements in the second low-frequency information according to the moving average filtering algorithm to obtain the fourth low-frequency information. And / or, The gain coefficient of each pixel in the third gain matrix represents: the difference between the gain coefficient of the pixel in the second gain matrix and the corresponding element in the second high-frequency information, and the sum of the gain coefficient and the corresponding element in the first high-frequency information; or, the gain coefficient of each pixel in the third gain matrix represents: the ratio of the gain coefficient of the pixel in the second gain matrix to the corresponding element in the second high-frequency information, and the product of the ratio of the gain coefficient and the corresponding element in the first high-frequency information. And / or, The device further includes: The fourth gain matrix calculation module is used to calculate the response output of each pixel to a uniform blackbody at the first temperature and the second temperature respectively under the first transmittance after replacing the second high-frequency information contained in the second gain matrix with the first high-frequency information to obtain a third gain matrix for correcting stripe noise in the second direction, using the third gain matrix and the two-point correction method, and based on the calculated response output, calculate the gain coefficient of each pixel according to the two-point correction method to obtain the fourth gain matrix. The fifth gain matrix calculation module is used to calculate the response output of each pixel to a uniform blackbody at the first temperature and the second temperature under the second transmittance using the third gain matrix and the two-point correction method, and to calculate the gain coefficient of each pixel according to the two-point correction method based on the calculated response output to obtain the fifth gain matrix. The high-frequency information determination module is used to extract the high-frequency information in the second direction from the fourth gain matrix and the fifth gain matrix respectively, to obtain the third high-frequency information and the fourth high-frequency information; The sixth gain matrix determination module is used to replace the fourth high-frequency information contained in the fifth gain matrix with the third high-frequency information to obtain a sixth gain matrix for correcting stripe noise in the first direction. And / or, The first transmittance characterizes the overall transmittance inside the thermal imaging device when the thermal imaging device is not equipped with a lens, or the overall transmittance inside the thermal imaging device when the first lens is equipped; the second transmittance characterizes the overall transmittance inside the thermal imaging device when the second lens is equipped.

12. A stripe noise correction device characterized by comprising: The device includes: A gain matrix acquisition module is used to acquire a gain matrix for correcting stripe noise; wherein the acquired gain matrix is ​​obtained based on the method described in any one of claims 1-7; The response output correction module is used to correct the initial response output detected by each pixel in the thermal imaging device using the acquired gain matrix and the two-point correction method, so as to obtain the corrected target response output. The infrared image generation module is used to generate an infrared image using the obtained target response output.

13. The apparatus of claim 12, wherein, The response output correction module is specifically used to correct the initial response output detected by each pixel in the thermal imaging device by using the obtained third gain matrix for correcting stripe noise in the second direction and the two-point correction method, so as to obtain the corrected target response output. or, Using the obtained third gain matrix for correcting stripe noise in the second direction and the two-point correction method, the initial response output detected by each pixel in the thermal imaging device is corrected to obtain the response output after one correction; using the obtained sixth gain matrix for correcting stripe noise in the first direction and the two-point correction method, the response output after one correction is corrected to obtain the corrected target response output.

14. An electronic device, comprising: include: Memory, used to store computer programs; A processor, when executing a program stored in memory, implements the method of any one of claims 1-7 or any one of claims 8-9.

15. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed by a processor, implements the method described in any one of claims 1-7 or any one of claims 8-9.

16. A computer program product, characterised in that, When the computer program product is run on a computer, it causes the computer to perform the method according to any one of claims 1-7 or any one of claims 8-9.

Images