CN111369552B - Infrared blind element detection method, device and computer-readable storage medium - Google Patents

Abstract

The application discloses an infrared blind element detection method, device and computer-readable storage medium. Wherein, the method includes, during the production process of the infrared imaging device, using a pre-generated blind cell table to perform infrared blind cell detection; during the use of the infrared imaging device, performing sliding window processing on the infrared image to be output according to a preset neighborhood value, Obtain the maximum and minimum values of each window; if the infrared image to be output satisfies the preset determination condition, then determine the random blind element and flash blind element in the scene based on the relationship between the variance of the infrared image to be output and the third preset threshold . The blind cell table is obtained by merging the model blind cells and/or the flash blind cells and/or the response blind cells to determine the fixed blind cells in the calibration process; the preset judgment condition is based on the maximum value, the minimum value and the fourth preset threshold Relationships between are generated. The present application can quickly and accurately detect any type of blind element without increasing the length of the process, and reduce the occurrence of blind element misjudgment.

Description

Infrared blind element detection method, device and computer-readable storage medium

technical field

The present application relates to the technical field of infrared imaging equipment preparation, in particular to an infrared blind element detection method, device and computer-readable storage medium.

Background technique

At present, cooled infrared detectors work in a refrigerated environment, and the detectors often experience temperature shocks between high and low temperatures during each power-on and power-off process. Due to the limitations of the existing manufacturing process and raw materials, infrared imaging often has dead pixels with a response rate less than 1/2 of the average response rate and overheating with a noise voltage greater than 2 times the average noise voltage described in the national military standard GB/T17444-2013 pixel. Among them, the invalid pixels in the national military standard are also called blind pixels, including dead pixels and overheated pixels.

In practical applications, the judgment of all blind elements cannot be realized by using the general judgment conditions in the national military standard, and there will be cases of missed judgments of blind elements. However, the existence of blind elements seriously restricts the infrared imaging effect, and has a serious impact on the application and promotion of infrared detectors. In actual engineering, blind elements appear as isolated or continuous bright spots and dark spots in infrared images. Some bright spots and dark spots do not change with the temperature of the scene, and only appear as points with a large difference in grayscale from the surrounding neighborhood in space. Form fixed blind cells; some blind cells appear as flickering pixels that change with time, forming flash blind cells; some blind cells appear as bright or dark spots randomly appearing with temperature or time changes, forming random blind Yuan. Through a certain method to accurately detect the blind pixels, and then use the appropriate blind pixel compensation algorithm to replace the detected blind pixels, so as to improve the quality of infrared imaging, which has important application value for the application and promotion of infrared detectors.

Related technologies usually detect blind pixels through blackbody calibration-based detection methods and scene-based detection methods. The blackbody calibration method uses the uniform blackbody image obtained during the calibration process, and then determines the blind pixel according to the response rate and noise of the pixel in the national standard. This method is widely used due to its simple principle, but it cannot deal with random blind elements. The scene-based detection method does not rely on the blackbody reference source, and can handle random blind elements, but it generally has the disadvantages of easy misjudgment, large amount of calculation, and great influence by image non-uniformity.

In view of this, how to perform high-accuracy blind cell detection on any blind cell without increasing the process time, and reduce the occurrence of blind cell misjudgment is a technical problem to be solved by those skilled in the art.

Contents of the invention

The application provides an infrared blind element detection method, device and computer-readable storage medium, which can quickly and accurately detect any type of blind element without increasing the length of the process, and reduce the risk of blind element misjudgment occur.

In order to solve the above technical problems, embodiments of the present invention provide the following technical solutions:

On the one hand, an embodiment of the present invention provides a method for detecting an infrared blind element, including:

In the production process of infrared imaging equipment, infrared blind element detection is performed using a pre-generated blind element table; the blind element table is obtained by combining model blind elements and/or flash blind elements and/or response blind elements for determining calibration Fixed blind elements in the process;

During the user's use of the infrared imaging device, sliding window processing is performed on the infrared image to be output according to the preset neighborhood value to obtain the maximum and minimum values of each window;

If the infrared image to be output satisfies a preset determination condition, determine a random flash blind element in the scene based on the relationship between the variance of the infrared image to be output and a third preset threshold;

Wherein, the model blind element is judged according to the gain value range of the non-uniformity correction model, and the flash blind element is the time-domain extreme value absolute value and the first A relationship determination between a preset threshold value, the response blind element is the relationship determination between the response matrix of the high temperature point and the low temperature point after non-uniformity correction and the second preset threshold value; the preset determination condition is based on the Determine the relationship between the maximum value, the minimum value and the fourth preset threshold.

Optionally, the determination of the blind element of the model includes:

Use the blackbody calibration method to collect multiple frames of continuous high-temperature images and multiple frames of continuous low-temperature images;

Calculating a cumulative probability density distribution function using the gain value calculated from the non-uniformity-corrected image to obtain a gain value whose interval proportion of the gain value distribution is within a preset value range;

A pixel whose current gain value is not within the preset value range is determined as the model pixel.

Optionally, the determination of the relationship between the absolute value of the time-domain extremum and the first preset threshold of the high-temperature point image and the low-temperature point image corrected by non-uniformity in the flash blind element includes:

Calculate the mean value matrix of the corresponding pixels and the time-domain extremum absolute value matrix of the corresponding pixels in the multi-frame high-temperature point image after non-uniformity correction, and obtain the high-temperature mean value matrix and the high-temperature time-domain extremum absolute value matrix;

Traversing each element in the high-temperature time-domain extreme value absolute value matrix, if the current element value in the high-temperature time-domain extreme value absolute value matrix is greater than the first preset threshold or any element in the mean value matrix value, the pixel corresponding to the current element value is classified into the first type of time-domain blind element set;

Calculate the mean value matrix of the corresponding pixel points and the time-domain extremum absolute value matrix of the corresponding pixel points in the non-uniformity-corrected multi-frame low-temperature point image, and obtain the low-temperature mean value matrix and the low-temperature time-domain extremum absolute value matrix;

Traversing through each element in the low temperature time domain extreme value absolute value matrix, if the current element value in the low temperature time domain extreme value absolute value matrix is greater than the fifth preset threshold or any element in the mean value matrix value, the pixel corresponding to the current element value is classified into the second type of temporal blind element set;

Combining the first type of time domain blind cell set and the second type of time domain blind cell set to obtain flash blind cells used for determining the calibration process.

Optionally, calculating the average value matrix of corresponding pixels and the time-domain extremum absolute value matrix of corresponding pixels in the non-uniformity-corrected multi-frame high-temperature point images to obtain the high-temperature average value matrix and the high-temperature time-domain extremum absolute value matrix include:

Utilize the mean value matrix relational expression to calculate the mean value matrix of the corresponding pixels of the multi-frame high-temperature point image after non-uniformity correction, obtain the high temperature mean value matrix; the described mean value matrix relational expression is:

The time-domain extremum absolute value matrix of the pixels corresponding to the multi-frame high-temperature point image after non-uniformity correction is calculated by using the time-domain extremum calculation relation to obtain the high-temperature time-domain extremum absolute value matrix, and the time-domain extremum calculation relationship The formula is:

In the formula, meanFigT1(i, j) is the high temperature mean value matrix, minmaxFigT1(i, j) is the absolute value matrix of the high temperature time domain extremum, Region={1,2,3,...,num}, num is The total number of frames of the high-temperature point image, figT1 _frame (i, j) is the value of the i-th row and j-th column of the frame-th frame of the high-temperature point image after correction when the temperature is T1.

Optionally, the determination of the relationship between the response matrix of the high temperature point and the low temperature point after non-uniformity correction and the second preset threshold value includes:

Calculate the gray value of the multi-frame high-temperature point image and low-temperature point image after non-uniformity correction, and obtain the high-temperature average value and low-temperature average value;

Subtracting the high temperature mean value from the low temperature mean value to obtain the response matrix;

Calculate the average value of the response matrix to all pixels of the entire array to obtain the average value of the array;

If the difference between the current element value in the response matrix and the mean value of the area array is greater than a preset sixth preset threshold, the pixel corresponding to the current element value is determined to be the response blind pixel.

Optionally, calculating the mean value of the response matrix for all pixels of the entire area array to obtain the mean value of the area array is:

The mean value of the area array is calculated using the mean calculation relation of the area array, and the mean value calculation relation of the area array is:

response(i,j)=meanFigT1(i,j)-meanFigT2(i,j);

In the formula, response(i,j) is the response matrix, meanResponse is the mean value of the array, m*n is the length and width of the array, meanFigT1(i,j) is the high temperature after non-uniformity correction The point image corresponds to the mean value matrix of the pixels when the temperature is T1, and meanFigT2(i,j) is the mean value matrix of the corresponding pixels of the high-temperature point image after non-uniformity correction when the temperature is T2.

Optionally, the preset neighborhood value is M*N, and if the infrared image to be output satisfies a preset determination condition, based on the relationship between the variance of the infrared image to be output and a third preset threshold The random flash blind elements in the judgment scene include:

If the maximum value and the minimum value of the M*N centered on (i, j) of the infrared image to be output are respectively tempMax and minimum value tempMin, the preset determination condition is:

Calculate the variance variance (i, j) after the central point of the infrared image to be output is removed, and the random flash blind element is judged according to the determination relational expression, and the determination relational expression is:

Wherein, threshold3 is the third preset threshold, and threshold4 is the fourth preset threshold; if map4(i, j) is 1, the horizontal and vertical coordinates of the infrared image to be output are (i, j) corresponding The pixel of is the random blinking element, if map4(i, j) is 0, then the pixel corresponding to (i, j) is not the random blinking element.

Another aspect of the embodiment of the present invention provides an infrared blind element detection device, including:

Calibrate the blind element detection module, used in the production process of infrared imaging equipment, use the pre-generated blind element table to perform infrared blind element detection; the blind element table is composed of model blind elements and/or flash blind elements and/or response Blind elements are merged to obtain fixed blind elements used to determine the calibration process; the model blind elements are determined according to the gain value range of the non-uniformity correction model, and the flash blind elements are high-temperature point images corrected by non-uniformity and determine the relationship between the absolute value of the time-domain extremum of the low-temperature point image and the first preset threshold, and the response blind element is the difference between the response matrix of the high-temperature point and the low-temperature point after non-uniformity correction and the second preset threshold determine the relationship between

A sliding window processing module, configured to perform sliding window processing on the infrared image to be output according to a preset neighborhood value during the user's use of the infrared imaging device, to obtain the maximum and minimum values of each window;

The random flash blind element detection module is used to determine the random flash blind element in the scene based on the relationship between the variance of the infrared image to be output and the third preset threshold if the infrared image to be output meets the preset determination condition; The preset determination condition is determined based on the relationship between the maximum value, the minimum value and a fourth preset threshold.

An embodiment of the present invention also provides an infrared blind element detection device, which includes a processor configured to implement the steps of the infrared blind element detection method described in any one of the preceding items when executing a computer program stored in a memory.

Finally, the embodiment of the present invention also provides a computer-readable storage medium, the computer-readable storage medium is stored with an infrared blind element detection program, and when the infrared blind element detection program is executed by a processor, the above-mentioned one can be realized. Describe the steps of the infrared blind element detection method.

The advantage of the technical solution provided by this application is that since the blind element table is calibrated during the production process, the computer program on which the implementation process depends is not written into the core, does not increase the power consumption of the system, and does not require additional data collection. The speed is faster and does not increase the production process. In the process of equipment production, the blind element table can be used to accurately locate the fixed blind element, and the corrected high temperature target and low temperature target are used for judgment, considering the influence of non-uniformity correction on the high and low temperature target, which can effectively avoid blind element misjudgment. Improve the accuracy of blind element detection; in actual use, by setting the threshold of scene-based blind element detection, dense flash points on a uniform surface can be detected without affecting other details of the scene. The joint use of scene-based blind element detection and calibration-based blind element detection makes the output infrared image cleaner and improves the quality of infrared image.

In addition, the embodiment of the present invention also provides a corresponding implementation device and computer-readable storage medium for the infrared blind element detection method, which further makes the method more practical, and the device and computer-readable storage medium have corresponding advantages.

It is to be understood that both the foregoing general description and the following detailed description are exemplary only and are not restrictive of the present disclosure.

Description of drawings

In order to more clearly illustrate the technical solutions of the embodiments of the present invention or related technologies, the following will briefly introduce the accompanying drawings that need to be used in the description of the embodiments or related technologies. Obviously, the accompanying drawings in the following description are only the present invention For some embodiments of the present invention, those of ordinary skill in the art can also obtain other drawings based on these drawings on the premise of not paying creative efforts.

FIG. 1 is a schematic flowchart of a method for detecting infrared blind elements provided by an embodiment of the present invention;

FIG. 2 is a schematic diagram of a K cumulative probability density distribution provided by an embodiment of the present invention;

FIG. 3 is a partially enlarged schematic diagram of FIG. 2 provided by an embodiment of the present invention;

Fig. 4 shows the signal values of the 200th row of pixels in a schematic image provided by the embodiment of the present invention at the same ambient temperature and target temperature of 14°C and 4°C respectively;

Fig. 5 is the difference between the signal values of the 200th line pixel of a schematic image provided by the embodiment of the present invention at the same ambient temperature and target temperature of 14°C and 4°C respectively;

Fig. 6 is a schematic image provided by the embodiment of the present invention, the signal values of the 200th row of pixels after non-uniformity correction at the same ambient temperature and the target temperature are 14°C and 4°C respectively;

Fig. 7 is a schematic image provided by an embodiment of the present invention, the difference between the signal values of the 200th row of pixels after non-uniformity correction at the same ambient temperature and the target temperature are 14°C and 4°C respectively;

Fig. 8 is a structural diagram of a specific implementation mode of an infrared blind element detection device provided by an embodiment of the present invention;

FIG. 9 is a structural diagram of another specific implementation manner of an infrared blind element detection device provided by an embodiment of the present invention.

Detailed ways

In order to enable those skilled in the art to better understand the solution of the present invention, the present invention will be further described in detail below in conjunction with the accompanying drawings and specific embodiments. Apparently, the described embodiments are only some of the embodiments of the present invention, but not all of them. Based on the embodiments of the present invention, all other embodiments obtained by persons of ordinary skill in the art without making creative efforts belong to the protection scope of the present invention.

The terms "first", "second", "third" and "fourth" in the specification and claims of this application and the above drawings are used to distinguish different objects, rather than to describe a specific order . Furthermore, the terms "comprising" and "having", and any variations thereof, are intended to cover a non-exclusive inclusion. For example, a process, method, system, product, or device comprising a series of steps or units is not limited to the listed steps or units, but may include unlisted steps or units.

After introducing the technical solutions of the embodiments of the present invention, various non-limiting implementations of the present application will be described in detail below.

First, referring to FIG. 1, FIG. 1 is a schematic flowchart of a method for detecting infrared blind elements provided by an embodiment of the present invention. The embodiment of the present invention may include the following:

S101: During the production process of the infrared imaging device, use the pre-generated blind element table to perform infrared blind element detection.

In the embodiment of the present invention, the production process of the infrared imaging device is also the calibration process before the infrared imaging device leaves the factory. The blind element table is calculated by the host computer program before the equipment leaves the factory. It can be obtained by combining model blind elements and/or flash blind elements and/or response blind elements to determine the fixed blind elements in the calibration process. Persons skilled in the art can combine model blind elements, flash blind elements, and response blind elements according to actual application scenarios and the unique properties of infrared detectors, which is not limited in this application.

It is understandable that pixel flicker is often caused by unstable detector responsivity. When facing a uniform surface with a certain temperature, the grayscale difference of the same pixel in consecutive multiple frames of images is relatively large. The instability of the detector pixel responsivity can be represented by the time domain variance or time domain extremum of the pixels corresponding to the continuous multi-frame images at a certain temperature. Time-domain variance can detect flash blind elements that change drastically over time, and time-domain extremum can detect flash blind elements that have certain periodic mutations, and can also detect flash blind elements that change rapidly over time. By comparing and calculating the time-domain variance or time-domain extremum of the pixels corresponding to the continuous multi-frame images under different ambient temperatures and different target temperatures and the set threshold, the flash blind pixels at each ambient temperature and each target temperature can be obtained . If the signal values of adjacent pixel points differ greatly under the same ambient temperature and the same target temperature, a fixed blind pixel is formed. Fixed blind cells are often caused by large differences in detector cell responsivity. The non-uniformity corrected gain and bias can somewhat mitigate the differences in responsivity between detector elements. Directly judging the difference between the original image or the original image at different target temperatures will easily lead to misjudgment of blind cells, and the number of blind cells is too large. Too many blind cells will directly lead to the deterioration of the blind cell replacement effect. For the image after two-point correction, multi-point correction or segment correction, the judgment of fixed blind pixels can reduce the number of fixed blind pixel judgments, and reduce the misjudgment of blind pixels to a certain extent. Based on this, the model blind element of this application can be determined based on the gain value range of the non-uniformity correction model, and the flash blind element can be the absolute value of the time-domain extremum of the high-temperature point image and the low-temperature point image after using the non-uniformity correction For the determination of the relationship between the first preset threshold and the response blind element, the response matrix of the high temperature point and the low temperature point after non-uniformity correction and the second preset threshold can be determined.

It should be noted that since the calibration is carried out during the production process, the computer program corresponding to this part is not written into the movement, which does not increase the power consumption of the system. Since this process is carried out during the routine calibration process, the calculation is performed based on the high temperature and low temperature pictures during the calibration process, no additional data collection is required, the calculation speed is fast, and the production process is not increased.

S102: During the user's use of the infrared imaging device, perform sliding window processing on the infrared image to be output according to a preset neighborhood value to obtain a maximum value and a minimum value of each window.

It can be understood that the quality of the output infrared imaging can be guaranteed after the infrared blind element is detected and replaced during the calibration process of the infrared imaging device. However, during the process of using the infrared imaging device by the user, that is, in the actual application process after the infrared imaging device leaves the factory, the infrared detector will generate random blind pixels with time or temperature changes. In order to deal with the random blind cells generated with time or temperature changes, To improve the quality of the output image, scene-based blind element detection is essential. This part of blind elements changes with the environment, real-time detection of flashing blind elements in a relatively uniform scene, and through the blind element replacement method to make the relatively uniform The scene is cleaner.

In the embodiment of the present invention, the preset neighborhood value can be determined according to the actual scene, for example, it can be 7*7 or 5*5, and the calculation method of the window size in the related sliding window processing technology can be used to calculate each window of this application values, and select the maximum and minimum values from these values.

S103: If the infrared image to be output satisfies the preset determination condition, determine the random flash blind element in the scene based on the relationship between the variance of the infrared image to be output and the third preset threshold.

Wherein, the preset judgment condition can be determined based on the relationship between the maximum value, the minimum value and the fourth preset threshold value, and the fourth preset threshold value and the third preset threshold value can be determined according to actual needs, and this application does not make any limited. The third preset threshold is a threshold for scene-based blind element detection set based on the density of uniform surface flash points in the scene. In the embodiment of the present invention, any image variance calculation method may be used to calculate the variance value of the infrared image to be output, which is not limited in this application. After S103 detects random flash blind elements in real time, any blind element replacement method such as the neighborhood mean value replacement method can be used to replace random flash blind elements, which can compensate for denser flash points on a uniform surface in the scene, and does not affect the scene other details are affected.

Finally, it should be noted that the computer program on which the implementation of S102 and S103 of the present application depends needs to be written into the core of the infrared imaging device, and these two steps can be realized by embedded software.

In the technical solution provided by the embodiment of the present invention, since the blind element table is calibrated during the production process, the computer program on which the implementation process depends is not written into the core, does not increase the power consumption of the system, and does not require additional data collection. The calculation speed is fast and does not increase the production process procedures. In the process of equipment production, the blind element table can be used to accurately locate the fixed blind element, and the corrected high temperature target and low temperature target are used for judgment, considering the influence of non-uniformity correction on the high and low temperature target, which can effectively avoid blind element misjudgment. Improve the accuracy of blind element detection; in actual use, by setting the threshold of scene-based blind element detection, dense flash points on a uniform surface can be detected without affecting other details of the scene. The combination of scene-based blind element detection and calibration-based blind element detection can make the output infrared image cleaner and improve image quality.

In the above-mentioned embodiment, there is no limitation on how to execute S103. In this embodiment, a method for detecting random flash blind elements is given. If the preset neighborhood value of S102 is M*N, S103 may include the following steps:

If the maximum value and minimum value of M*N centered on (i, j) of the infrared image outFig to be output are tempMax and minimum value tempMin respectively, when the maximum value and the minimum value are not equal and the absolute value of the difference between the maximum value and the minimum value When the value is less than the fourth encounter and the central point is the maximum value or the minimum value, the subsequent variance calculation steps can be performed, that is to say, the preset judgment condition of the present application can be expressed as:

When the above conditions are met, in order to improve the detection accuracy of the blind element, before calculating the variance, the center point can be removed first, that is, the variance variance(i,j) of the infrared image to be output after removing the center point can be calculated, and the random flash blind element is determined according to The relational expression is judged, and the judgment relational expression is:

Among them, threshold3 is the third preset threshold, and threshold4 is the fourth preset threshold; if map4(i, j) is 1, the pixels corresponding to the horizontal and vertical coordinates of the infrared image to be output (i, j) are randomly flashing Blind element, if map4(i, j) is 0, then the pixel corresponding to (i, j) is not a random flash blind element, that is, it has rich texture.

As an optional implementation, the present application also provides a method for determining model blind elements, which may include the following content:

The black body calibration method is used to collect multiple frames of continuous high-temperature images and multiple frames of continuous low-temperature images; the number of continuous high-temperature images and continuous low-temperature images may be the same or different, and this application does not make any restrictions on this. Use a non-uniformity correction method such as a two-point correction method or a multi-point correction method to process the image, and then calculate the corrected gain value, that is, the K value, and use the K value to calculate the cumulative probability density distribution function or probability density function, K cumulative probability The schematic diagram of the density distribution can be shown in Figure 2, so as to obtain the gain value whose interval proportion of the gain value distribution is within the preset value range; the preset value range can be, for example, 1% to 99.9%, or Values are selected according to actual requirements, which will not affect the implementation of the present application. Since the infrared imaging equipment is developed based on the FPGA (Field Programmable Gate Array, field programmable logic gate array) platform, the integer calculation is performed on the FPGA. In order to improve the accuracy of the K value, it can be adjusted on the basis of the preset value range to enlarge the preset value. The value range is set as the final preset value range, for example, the end point of the range is multiplied by the same coefficient value. Determine the pixel whose current gain value is not within the preset value range as the model pixel.

Specifically, the K value calculated after non-uniformity correction is converted into a 1-dimensional array. Since the distribution function of K is a continuous random variable and K∈(-∞,+∞), the probability density function of K can be f(K), and the cumulative probability density distribution function can be F(K). When F(K)<f1 or F(K)>f2, the corresponding K value is replaced by the nearest threshold K1 or K2, as shown in the following relationship:

Therefore, when the gain of the pixel satisfies the following conditions, the model blind element is generated. The pixel point of map1(i, j) is the blind element point, where i, j are the horizontal and vertical coordinates corresponding to the pixel in the image. This condition can be expressed as :

If f1=0.0021, f2=0.9991, corresponding K1=0.8931, K2=1.22, as shown in Figure 3, K1 and K2 can be translated left and right according to the actual situation, and a similar effect can be obtained.

In addition, the model blind element can also use the probability density function or histogram, cumulative distribution histogram to obtain the gain value whose interval proportion of the gain value distribution is within the preset value range, and judge the threshold according to the distribution of K.

In the embodiment of the present invention, the model blind elements introduced by the limited coefficients of the blackbody-based two-point calibration model, multi-point calibration model and segment calibration model are fully considered. If the gain calculated by the existing method is directly used without considering the blind elements of the model, the value range of the gain will be expanded. However, due to the bit width limitation of data storage in the FPGA, the accuracy of the calculated gains will be greatly reduced, and even lead to the failure of non-uniformity correction. The model blind element is used to determine the distribution range of the gain according to the distribution and proportion of the gain. The gain can be limited to a small value range. Through the shift, the accuracy of the gain can be guaranteed to the greatest extent.

Optionally, the present application also provides a calculation process under an implementation mode of flash blind element, which may include the following content:

Calculate the mean value matrix of corresponding pixels and the time-domain extremum absolute value matrix of corresponding pixels in the multi-frame high-temperature point image after non-uniformity correction, and obtain the high-temperature mean value matrix and high-temperature time-domain extremum absolute value matrix. In this step, the average value matrix and the absolute value matrix of time-domain extremum values of each corrected high-temperature point image and low-temperature point image can be calculated. The total number of high-temperature point images and low-temperature point images may be the same or different, which does not affect the implementation of the present application. That is to say, the calibration form of N segments of K can be used, and the calibration form of one segment of K can also be used, then N=1. For multiple groups of K, it is only necessary to combine the blind element tables obtained from the images of the target temperature points used in the calibration of different groups of K. Among them, the average value matrix of the corresponding pixels of the non-uniformity-corrected multi-frame high-temperature point image can be calculated by using the mean-value matrix relational expression to obtain the high-temperature mean-value matrix; the non-uniformity-corrected multi-frame The time-domain extremum absolute value matrix of the pixels corresponding to the high-temperature point image is obtained to obtain the high-temperature time-domain extremum absolute value matrix. The mean matrix relation can be expressed as:

The time-domain extremum calculation relation can be expressed as:

In the formula, meanFigT1(i,j) is the high temperature mean matrix, minmaxFigT1(i,j) is the absolute value matrix of high temperature time domain extremum, Region={1,2,3,…,num}, num is the number of high temperature point images The total number of frames, figT1 _frame (i, j) is the value of row i and column j of the corrected high-temperature point image in the frame frame when the temperature is T1.

Traversing through each element in the absolute value matrix of extreme values in the high temperature time domain, if the current element value in the absolute value matrix of high temperature time domain is greater than the first preset threshold or a multiple of any element value in the mean value matrix, the current element The pixels corresponding to the values are classified into the first type of temporal blind pixel set. The value of a certain point in the time-domain extreme value absolute value matrix minmaxFigT1 is greater than the first preset threshold A1 or a multiple of any element value in the mean value matrix, then the point is a blind element, and the entire time-domain extreme value absolute value matrix minmaxFigT1 is traversed, All the blind cells obtained are time-domain blind cells 1, which can be recorded as map21, and map21 can be expressed as:

或

or

Calculate the mean value matrix of the corresponding pixels and the time-domain extremum absolute value matrix of the corresponding pixels in the non-uniformity-corrected multi-frame low-temperature point images, and obtain the low-temperature mean value matrix and the low-temperature time-domain extremum absolute value matrix. Among them, the low-temperature average value matrix can be used to calculate the average value matrix of the pixels corresponding to the multi-frame high-temperature point images after non-uniformity correction, and the low-temperature average value matrix can be obtained; the non-uniformity-corrected The time-domain extremum absolute value matrix of pixels corresponding to multiple frames of high-temperature point images is obtained to obtain the low-temperature time-domain extremum absolute value matrix. The low temperature mean matrix relation can be expressed as:

The absolute value matrix of low temperature time domain extremum can be expressed as:

In the formula, figT2 _frame (i, j) indicates the value of row i, column j of the corrected picture in the frame frame when the temperature is T2, Region={1,2,3,...,num}.

Traversing through each element in the low temperature time domain extreme value absolute value matrix, if the current element value in the low temperature time domain extreme value absolute value matrix is greater than the fifth preset threshold value or a multiple of any element value in the mean value matrix, the current element The pixels corresponding to the values are classified into the second type of temporal blind pixel set. The value of a certain point in the time-domain extreme value absolute value matrix minmaxFigT2 is greater than the fifth preset threshold A2 or a multiple of any element value in the mean value matrix, then the point is a blind element, and the entire time-domain extreme value absolute value matrix is traversed to obtain All blind cells are time-domain blind cells 2, which can be marked as map22, and map22 can be expressed as:

或者

or

The first type of time-domain blind element set and the second type of time-domain blind element set are combined to obtain flash blind elements used for determining the calibration process.

In addition, the embodiment of the present invention can also use the time domain to calculate the average value, and then subtract the average value from each number to obtain a similar effect by calculating the maximum value of the absolute value, and can also detect the periodic sudden change in the calibration process. Blind Yuan.

It can be seen from the above that the embodiment of the present invention aims at the judgment of flashing blind elements in the related art, which is often calculated and compared according to the variance of each point in the time domain, and the amount of calculation is relatively large, and because the variance of the blind element points with periodic mutations is overwhelmed by many The value of the points after averaging is small, and such blind elements are often missed. However, the method of calculating the absolute value of the difference between the maximum and minimum values in the time domain can not only reduce the calculation amount, but also effectively detect the periodic mutation blind element.

As another optional implementation, this application also provides a calculation method for responding to blind elements, which may include the following content:

The embodiment of the present invention can calculate the average gray value of multiple frames of high-temperature point images and low-temperature point images after non-uniformity correction, so as to obtain the high-temperature average value and the low-temperature average value. Then the high temperature mean value and the low temperature mean value are subtracted to obtain a response matrix, and the response matrix can be calculated by using the following relational expression response(i,j)=meanFigT1(i,j)-meanFigT2(i,j), for example, the response matrix Calculate the mean value of all the pixels of the entire area array to obtain the mean value of the area array; for example, the mean value of the area array can be calculated using the formula for calculating the mean value of the area array, and the formula for calculating the mean value of the area array can be expressed as:

In the formula, response(i,j) is the response matrix, meanResponse is the mean value of the surface array, m*n is the length and width of the surface array, meanFigT1(i,j) is the high-temperature point image after non-uniformity correction at temperature T1 When is the mean value matrix of the corresponding pixel points, meanFigT2(i,j) is the mean value matrix of the corresponding pixel points of the high-temperature point image after non-uniformity correction when the temperature is T2.

If the difference between the current element value in the response matrix and the mean value of the surface array is greater than the preset sixth preset threshold threshold6, the pixel corresponding to the current element value is determined to be a response blind element, and the response blind element map3(i, j) The judgment process can be expressed as:

In addition, in the embodiment of the present invention, when the signal value of a certain pixel in the response matrix>sixth threshold*mean value of the area array, it can be determined as a blind response pixel.

It can be seen from the above that the embodiment of the present invention aims at the judgment based on the response blind element in the related art, which is often judged according to the gray level difference between the original uncorrected high-temperature target and the low-temperature target, without considering the impact of correction on the high-temperature target and low-temperature target, the threshold Improper setting can easily lead to misjudgment of blind elements. This application uses the corrected high-temperature target and low-temperature target for judgment. Comparing Figures 4-7, it can be seen that considering the impact of non-uniformity correction on high and low temperature targets, to a certain extent It can avoid the misjudgment of blind elements and improve the detection accuracy of blind elements.

It should be noted that there is no strict order of execution among the steps in this application, as long as they conform to the logical order, these steps can be executed at the same time, or in a certain preset order, and Figure 1 is only a schematic way , does not mean that it can only be executed in this order.

The embodiment of the present invention also provides a corresponding device for the infrared blind element detection method, which further makes the method more practical. Wherein, the device can be described separately from the perspective of functional modules and hardware. The following is an introduction to the blind infrared pixel detection device provided by the embodiment of the present invention. The blind infrared pixel detection device described below and the blind infrared pixel detection method described above can be referred to in correspondence.

Based on the perspective of functional modules, refer to FIG. 8, which is a structural diagram of an infrared blind element detection device provided in an embodiment of the present invention in a specific implementation manner. The device may include:

The calibration blind element detection module 801 is used to perform infrared blind element detection using a pre-generated blind element table during the production process of the infrared imaging device; the blind element table is composed of model blind elements and/or flash blind elements and/or response blind elements The fixed blind elements are combined to determine the fixed blind elements in the calibration process; the model blind elements are determined according to the gain value range of the non-uniformity correction model, and the flash blind elements are the high-temperature point images and low-temperature point images after non-uniformity correction The determination of the relationship between the absolute value of the extreme value in the time domain and the first preset threshold, and the response blind element is the determination of the relationship between the response matrix of the high temperature point and the low temperature point after non-uniformity correction and the second preset threshold.

The sliding window processing module 802 is configured to perform sliding window processing on the infrared image to be output according to preset neighborhood values during the user's use of the infrared imaging device, to obtain the maximum and minimum values of each window.

The random flash blind element detection module 803 is used to determine the random flash blind element in the scene based on the relationship between the variance of the infrared image to be output and the third preset threshold if the infrared image to be output meets the preset judgment condition; the preset judgment condition is based on The relationship among the maximum value, the minimum value and the fourth preset threshold value is determined.

Optionally, in some implementations of this embodiment, the calibration blind element detection module 801 includes a blind element table construction submodule, and the blind element table construction submodule may include a model blind element determination unit, and the model blind element The element determination unit is used to collect multiple frames of continuous high-temperature images and multiple frames of continuous low-temperature images using the black body calibration method; the gain value calculated by using the non-uniformity corrected image is used to calculate the cumulative probability density distribution function to obtain the interval of the gain value distribution. Compared with the gain value within the preset value range; the pixel whose current gain value is not within the preset value range is determined as the model pixel.

In other implementations of this embodiment, the blind element table construction submodule may include a flash blind element determination unit, and the flash blind element determination unit may be used to calculate pixels corresponding to multi-frame high-temperature spot images after non-uniformity correction Point mean matrix and time-domain extremum absolute value matrix of corresponding pixels, to obtain high-temperature mean matrix and high-temperature time-domain extremum absolute value matrix; traverse each element in high-temperature time-domain extremum absolute value matrix, if high-temperature time-domain If the current element value in the extreme value absolute value matrix is greater than the first preset threshold value or a multiple of any element value in the mean value matrix, then the pixel corresponding to the current element value is classified into the first type of time-domain blind element set; The uniformity-corrected multi-frame low-temperature point image corresponds to the mean value matrix of the pixels and the time-domain extremum absolute value matrix of the corresponding pixels to obtain the low-temperature mean value matrix and the low-temperature time-domain extremum absolute value matrix; traversing the low-temperature time-domain extremum absolute value matrix For each element in the value matrix, if the current element value in the low temperature time domain extreme value absolute value matrix is greater than the fifth preset threshold or the multiple of any element value in the mean value matrix, the pixel corresponding to the current element value is classified to the second type of time domain blind cell set; merging the first type of time domain blind cell set and the second type of time domain blind cell set to obtain flash blind cells used for determining the calibration process.

Optionally, in some further implementations of the embodiments of the present invention, the blind element table construction submodule may include a response blind element determination unit, and the response blind element determination unit may be used to calculate multi-frame high temperature after non-uniformity correction The gray value of the point image and the low temperature point image is obtained to obtain the high temperature average value and the low temperature average value; the high temperature average value and the low temperature average value are subtracted to obtain a response matrix; the response matrix is calculated for all pixels of the entire area array to obtain the area array average value; If the difference between the current element value in the response matrix and the mean value of the area array is greater than a preset sixth preset threshold, the pixel corresponding to the current element value is determined to be a response blind pixel.

As another optional implementation manner, the random flash blind element detection module 803 can also be used if the maximum value and the minimum value of M*N centered on (i, j) of the infrared image outFig to be output are tempMax, The minimum value tempMin, the default judgment condition is:

Calculate the variance variance(i,j) of the infrared image to be output after removing the center point, and the random flash blind element is judged according to the judgment relation, which is:

Among them, threshold3 is the third preset threshold, and threshold4 is the fourth preset threshold; if map4(i, j) is 1, the pixels corresponding to the horizontal and vertical coordinates of the infrared image to be output (i, j) are randomly flashing Blind element, if map4(i, j) is 0, then the pixel corresponding to (i, j) is not a random flash blind element.

The functions of each functional module of the infrared blind element detection device described in the embodiment of the present invention can be specifically implemented according to the method in the above method embodiment, and the specific implementation process can refer to the relevant description of the above method embodiment, and will not be repeated here.

It can be seen from the above that the embodiments of the present invention can quickly and accurately detect any type of blind element without increasing the process time, and reduce the occurrence of blind element misjudgment.

The infrared blind element detection device mentioned above is described from the perspective of functional modules. Further, the present application also provides an infrared blind element detection device, which is described from the perspective of hardware. FIG. 9 is a structural diagram of another infrared blind element detection device provided by an embodiment of the present application. As shown in Figure 9, the device includes a memory 90 for storing computer programs;

The processor 91 is configured to implement the steps of the infrared blind element detection method mentioned in the above embodiment when executing the computer program.

Wherein, the processor 91 may include one or more processing cores, such as a 4-core processor, an 8-core processor, and the like. Processor 91 can adopt at least one hardware form in DSP (Digital Signal Processing, digital signal processing), FPGA (Field-Programmable Gate Array, field programmable gate array), PLA (Programmable Logic Array, programmable logic array) accomplish. Processor 91 may also include a main processor and a coprocessor, and the main processor is a processor for processing data in a wake-up state, also known as a CPU (Central Processing Unit, central processing unit); the coprocessor is used to Low-power processor for processing data in standby state. In some embodiments, the processor 91 may be integrated with a GPU (Graphics Processing Unit, image processor), and the GPU is used for rendering and drawing the content to be displayed on the display screen. In some embodiments, the processor 91 may further include an AI (Artificial Intelligence, artificial intelligence) processor, where the AI processor is configured to process computing operations related to machine learning.

Memory 90 may include one or more computer-readable storage media, which may be non-transitory. The memory 90 may also include high-speed random access memory, and non-volatile memory, such as one or more magnetic disk storage devices, flash memory storage devices. In this embodiment, the memory 90 is at least used to store the following computer program 901 , wherein, after the computer program is loaded and executed by the processor 91 , it can realize the relevant steps of the infrared blind element detection method disclosed in any of the foregoing embodiments. In addition, the resources stored in the memory 90 may also include an operating system 902 and data 903, etc., and the storage method may be temporary storage or permanent storage. Wherein, the operating system 902 may include Windows, Unix, Linux and so on. The data 903 may include, but is not limited to, data corresponding to the infrared blind element detection result and the like.

In some embodiments, the infrared blind element detection device may further include a display screen 92 , an input/output interface 93 , a communication interface 94 , a power supply 95 and a communication bus 96 .

Those skilled in the art can understand that the structure shown in FIG. 9 does not constitute a limitation to the infrared blind element detection device, and may include more or less components than shown in the figure, such as a sensor 97 .

The functions of each functional module of the infrared blind element detection device described in the embodiment of the present invention can be specifically implemented according to the method in the above method embodiment, and the specific implementation process can refer to the relevant description of the above method embodiment, and will not be repeated here.

It can be seen from the above that the embodiments of the present invention can quickly and accurately detect any type of blind element without increasing the process time, and reduce the occurrence of blind element misjudgment.

It can be understood that if the infrared blind element detection method in the above embodiments is implemented in the form of a software function unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application is essentially or part of the contribution to the prior art or all or part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a storage medium , executing all or part of the steps of the methods in the various embodiments of the present application. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (Read-Only Memory, ROM), random access memory (Random Access Memory, RAM), electrically erasable programmable ROM, registers, hard disk, programmable Various media that can store program codes such as removable disks, CD-ROMs, magnetic disks, or optical disks.

Based on this, an embodiment of the present invention also provides a computer-readable storage medium, which stores a blind infrared element detection program, and when the blind infrared element detection program is executed by a processor, it is as described in any one of the above embodiments. A step of.

The functions of each functional module of the computer-readable storage medium in the embodiments of the present invention can be specifically implemented according to the methods in the above-mentioned method embodiments, and the specific implementation process can refer to the relevant descriptions of the above-mentioned method embodiments, which will not be repeated here.

It can be seen from the above that the embodiments of the present invention can quickly and accurately detect any type of blind element without increasing the process time, and reduce the occurrence of blind element misjudgment.

Each embodiment in this specification is described in a progressive manner, each embodiment focuses on the difference from other embodiments, and the same or similar parts of each embodiment can be referred to each other. As for the device disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and for relevant details, please refer to the description of the method part.

Professionals can further realize that the units and algorithm steps of the examples described in conjunction with the embodiments disclosed herein can be implemented by electronic hardware, computer software or a combination of the two. In order to clearly illustrate the possible For interchangeability, in the above description, the composition and steps of each example have been generally described according to their functions. Whether these functions are executed by hardware or software depends on the specific application and design constraints of the technical solution. Skilled artisans may use different methods to implement the described functions for each specific application, but such implementation should not be regarded as exceeding the scope of the present invention.

The above is a detailed introduction of the infrared blind element detection method, device and computer-readable storage medium provided by the present application. In this paper, specific examples are used to illustrate the principle and implementation of the present invention, and the descriptions of the above embodiments are only used to help understand the method and core idea of the present invention. It should be pointed out that those skilled in the art can make several improvements and modifications to the application without departing from the principle of the present invention, and these improvements and modifications also fall within the protection scope of the claims of the application.

Claims (9)

1. A blind infrared element detection method, characterized in that, comprising: In the production process of infrared imaging equipment, infrared blind element detection is performed using a pre-generated blind element table; the blind element table is obtained by combining model blind elements and/or flash blind elements and/or response blind elements for determining calibration Fixed blind elements in the process; During the user's use of the infrared imaging device, sliding window processing is performed on the infrared image to be output according to the preset neighborhood value to obtain the maximum and minimum values of each window; If the infrared image to be output satisfies a preset determination condition, determine a random flash blind element in the scene based on the relationship between the variance of the infrared image to be output and a third preset threshold; Among them, the blind element of the model is determined according to the gain value range of the non-uniformity correction model, and the blind element of the flash is the absolute value of the time-domain extremum and the first preset threshold of the high-temperature point image and the low-temperature point image after using the non-uniformity correction The relationship judgment between the response blind element is the judgment of the relationship between the response matrix of the high temperature point and the low temperature point after non-uniformity correction and the second preset threshold; the preset judgment condition is based on the maximum value, the determining the relationship between the minimum value and the fourth preset threshold; Wherein, the response blind element is the determination of the relationship between the response matrix of the high temperature point and the low temperature point after non-uniformity correction and the second preset threshold value, including: Calculate the gray value of the multi-frame high-temperature point image and low-temperature point image after non-uniformity correction, and obtain the high-temperature average value and low-temperature average value; Subtracting the high temperature mean value from the low temperature mean value to obtain the response matrix; Calculate the average value of the response matrix to all pixels of the entire array to obtain the average value of the array; If the difference between the current element value in the response matrix and the mean value of the area array is greater than a preset sixth preset threshold, the pixel corresponding to the current element value is determined to be the response blind pixel. 2. The infrared blind element detection method according to claim 1, wherein the model blind element is determined according to the range of gain values of the non-uniformity correction model and includes: Use the blackbody calibration method to collect multiple frames of continuous high-temperature images and multiple frames of continuous low-temperature images; Calculating a cumulative probability density distribution function using the gain value calculated from the non-uniformity-corrected image to obtain a gain value whose interval proportion of the gain value distribution is within a preset value range; A pixel whose current gain value is not within the preset value range is determined as the blind model pixel. 3. The infrared blind element detection method according to claim 1, characterized in that, the blind flash element is the absolute value of the time domain extremum and the first preset value of the high-temperature point image and the low-temperature point image after using the non-uniformity correction. The determination of the relationship between the thresholds includes: Calculate the mean value matrix of the corresponding pixels and the time-domain extremum absolute value matrix of the corresponding pixels in the multi-frame high-temperature point image after non-uniformity correction, and obtain the high-temperature mean value matrix and the high-temperature time-domain extremum absolute value matrix; Traversing each element in the high-temperature time-domain extreme value absolute value matrix, if the current element value in the high-temperature time-domain extreme value absolute value matrix is greater than the first preset threshold or any element in the mean value matrix value, the pixel corresponding to the current element value is classified into the first type of time-domain blind element set; Calculate the mean value matrix of the corresponding pixel points and the time-domain extremum absolute value matrix of the corresponding pixel points in the non-uniformity-corrected multi-frame low-temperature point image, and obtain the low-temperature mean value matrix and the low-temperature time-domain extremum absolute value matrix; Traversing each element in the low-temperature time-domain extreme value absolute value matrix, if the current element value in the low-temperature time-domain extreme value absolute value matrix is greater than the fifth preset threshold or any element value in the mean value matrix multiple, the pixel corresponding to the current element value is classified into the second type of temporal blind element set; Combining the first type of time domain blind cell set and the second type of time domain blind cell set to obtain flash blind cells used for determining the calibration process. 4. The infrared blind element detection method according to claim 3, wherein said calculating the mean value matrix of the corresponding pixels of the multi-frame high-temperature point images after the non-uniformity correction and the absolute value of the time-domain extremum of the corresponding pixels Matrix, to obtain the high temperature mean value matrix and the high temperature time domain extreme value absolute value matrix include: Utilize the mean value matrix relational expression to calculate the mean value matrix of the corresponding pixels of the multi-frame high-temperature point image after non-uniformity correction, obtain the high temperature mean value matrix; the described mean value matrix relational expression is:

The time-domain extremum absolute value matrix of the pixels corresponding to the multi-frame high-temperature point image after non-uniformity correction is calculated by using the time-domain extremum calculation relation to obtain the high-temperature time-domain extremum absolute value matrix, and the time-domain extremum calculation relationship The formula is:

In the formula, meanFigT1(i, j) is the high temperature mean value matrix, minmaxFigT1(i, j) is the absolute value matrix of the high temperature time domain extremum, Region={1,2,3,...,num}, num is The total number of frames of the high-temperature point image, figT1 _frame (i, j) is the value of the i-th row and j-th column of the frame-th frame of the high-temperature point image after correction when the temperature is T1. 5. infrared blind element detection method according to claim 1, is characterized in that, described response matrix calculates mean value to all pixels of whole area array and obtains area array mean value as: The mean value of the area array is calculated using the mean calculation relation of the area array, and the mean value calculation relation of the area array is:

response(i,j)=meanFigT1(i,j)-meanFigT2(i,j); In the formula, response(i,j) is the response matrix, meanResponse is the mean value of the array, m*n is the length and width of the array, meanFigT1(i,j) is the high temperature after non-uniformity correction The point image corresponds to the mean value matrix of the pixels when the temperature is T1, and meanFigT2(i,j) is the mean value matrix of the corresponding pixels of the high-temperature point image after non-uniformity correction when the temperature is T2. 6. The infrared blind element detection method according to any one of claims 1 to 5, wherein the preset neighborhood value is M*N, and if the infrared image to be output meets the preset judgment condition , the random flash blind element in the scene determined based on the relationship between the variance of the infrared image to be output and the third preset threshold includes: If the maximum value and the minimum value of the M*N centered on (i, j) of the infrared image to be output are respectively tempMax and minimum value tempMin, the preset determination condition is:

Calculate the variance variance (i, j) after the central point of the infrared image to be output is removed, and the random flash blind element is judged according to the determination relational expression, and the determination relational expression is:

Wherein, threshold3 is the third preset threshold, and threshold4 is the fourth preset threshold; if map4(i, j) is 1, the horizontal and vertical coordinates of the infrared image to be output are (i, j) corresponding The pixel of is the random blinking element, if map4(i, j) is 0, then the pixel corresponding to (i, j) is not the random blinking element. 7. An infrared blind element detection device, characterized in that, comprising: Calibrate the blind element detection module, used in the production process of infrared imaging equipment, use the pre-generated blind element table to perform infrared blind element detection; the blind element table is composed of model blind elements and/or flash blind elements and/or response Blind elements are merged to obtain fixed blind elements used to determine the calibration process; model blind elements are determined based on the gain value range of the non-uniformity correction model, and flash blind elements are high-temperature point images and low-temperature point images after non-uniformity correction The determination of the relationship between the absolute value of the time domain extreme value and the first preset threshold, and the response blind element is the determination of the relationship between the response matrix of the high temperature point and the low temperature point after non-uniformity correction and the second preset threshold; A sliding window processing module, configured to perform sliding window processing on the infrared image to be output according to a preset neighborhood value during the user's use of the infrared imaging device, to obtain the maximum and minimum values of each window; The random flash blind element detection module is used to determine the random flash blind element in the scene based on the relationship between the variance of the infrared image to be output and the third preset threshold if the infrared image to be output meets the preset determination condition; The preset determination condition is determined based on the relationship between the maximum value, the minimum value and a fourth preset threshold; Wherein, the calibration blind element detection module is further used for: Calculate the gray value of the multi-frame high-temperature point image and low-temperature point image after non-uniformity correction, and obtain the high-temperature average value and low-temperature average value; Subtracting the high temperature mean value from the low temperature mean value to obtain the response matrix; Calculate the average value of the response matrix to all pixels of the entire array to obtain the average value of the array; If the difference between the current element value in the response matrix and the mean value of the area array is greater than a preset sixth preset threshold, the pixel corresponding to the current element value is determined to be the response blind pixel. 8. An infrared blind element detection device, characterized in that it includes a processor, and the processor is used to implement the steps of the infrared blind element detection method according to any one of claims 1 to 6 when executing the computer program stored in the memory . 9. A computer-readable storage medium, characterized in that, the computer-readable storage medium is stored with an infrared blind element detection program, and when the infrared blind element detection program is executed by a processor, it realizes any of claims 1 to 6. A step of the infrared blind element detection method.

Images