Method and device for adjusting working point of uncooled thermal infrared imager and medium

Abstract

The invention relates to the field of monitoring, in particular to an uncooled thermal infrared imager working point adjusting method and device and a medium. According to the method, variables such as environment temperature change, response mean value change, environment temperature change, image quality evaluation coefficient and the like are integrated, a data set is generated through fusion by adopting a data set automatic generation algorithm, then an optimal TEC set value is given by adopting a fusion K nearest neighbor algorithm, and an optimal working point adjustment strategy is determined. A TEC control algorithm based on a machine learning technology is provided by integrating variables such as detector output change, detector focal plane temperature change and temperature change of an environment where a detector is located from the perspective of detector use characteristics by fusing a K-nearest neighbor algorithm. The optimal TEC temperature and switching opportunity are selected according to input data, it is ensured that the image quality can keep high contrast and high definition all day long, power consumption control is more reasonable, the over-range temperature difference between the focal plane temperature and the environment temperature cannot occur, and the heat productivity is obviously reduced.

Description

Technical Field

[0001] The invention relates to the field of monitoring, and in particular to a method, a device and a medium for adjusting the working point of an uncooled infrared thermal imager. Background Art

[0002] The detectors used in uncooled infrared thermal imagers are made of materials such as vanadium oxide and polysilicon. As the temperature changes and time passes, the output response drifts, which is manifested as poor image uniformity and the inability to present high-contrast, high-definition images. In order to overcome this drift characteristic, from the perspective of temperature change, domestic head uncooled detector manufacturers use electronic cooling devices (TEC for short) to control the temperature of the focal plane, keep the focal plane temperature stable over a relatively long period of time, and thus reduce the drift rate.

[0003] The traditional TEC control method adopts the principle of zone temperature control, dividing the working temperature range of the uncooled thermal imager into several zones. The zone division adopts a fixed allocation method, or is divided equally, or is allocated according to the characteristics of experimental data. A working point (working temperature) is set for each zone. Through TEC control, the focal plane temperature is limited to the working point temperature. When the ambient temperature of the detector changes, the temperature zone is switched. The above process is called the working point adjustment of the uncooled infrared thermal imager. When the ambient temperature and the working point temperature are close, the TEC cooling power is low, the device heats up less, and the image quality is normal. When the ambient temperature and the working point temperature deviate greatly, the TEC cooling power is high, the device heats up a lot, which affects the radiation distribution of the optical system and causes the image quality to deteriorate. This partitioning method is simple to control and easy to implement in engineering. The temperature zone is switched according to the real-time temperature, but the disadvantages are obvious. The fixed temperature zone allocation and the fixed value operation of the working point have a serious impact on the image quality, and it cannot ensure that the detector always works in the best state. Summary of the invention

[0004] In view of the defects of the prior art, the present invention provides a method, device and medium for adjusting the working point of an uncooled infrared thermal imager, and designs a set of TEC setting value generation and switching system based on the fusion K-nearest neighbor algorithm, that is, the working point adjustment method of an uncooled thermal imager based on the fusion K-nearest neighbor will no longer rely on fixed temperature zones and temperature working points, and the temperature point setting (TEC working temperature setting) will be performed according to the dynamic data of the thermal imager.

[0005] In order to solve the technical problem, the technical solution adopted by the present invention is: a method for adjusting the working point of an uncooled infrared thermal imager, comprising the following steps: S01, real-time acquisition of the ambient temperature, shutter mean, response mean and current TEC setting value of the uncooled infrared thermal imager, and calculation of the image quality evaluation coefficient; S02: preprocessing, generating initial data and dynamically updating the data collected in step S01 to form an original data set ; S03. For the original data set Use function Processing is performed to obtain a valid data set D and a valid data set parameter P. The valid data set parameter P is a limiting condition for the valid data set D. The function The expression formula is: , function The original data set Analyze and calculate the data elements, calculate the maximum value of each dimension parameter and define it as the effective data set parameter P, then normalize each data element based on the effective data set parameter P to obtain the effective data set D; S04. Extract the ambient temperature, shutter mean, and response mean from the valid data set to form a valid data set subset ; S05, extract the ambient temperature, shutter mean and response mean in the pre-processed data of step S02, and pass the extracted data through the function Get input data , a subset of the valid dataset The elements of the input data ,function The expression formula is: , in Represents preprocessed data, function According to the effective data set parameters , for preprocessed data The ambient temperature, shutter mean and response mean in are normalized; S06, K value allocation, the adjustable parameters , , The K value in the K nearest neighbor algorithm is obtained through proportional operation: ; S07. Calculate input data and input data The distance between , will be less than the distance K Recorded as ,from Randomly select 5 data As a fusion reference, five elements with subscripts a, b, c, d, and e are searched from the valid data set D, and the image quality evaluation coefficient of each element is taken out. , calculate the fusion coefficient , , , , ; S08: The fusion coefficient obtained in step S07 and its corresponding valid data set Substitute the TEC setting value in the following formula to calculate the initial value of the TEC setting value : , , , , , They represent the fusion coefficients corresponding to the five elements with subscripts a, b, c, d, and e in the valid data set D, respectively. , , , , They represent the TEC setting values ​​corresponding to the five elements with subscripts a, b, c, d, and e in the valid data set D; Set the TEC setpoint to the initial value Perform fusion correction according to the following formula to obtain the final TEC setting value , , in , is the feedback coefficient.

[0006] Furthermore, the preprocessing in step S02 is to average multiple data before the clock beat according to the clock beat T, and the data involved in the preprocessing include the ambient temperature, the shutter mean, the response mean and the current TEC setting value; Initial data generation refers to performing data acquisition and preprocessing within multiple clock beats of the initial operation of the uncooled infrared thermal imager, and judging whether the image quality evaluation coefficient of each clock beat meets the set requirements. If so, the preprocessed data at that moment is valid, and the preprocessed data at that moment and its corresponding image quality evaluation coefficient are added to the original data set until the quantity requirement of the original data set is met; Dynamic update occurs during the entire life cycle of the uncooled infrared thermal imager, which includes two processes. The first is to sort the existing original data set according to the image quality evaluation coefficient, find the preprocessed data with the largest image quality evaluation coefficient, and record its image quality evaluation coefficient as A. Then, obtain the preprocessed data of the current clock beat, and record the image quality evaluation coefficient of the current clock beat as B. Compare the sizes of A and B. If B is smaller than A, replace A and its corresponding preprocessed data with B and its corresponding preprocessed data.

[0007] Furthermore, the feedback coefficient , The calculation process is: Get TEC setpoint Image quality evaluation function before setting , set TEC setpoint After the temperature stabilizes, obtain the image quality evaluation function , get the valid data set parameters The maximum value of the image quality evaluation coefficient in , set two parameters and , represent positive and negative feedback respectively, and the calculation formula is: .

[0008] Furthermore, the image quality evaluation coefficient is obtained through the image quality evaluation algorithm, and the specific process is as follows: Define the image array and calculate its mean. Arrange the response values ​​of each pixel of the infrared thermal imager in sequence to form an array with row and column distribution, which is called the image array. represents the mean value of the image array at time T, then is the mean response value of the infrared thermal imager at time T; Define the local array and calculate its mean. Use a rectangular box to select a part of the pixels in the image array. The array composed of the response values ​​of the selected pixels is called the local array. represents the local array mean at time T; use represents the image quality evaluation coefficient at time T, then: , if More than 16383, =16383, The minimum value is 0.

[0009] Furthermore, the method further includes step S09, TEC switching, obtaining the current image quality evaluation coefficient , get the current focal surface temperature , get the valid data set parameters The maximum value of the image quality evaluation coefficient in , calculate the switching decision coefficient , when the determination coefficient When the value exceeds the set threshold, TEC switching is executed, that is, the TEC set value Transmitted to the TEC control program, start executing the TEC switching task, switch the decision coefficient The calculation formula is: , in Indicates the current ambient temperature. represents the current response mean, The current shutter average value.

[0010] Furthermore, it is determined whether the image quality evaluation coefficient of each clock beat meets the set requirements: It is determined whether the image quality evaluation coefficient of each clock beat is less than 5000. If it is less than 5000, the requirement is met.

[0011] Furthermore, the fusion coefficient , , , , The calculation formula is: .

[0012] Further, adjustable parameters Characterizes the ambient temperature, Characterizes the response mean, Characterize shutter mean and set adjustable parameters according to individual characteristics of uncooled infrared thermal imagers , , ,in The default value is 1. The default value is 1. The default value is 500.

[0013] A device for adjusting the working point of an uncooled infrared thermal imager comprises a processor and a memory storing program instructions, wherein the processor is configured to execute the method for adjusting the working point of an uncooled infrared thermal imager as described above when running the program instructions.

[0014] A storage medium stores program instructions, which, when running, execute the above-mentioned method for adjusting the working point of an uncooled infrared thermal imager.

[0015] Beneficial effects of the present invention: As described in the technical background, the traditional method of generating TEC set values ​​is implemented by partitioning, and all working points and partition endpoints are pre-set according to the ambient temperature switching, which has problems such as high power consumption and poor image quality. The present invention integrates variables such as ambient temperature change, response mean change, ambient temperature change, image quality evaluation coefficient, etc., adopts the data set automatic generation algorithm (DSAC algorithm for short) to fuse and generate data sets, and then adopts the fusion K nearest neighbor algorithm (PKNN algorithm) to give the optimal TEC set value and determine the optimal working point adjustment strategy.

[0016] The fusion K nearest neighbor algorithm starts from the perspective of detector usage characteristics, integrates variables such as detector output changes, detector focal surface temperature changes, and detector ambient temperature changes, and proposes a TEC control algorithm based on machine learning technology: First, the automatic data set generation algorithm is adopted, combined with the image quality evaluation algorithm, and no fixed temperature zone division and temperature working point are set. An effective data set is dynamically established based on the dynamic data of the thermal imager to provide a basis for neighborhood fusion technology; Second, the improved K nearest neighbor algorithm-fused K nearest neighbor algorithm (KNN for short, PKNN after improvement) is adopted to realize the intelligent fusion processing of ambient temperature changes, focal surface temperature changes, and detector response output changes, and the optimal working point adjustment strategy is given, and the TEC setting value is given. Third, the TEC setting processing algorithm is used to realize temperature switching. The switching is not based on temperature as the only reference variable, but based on ambient temperature changes, focal surface temperature changes, detector response output changes, etc., and the switching decision coefficient is calculated as the switching trigger condition.

[0017] The RKNN algorithm itself belongs to the category of machine learning algorithms. During operation, the device will continuously enrich the sample library and update the data set, making the PKNN algorithm more and more "smart". It selects the optimal TEC temperature and switching time according to the input data to ensure that the picture quality can maintain high contrast and high definition all day long. The power consumption control is more reasonable, and there will be no out-of-range temperature difference between the focal surface temperature and the ambient temperature, and the heat generation is significantly reduced. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1 A flowchart of the method described in the embodiment; Figure 2 Schematic diagram of selecting a local matrix in an image matrix. DETAILED DESCRIPTION

[0019] The present invention will be further described below in conjunction with the accompanying drawings and specific embodiments.

[0020] Example 1 This embodiment discloses a method for adjusting the working point of an uncooled infrared thermal imager. Figure 1 As shown, the following steps are included: S01, collect the ambient temperature, shutter mean, response mean and current TEC setting value of the uncooled infrared thermal imager in real time, and calculate the image quality evaluation coefficient based on the image quality evaluation algorithm (KIMG); S02, Generate a valid data set based on the data set automatic generation algorithm (DASC); S03, generating TEC set values ​​based on a fusion K-nearest neighbor algorithm (PKNN); S04. Execute TEC switching and feedback acquisition based on the TEC set value processing algorithm.

[0021] There are many methods for evaluating image quality in the prior art, such as peak signal-to-noise ratio and mean square error based on image pixel statistics, as well as MAE and signal-to-noise ratio SNR, which are relatively common quality evaluation methods. They measure the quality of the image to be evaluated from a statistical perspective by calculating the difference between the grayscale values ​​of the corresponding pixels of the image to be evaluated and the reference image. This embodiment discloses a real-time output image quality evaluation coefficient after shutter occlusion, which characterizes the uniformity of the picture and provides a judgment basis for the DSAC and PKNN algorithms. The calculation process is: (1) Image array definition: Infrared thermal imagers have the ability to sense heat. The response signal of each pixel is digitized by a 14-bit D / A converter. We call the digitized pixel response signal the response value. Since the D / A converter is 14 bits, the range of the response value is [0,16383]. The response values ​​of each pixel of the infrared thermal imager are arranged in sequence to form an array with rows and columns. We call this array an image array. Represents the response output value of a pixel in the image array, m represents the number of columns of the image, and its value range is [0,639], and n represents the number of rows of the image, and its value range is [0,511]. Represents the shutter response value of the pixel in the nth row and mth column.

[0022] (2) Local array definition: A rectangular frame is used to select a part of the pixels in the image array. The array composed of the response values ​​of the selected pixels is called a local array. represents the response value of a pixel in the local array, j and k represent the array size (j / k=1.25, j≤m, k≤n), p and q represent the position of the starting pixel (yellow dot position) of the local array in the image array (p≤m, q≤n), such as Figure 2 shown.

[0023] (3) Definition and calculation of response mean: use represents the mean value of the image array at time T, then is the response mean of the infrared thermal imager at time T. Since the response value of each pixel ranges from [0,16383], the response mean range is also [0,16383], which can be expressed by the formula: (1).

[0024] (4) Definition and calculation of local array mean: use represents the local array mean at time T, calculated according to formula (2), then The value range is [0,16383]: (2); Among them, w is an open parameter of the algorithm, which determines the size of the local array. p and q are generated by a random number function with a random number generation frequency of 1000 Hz. The random function must ensure that p and q are legal, that is, the selected array range is within the image array.

[0025] (5) Obtain image quality evaluation coefficient: use represents the image quality evaluation coefficient at time T, and its value range is [0,16383], then: (3) Where t is 1000 milliseconds, matching the frame rate of 50 Hz, that is, continuous integration is performed within a time range of 1 second. The coefficient evaluates the image uniformity in the time dimension and the spatial dimension within 1 second.

[0026] (6) Restriction treatment evaluation coefficient: if More than 16383, then =16383, The minimum value of is 0. If the image quality evaluation coefficient is less than 5000, it meets the uniformity requirement and is judged as a qualified image. Inversely proportional to image quality, that is, image quality evaluation coefficient The larger the image is, the worse the image uniformity is and the worse the quality is. Images with less than 5,000 are qualified.

[0027] After collecting data, it is necessary to generate a valid data set. This process is implemented through the effective data set automatic generation algorithm (DSAC). The algorithm automatically generates a valid data set according to the clock beat T. The effective data set includes the effective data set and the effective set parameters, both of which are obtained by fusion and shaping of the original data set. The original data set is obtained by the pre-processed data elements through the following three processing processes A, B, and C. The ambient temperature, response mean, and shutter mean in the effective data set actually form a three-dimensional discrete distribution of points. The KPNN algorithm will use these discrete distribution points to generate the final ideal result. One of the innovations of the effective data set automatic generation algorithm (DSAC) is that it has the ability to continuously update itself. Compared with traditional calibration algorithms, the accuracy brought by self-update is essentially improved.

[0028] The implementation process of the effective data set automatic generation algorithm (DSAC) is as follows: (1) Preprocessing data generation: Ambient temperature , shutter average , response mean These are all internal data generated by the thermal imager in real time. is the current TEC setting value. Formula (4) is used to preprocess the data, and the average value of 50 times (or other values) is obtained to eliminate the negative impact of fluctuation errors and preprocess the data into data related to time T.

[0029] (4), in, , , , They are the TEC set value, ambient temperature value, response mean, and shutter mean corresponding to time T respectively.

[0030] The data generated in formula (4) is combined into preprocessed data elements related to the time beat T : (5).

[0031] because It is obtained by calculating the average value of the internal data generated by the thermal imager in real time according to the time beat, so It is also generated in real time and updated every time tick T. The data will be used to generate the original data set and also used as the original data input for the PKNN algorithm. The data is normalized and participates in the operation of the PKNN algorithm.

[0032] (2) Original dataset generation: The original data set is the basic data of the present invention, which is composed of 2000 dynamically updated data. When the system is run for the first time, the original data set is empty. At this time, a set consisting of 2000 initial data will be established after multiple T clock beats. During the life cycle of the subsequent algorithm operation, the original data set will be continuously updated with the time beat T to ensure the accuracy of the effective data set generated from the original data set.

[0033] The formula for generating the original data is: (6), in represents an element in the original dataset, represents the preprocessed data at time T, for The corresponding image quality evaluation coefficient. Function It includes three processes: A, B, and C. The input is preprocessed data and image quality evaluation coefficient at time T The ultimate goal of the function is to establish and dynamically update the original data set. The three processes A, B, and C are: A. Determine the validity of preprocessed data according to the time beat T The value of <5000, then determine the preprocessed data at time T Valid and Add image quality evaluation coefficient to preprocessing data elements Then get the valid data after preprocessing , written as: (7), in Represents the preprocessed data that has been judged for validity. , , , , They respectively represent the elements of the pre-processed data after validity determination, namely, the pre-processed TEC set value, ambient temperature, response mean and shutter mean after validity determination.

[0034] B. Sort the data in the original data set according to the image quality evaluation coefficient of each element, and find the data element with the largest image quality evaluation coefficient in the original data set. , temporarily stored and defined as , and its corresponding image quality evaluation coefficient is defined as .

[0035] C. Comparison Image quality evaluation coefficient in and Medium image quality evaluation coefficient If the size < , then use the effective preprocessing data obtained in the A process Replace the original data set , forming a new original data set, otherwise no processing is done.

[0036] The initial data generation and dynamic update are completed through the three processes A, B, and C. The initial data generation occurs in the initial state of the system operation. At this time, the original data set is empty. After running this process, 2000 initial data will be obtained in the original data set. The detailed steps of the process are: start the counter, count from 0, call process A cyclically, and get a After that, the counter automatically increases by 1, and then Feed When the counter is equal to 2000, the loop and counting ends. There will be 2000 initial data in.

[0037] Dynamic update is the real-time update of the data in the original data set. This process occurs during the entire life cycle of the device. The purpose is to keep the data set real-time at all times through continuous real-time updates, and to provide the most accurate data support for the PKNN algorithm to the maximum extent.

[0038] The detailed steps of the process are: 1. The A processing process in the call obtains a ; 2. Call the B process to sort the original data set and get and ; 3. Call the C processing procedure. If the replacement condition is met, and Replace it and get a new original data set.

[0039] 4. According to the clock beat T, execute the three processes 1, 2, and 3 once in each beat.

[0040] After the above processing, it can be ensured that the data in the original data set are composed of elements with high image quality. As time accumulates, the original data set will become more and more reasonable.

[0041] (3) Generation of effective data sets: The effective data set contains two components: the effective data set and the effective data set parameters. The effective data set parameters are the limiting conditions of the effective data set. The effective data set parameters and the effective data set are obtained by fusion and shaping of the original data set. The purpose is to facilitate the subsequent PKNN algorithm operation and data storage, reduce the computing resource occupancy rate, and reduce the data storage space occupancy.

[0042] The effective data set parameters can be expressed as: (8), in Set the maximum value of TEC in the valid data set. The minimum value of TEC set value in the valid data set. is the maximum value of the response mean in the valid data set, is the minimum value of the response mean in the valid data set, is the maximum shutter mean value in the valid data set, is the minimum shutter mean value in the valid data set, is the maximum value of the image quality evaluation coefficient.

[0043] The valid data set can be expressed as: (9), in , , , , Represent each element in the valid data set.

[0044] In this embodiment, the original data set Use function Processing is performed to obtain a valid data set D and a valid data set parameter P. The valid data set parameter P is a limiting condition for the valid data set D. The function The expression formula is: (10), function The original data set The data elements are analyzed and calculated, the maximum value of each dimension parameter is calculated and defined as the effective data set parameter P, and then the data elements are normalized in combination with the effective data set parameter P to obtain the effective data set D.

[0045] Extract the ambient temperature, shutter mean, and response mean from the valid data set to form a valid data set subset Different from the subset in the concept of set, this embodiment uses the valid data set The elements in each vector , , Valid dataset subset ,Right now .

[0046] In this embodiment, the purpose of the fusion K nearest neighbor algorithm is to find several points within the K range from the discrete valid data set, and generate a fusion coefficient in proportion to the image quality evaluation coefficient of each point, and generate the final TEC setting value in combination with the feedback of the TEC setting value generation algorithm. The implementation process is: (1) Preprocessing data normalization Fusion of K-nearest neighbor algorithm to obtain pre-processed data in DSAC algorithm in real time ,Will The ambient temperature, shutter mean, and response mean in the function Get valid input data that can be recognized by the fusion K nearest neighbor algorithm , while the effective dataset subset The elements of are also input into the fusion K nearest neighbor algorithm to form the input data ,function The expression formula is: (11), in Represents preprocessed data, function The role is to base the effective data set parameters on , for preprocessed data The ambient temperature, shutter mean and response mean in the data are normalized; the normalized data can be included in the valid data set Find the approximate value in and then match the valid data set Internal data element.

[0047] (2) K value allocation: In the traditional K-nearest neighbor algorithm, the K value is a specific value, which is closely related to the algorithm complexity and computational complexity. The K value of this algorithm will be calculated by proportional distribution. The K value is based on the adjustable parameter (K value representing ambient temperature), (K representing the corresponding mean), (K representing the shutter mean) is obtained through proportional operation, and the calculation formula is: (12); Should The value will be used in the (3) nearest neighbor vector search algorithm to determine the nearest neighbor vector set. , , is the system input parameter, which is set according to the individual characteristics of the thermal imager. The default value is 1. The default value is 1. The default value is 500.

[0048] (3) Search for nearest neighbor vectors: Will and Substituting into formula (13), we get and The distance array of each element in ,in : (13), in , , are the ambient temperature, shutter mean, and response mean processed by formula (11), , , For the dataset Ambient temperature, shutter mean, and response mean.

[0049] The data obtained (in ) distance to filter and get The data is recorded as , these data have the same status, and in the following algorithm, they can be randomly selected for calculation. The x and as well as The x in the table remains consistent and corresponds one to one. Subsequent algorithms will use x for data search.

[0050] (4) Sort neighbors and determine fusion ratio: Random Selection The smallest 5 data in As a fusion reference, from the effective dataset In the search, five elements with subscripts a, b, c, d, and e are found, and the image quality evaluation coefficient of each element is taken out. , using this value, calculate the fusion coefficient: (14).

[0051] (5) Fusion correction, calculate the TEC setting value S: The fusion coefficient obtained by formula (14) and its corresponding valid data set Substitute the TEC setting value in into formula (15) to calculate the initial value of the TEC setting value : (15), , , , , They represent the fusion coefficients corresponding to the five elements with subscripts a, b, c, d, and e in the valid data set D, respectively. , , , , They respectively represent the TEC setting values ​​corresponding to the five elements with subscripts a, b, c, d, and e in the valid data set D.

[0052] In order to make the TEC setting value more accurate, the initial value of the TEC setting value According to formula (16), the fusion correction is performed to obtain the final TEC setting value , (16), in , is the feedback coefficient.

[0053] In this embodiment, the TEC set value processing algorithm includes: (1) TEC switching: Get the current image quality evaluation coefficient , get the current focal surface temperature , get the valid data set parameters The maximum value of the image quality evaluation coefficient in , calculate the switching decision coefficient , when the determination coefficient When the value exceeds the set threshold 5, TEC switching is executed, that is, the TEC set value Transmitted to the TEC control program, start executing the TEC switching task, switch the decision coefficient The calculation formula is: (17), in Indicates the current ambient temperature. represents the current response mean, The current shutter average value.

[0054] (2) Feedback acquisition Get TEC setpoint Image quality evaluation function before setting , set TEC setpoint After the temperature stabilizes, obtain the image quality evaluation function , get the valid data set parameters The maximum value of the image quality evaluation coefficient in , set two parameters and , represent positive and negative feedback respectively, and the calculation formula is: (18).

[0055] Example 2 The embodiment of the present disclosure provides a device for adjusting the working point of an uncooled infrared thermal imager, including a processor and a memory. Optionally, the device may also include a communication interface and a bus. The processor, the communication interface, and the memory may communicate with each other through the bus. The communication interface may be used for information transmission. The processor may call the logic instructions in the memory to execute the method for adjusting the working point of the uncooled infrared thermal imager of the above embodiment.

[0056] In addition, the logic instructions in the above-mentioned memory can be implemented in the form of software functional units and can be stored in a computer-readable storage medium when sold or used as an independent product.

[0057] The memory, as a computer-readable storage medium, can be used to store software programs and computer executable programs, such as program instructions / modules corresponding to the method in the embodiment of the present disclosure. The processor executes the function application and data processing by running the program instructions / modules stored in the memory, that is, the working point adjustment method of the uncooled infrared thermal imager in the above embodiment is realized.

[0058] The memory may include a program storage area and a data storage area, wherein the program storage area may store an operating system and an application required for at least one function; the data storage area may store data created according to the use of the terminal device, etc. In addition, the memory may include a high-speed random access memory and may also include a non-volatile memory.

[0059] Example 3 An embodiment of the present disclosure provides a computer-readable storage medium storing computer-executable instructions, wherein the computer-executable instructions are configured to execute the above-mentioned method for adjusting the working point of an uncooled infrared thermal imager.

[0060] The computer-readable storage medium mentioned above may be a transient computer-readable storage medium or a non-transitory computer-readable storage medium.

[0061] The technical solution of the embodiment of the present disclosure can be embodied in the form of a software product, which is stored in a storage medium and includes one or more instructions for enabling a computer device (which may be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the method described in the embodiment of the present disclosure. The aforementioned storage medium may be a non-transient storage medium, including: a USB flash drive, a mobile hard disk, a read-only memory (ROM, Read-Only Memory), a random access memory (RAM, Random Access Memory), a disk or an optical disk, and other media that can store program codes, or a transient storage medium.

[0062] The above description and the accompanying drawings fully illustrate the embodiments of the present disclosure so that those skilled in the art can practice them. Other embodiments may include structural, logical, electrical, process and other changes. The embodiments represent only possible changes. Unless explicitly required, separate components and functions are optional, and the order of operation may vary. The parts and features of some embodiments may be included in or replace the parts and features of other embodiments. Moreover, the words used in this application are only used to describe the embodiments and are not used to limit the scope of protection. As used in the description in the text, unless the context clearly indicates, the singular forms of "a", "an" and "the" are intended to include plural forms as well. Similarly, the term "and / or" as used in this application refers to any and all possible combinations of one or more associated listings. In addition, when used in this application, the term "comprise" and its variants "comprises" and / or comprising refer to the presence of stated features, wholes, steps, operations, elements, and / or components, but do not exclude the presence or addition of one or more other features, wholes, steps, operations, elements, components and / or groups thereof. In the absence of further restrictions, the elements defined by the sentence "including one..." do not exclude the presence of other identical elements in the process, method or device including the elements. In this article, each embodiment may focus on the differences from other embodiments, and the same and similar parts between the embodiments may refer to each other. For the methods, products, etc. disclosed in the embodiments, if they correspond to the method part disclosed in the embodiments, then the relevant parts can refer to the description of the method part.

[0063] Those skilled in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software may depend on the specific application and design constraints of the technical solution. The technicians may use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of the embodiments of the present disclosure. The technicians may clearly understand that for the convenience and simplicity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments, and will not be repeated here.

Claims

1. A method for adjusting the working point of an uncooled infrared thermal imager, characterized in that: The following steps are involved: S01, real-time acquisition of the ambient temperature, shutter mean, response mean and current TEC setting value of the uncooled infrared thermal imager, and calculation of the image quality evaluation coefficient; S02: preprocessing, generating initial data and dynamically updating the data collected in step S01 to form an original data set ; S03. For the original data set Use function Processing is performed to obtain a valid data set D and a valid data set parameter P. The valid data set parameter P is a limiting condition for the valid data set D. The function The expression formula is: ， function The original data set Analyze and calculate the data elements, calculate the maximum value of each dimension parameter and define it as the effective data set parameter P, then normalize each data element based on the effective data set parameter P to obtain the effective data set D; S04. Extract the ambient temperature, shutter mean, and response mean from the valid data set to form a valid data set subset ; S05, extract the ambient temperature, shutter mean and response mean in the pre-processed data of step S02, and pass the extracted data through the function Get input data , a subset of the valid dataset The elements of the input data ,function The expression formula is: ， in Represents preprocessed data, function According to the effective data set parameters , for preprocessed data The ambient temperature, shutter mean and response mean in are normalized; S06, K value allocation, the adjustable parameters , , The K value in the K nearest neighbor algorithm is obtained through proportional operation: ； S07. Calculate input data and input data The distance between , will be less than the distance K Recorded as ,from Randomly select 5 data from As a fusion reference, five elements with subscripts a, b, c, d, and e are searched from the valid data set D, and the image quality evaluation coefficient of each element is taken out. , calculate the fusion coefficient , , , , ; S08: The fusion coefficient obtained in step S07 and its corresponding valid data set Substitute the TEC setting value in the following formula to calculate the initial value of the TEC setting value : ， , , , , They represent the fusion coefficients corresponding to the five elements with subscripts a, b, c, d, and e in the valid data set D, respectively. , , , , They represent the TEC setting values ​​corresponding to the five elements with subscripts a, b, c, d, and e in the valid data set D; Set the TEC setpoint to the initial value Perform fusion correction according to the following formula to obtain the final TEC setting value , , in , is the feedback coefficient.

2. The method for adjusting the working point of an uncooled infrared thermal imager according to claim 1, characterized in that: The preprocessing in step S02 is to average multiple data before the clock beat according to the clock beat T, and the data involved in the preprocessing include ambient temperature, shutter mean, response mean and current TEC setting value; Initial data generation refers to performing data acquisition and preprocessing within multiple clock beats of the initial operation of the uncooled infrared thermal imager, and judging whether the image quality evaluation coefficient of each clock beat meets the set requirements. If so, the preprocessed data at that moment is valid, and the preprocessed data at that moment and its corresponding image quality evaluation coefficient are added to the original data set until the quantity requirement of the original data set is met; Dynamic update occurs during the entire life cycle of the uncooled infrared thermal imager, which includes two processes. The first is to sort the existing original data set according to the image quality evaluation coefficient, find the preprocessed data with the largest image quality evaluation coefficient, and record its image quality evaluation coefficient as A. Then, obtain the preprocessed data of the current clock beat, and record the image quality evaluation coefficient of the current clock beat as B. Compare the sizes of A and B. If B is smaller than A, replace A and its corresponding preprocessed data with B and its corresponding preprocessed data.

3. The method for adjusting the working point of an uncooled infrared thermal imager according to claim 1, characterized in that: Feedback Factor , The calculation process is: Get TEC setpoint Image quality evaluation function before setting , set TEC setpoint After the temperature stabilizes, obtain the image quality evaluation function , get the valid data set parameters The maximum value of the image quality evaluation coefficient in , set two parameters and , represent positive and negative feedback respectively, and the calculation formula is: 。 4. The method for adjusting the working point of an uncooled infrared thermal imager according to claim 1, characterized in that: The image quality evaluation coefficient is obtained through the image quality evaluation algorithm. The specific process is as follows: Define the image array and calculate its mean. Arrange the response values ​​of each pixel of the infrared thermal imager in sequence to form an array with row and column distribution, which is called the image array. represents the mean value of the image array at time T, then is the mean response value of the infrared thermal imager at time T; Define the local array and calculate its mean. Use a rectangular box to select a part of the pixels in the image array. The array composed of the response values ​​of the selected pixels is called the local array. represents the local array mean at time T; use represents the image quality evaluation coefficient at time T, then: ， if More than 16383, =16383, The minimum value is 0.

5. The method for adjusting the working point of an uncooled infrared thermal imager according to any one of claims 1 to 4, characterized in that: The process also includes step S09, TEC switching, and obtaining the current image quality evaluation coefficient. , get the current focal surface temperature , get the valid data set parameters The maximum value of the image quality evaluation coefficient in , calculate the switching decision coefficient , when the determination coefficient When the value exceeds the set threshold, TEC switching is executed, that is, the TEC set value Transmitted to the TEC control program, start executing the TEC switching task, switch the decision coefficient The calculation formula is: ， in Indicates the current ambient temperature. represents the current response mean, The current shutter average value.

6. The method for adjusting the working point of an uncooled infrared thermal imager according to claim 2, characterized in that: The image quality evaluation coefficient of each clock beat is judged to see whether it meets the set requirements: It is determined whether the image quality evaluation coefficient of each clock beat is less than 5000. If it is less than 5000, the requirement is met.

7. The method for adjusting the working point of an uncooled infrared thermal imager according to claim 1, characterized in that: Fusion coefficient , , , , The calculation formula is: 。 8. The method for adjusting the working point of an uncooled infrared thermal imager according to claim 1, characterized in that: Tunable parameters Characterizes the ambient temperature, Characterizes the response mean, Characterize shutter mean and set adjustable parameters according to individual characteristics of uncooled infrared thermal imagers , , ,in The default value is 1. The default value is 1. The default value is 500.

9. An uncooled infrared thermal imager working point adjustment device, comprising a processor and a memory storing program instructions, characterized in that: The processor is configured to execute the method for adjusting the working point of an uncooled infrared thermal imager according to any one of claims 1 to 8 when running the program instructions.

10. A storage medium storing program instructions, characterized in that: When the program instructions are executed, the method for adjusting the working point of an uncooled infrared thermal imager according to any one of claims 1 to 8 is executed.

Images