Uncooled detector cross grain correction method

Abstract

The invention relates to the field of image processing, in particular to an uncooled detector cross grain correction method and device and a medium. According to the method, firstly, a 3 * 5 convolution template is adopted to traverse the image, the position where the cross grain appears in the image for the first time is obtained through the convolution template, then the 3 * 3 convolution template is adopted to distinguish whether the cross grain is a continuous cross grain or not, and different weight coefficients are given to different rows in the convolution process to reflect different influences caused by different positions. According to the method, the images are processed by setting different convolution operators, a small section of cross grain appearing in one row of images can be processed in a targeted mode, meanwhile, cross grain noise crossing multiple rows of pixels can be removed, appearing different cross grain noise can be processed more flexibly and specifically, and meanwhile interference on the images is reduced.

Description

Technical Field

[0001] The invention relates to the field of image processing, and in particular to a method for correcting horizontal stripes of an uncooled detector. Background Art

[0002] The existing algorithms for removing horizontal stripes are more focused on the horizontal stripe noise caused by the different responses of each row of the detector. This noise is static and does not change with the scene, so the existing algorithms cannot effectively remove the dynamic horizontal stripe noise caused by various interferences.

[0003] The existing algorithm for removing horizontal stripes must first filter the image, and then use the difference between the original image and the filtered image to obtain the horizontal component to obtain the horizontal stripes. However, the horizontal edge of the image may interfere with the removal of noise, so it is necessary to select a suitable threshold to distinguish between horizontal stripes and edges. This method removes horizontal stripes in the entire line, which may weaken the horizontal details of the image. At the same time, it requires multiple steps to implement, consumes more resources, and cannot be applied to different hardware. Summary of the invention

[0004] In view of the defects of the prior art, the present invention provides a method for correcting horizontal stripes of an uncooled detector, which uses an algorithm to effectively remove horizontal stripe noise caused by interference during the imaging process of the detector.

[0005] In order to solve the technical problem, the technical solution adopted by the present invention is: a method for correcting horizontal stripes of an uncooled detector, comprising the following steps: S01, receiving an input image, dividing the input image by pixels, taking the first 5 rows of the input image and performing convolution operation with a 3*5 convolution template, judging the convolution result F, if the convolution result F of a row is less than or equal to the threshold TH1, it is considered that the pixel of the row undergoing convolution operation is horizontal stripe noise, if the convolution result F of a row is greater than the threshold TH1, it is considered that the pixel of the row undergoing convolution operation is an edge; through this step, the position where the horizontal stripes first appear in the input image is obtained; S02, determine whether horizontal stripe noise occurs in the last convolution operation. If horizontal stripe noise does not occur, continue to use the 3*5 convolution template to perform convolution operation and convolution result judgment with the next 5 adjacent rows of the image. If horizontal stripe noise occurs, take the row where the horizontal stripe noise occurs last as the starting position, take 3 rows of images from the starting position to perform convolution operation with the 3*3 convolution template, and judge the convolution result H. If the convolution result H of a row is less than or equal to the threshold TH2, it is considered that the pixel of the row undergoing convolution operation is horizontal stripe noise. If the convolution result H of a row is greater than the threshold TH2, it is considered that the pixel of the row undergoing convolution operation is an edge. The convolution result F represents the result of the convolution operation between the input image and the 3*5 convolution template, and the convolution result H represents the result of the convolution operation between the input image and the 3*3 convolution template; S03, repeating step S02 to obtain all the horizontal stripe noise on the input image, and superimposing the obtained horizontal stripe noise result with the input image data to obtain an image without the horizontal stripe noise.

[0006] Furthermore, when the input image is convolved with the 3*5 convolution template, the weight of each row of the input image participating in the convolution operation is calculated. The calculation process is: S11. Calculate the pixel mean and variance of each row of the input image, and set the variance weight according to the variance. If the variance of a row is greater than the threshold TH3, the variance weight of the row is 0. If the variance of a row is less than the threshold TH3, the variance weight of the row is calculated according to the following formula: Variance weight = 1 / (variance + 1); The mean weight is set according to the mean difference. If the mean difference between a row and the middle row is greater than or equal to the threshold TH4, the mean weight of the row is 0. If the mean difference between a row and the middle row is less than the threshold TH4, the mean weight of the row is calculated according to the following formula: Mean weight = (threshold TH4-mean difference) / threshold TH4; S12. Calculate the weight of each row participating in the convolution operation based on the variance weight and mean weight. Weight 1 = variance weight 1 + mean weight 1; Weight 2 = variance weight 2 + mean weight 2 + 1; Weight 3 = Weight 1 + Weight 2 + Weight 4 + Weight 5; Weight 4 = variance weight 4 + mean weight 4 + 1; Weight 5 = variance weight 5 + mean weight 5; Among them, weight 1, variance weight 1, and mean weight 1 respectively represent the weight, variance weight, and mean weight of the first row of the input image participating in the convolution operation; weight 2, variance weight 2, and mean weight 2 respectively represent the weight, variance weight, and mean weight of the second row of the input image participating in the convolution operation; weight 4, variance weight 4, and mean weight 4 respectively represent the weight, variance weight, and mean weight of the fourth row of the input image participating in the convolution operation; weight 5, variance weight 5, and mean weight 5 respectively represent the weight, variance weight, and mean weight of the fifth row of the input image participating in the convolution operation; and weight 3 represents the weight of the third row, i.e., the middle row, of the input image participating in the convolution operation.

[0007] The weight of the first row of the 3*3 convolutional template is 2, the circle in the second row is -3, and the weight of the third row is 1.

[0008] Furthermore, the horizontal noise result obtained in step S03 is a negative value.

[0009] Furthermore, threshold TH1 and threshold TH4 are horizontal noise, with a default value of 20; threshold TH2 is a 3*3 judgment threshold, with a default value of 10; and threshold TH3 is a variance threshold, with a default value of 10.

[0010] The present invention also discloses a non-refrigerated detector horizontal stripe correction device, comprising: An input image receiving module, configured to receive an input image and divide the input image into pixels; A first convolution module is configured to perform a convolution operation on the input image and a 3*5 convolution template; A first convolution result judgment module is configured to judge the convolution result of the first convolution module. If the convolution result F of a row is less than or equal to the threshold TH1, it is considered that the pixels in the row undergoing the convolution operation are horizontal stripe noise; if the convolution result F of a row is greater than the threshold TH1, it is considered that the pixels in the row undergoing the convolution operation are edges; A second convolution module is configured to perform a convolution operation on the input image and the 3*3 convolution template; The second convolution result judgment module is configured to judge the convolution result of the second convolution module. If the convolution result H of a row is less than or equal to the threshold TH2, the pixel of the row subjected to the convolution operation is considered to be a horizontal stripe noise. If the convolution result H of a row is greater than the threshold TH2, the pixel of the row subjected to the convolution operation is considered to be an edge. The convolution scheduling module is configured to determine the convolution template for convolution operation with the input image, and the determination is based on the determination results of the first convolution result determination module and / or the second convolution result determination module. If the determination result is that no horizontal stripe noise occurs, the 3*5 convolution template is continued to be used to perform convolution operation and convolution result determination with the adjacent last 5 rows of the image. If the determination result is that horizontal stripe noise occurs, the row where the horizontal stripe noise occurs last is used as the starting position, and 3 rows of images are taken from the starting position backward to perform convolution operation with the 3*3 convolution template.

[0011] The present invention also discloses a non-cooled detector horizontal stripe correction device, comprising a processor and a memory storing program instructions, wherein the processor is configured to execute the non-cooled detector horizontal stripe correction method as described above when running the program instructions.

[0012] The present invention also discloses a storage medium storing program instructions, wherein when the program instructions are run, the non-refrigerated detector horizontal stripe correction method as described above is executed.

[0013] Beneficial effects of the present invention: This method processes images by setting different convolution operators, and can specifically process a small section of horizontal stripes appearing in a row of images, and can remove horizontal stripe noise across multiple rows of pixels, and can more flexibly and specifically process different horizontal stripe noises that appear, while reducing interference to the image. It can effectively remove horizontal stripe noise caused by interference during detector imaging, and remove dynamic and static horizontal stripe noise in images while occupying very few resources, which is convenient to implement on a hardware platform, and can be applied to the needs of different scenarios. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] Figure 1 is a flow chart of the method described in Example 1; Figure 2 is a schematic diagram of the device described in Example 2; Figure 3 Schematic diagram of the device described in Example 3. DETAILED DESCRIPTION

[0015] In order to be able to understand the features and technical contents of the embodiments of the present disclosure in more detail, the implementation of the embodiments of the present disclosure is described in detail below in conjunction with the accompanying drawings. The attached drawings are for reference only and are not used to limit the embodiments of the present disclosure. In the following technical description, for the convenience of explanation, a full understanding of the disclosed embodiments is provided through multiple details. However, one or more embodiments can still be implemented without these details. In other cases, to simplify the drawings, well-known structures and devices can be simplified for display.

[0016] The terms "first", "second", etc. in the embodiments of the present disclosure are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the terms used in this way can be interchanged where appropriate, so as to describe the embodiments of the present disclosure described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions.

[0017] Unless otherwise stated, the term "plurality" means two or more.

[0018] In the embodiment of the present disclosure, the character " / " indicates that the preceding and following objects are in an "or" relationship. For example, A / B indicates: A or B.

[0019] The term "and / or" is a description of the association relationship between objects, indicating that three relationships can exist. For example, A and / or B means: A or B, or, A and B.

[0020] The term "correspondence" may refer to an association relationship or a binding relationship. The correspondence between A and B means that there is an association relationship or a binding relationship between A and B.

[0021] Example 1 This embodiment discloses a method for correcting horizontal stripes of an uncooled detector. Figure 1 As shown, the method comprises the following steps: S01, receiving an input image, dividing the input image by pixels, taking the first 5 rows of the input image and performing convolution operation with a 3*5 convolution template, judging the convolution result F, if the convolution result F of a row is less than or equal to the threshold TH1, it is considered that the pixel of the row undergoing convolution operation is horizontal stripe noise, if the convolution result F of a row is greater than the threshold TH1, it is considered that the pixel of the row undergoing convolution operation is an edge; through this step, the position where the horizontal stripes first appear in the input image is obtained; S02, determine whether horizontal stripe noise occurs in the last convolution operation. If horizontal stripe noise does not occur, continue to use the 3*5 convolution template to perform convolution operation and convolution result judgment with the next 5 adjacent rows of the image. If horizontal stripe noise occurs, take the row where the horizontal stripe noise occurs last as the starting position, take 3 rows of images from the starting position to perform convolution operation with the 3*3 convolution template, and judge the convolution result H. If the convolution result H of a row is less than or equal to the threshold TH2, it is considered that the pixel of the row undergoing convolution operation is horizontal stripe noise. If the convolution result H of a row is greater than the threshold TH2, it is considered that the pixel of the row undergoing convolution operation is an edge. The convolution result F represents the result of the convolution operation between the input image and the 3*5 convolution template, and the convolution result H represents the result of the convolution operation between the input image and the 3*3 convolution template; S03, repeating step S02 to obtain all the horizontal stripe noise on the input image, and superimposing the obtained horizontal stripe noise result with the input image data to obtain an image without the horizontal stripe noise.

[0022] Since the horizontal noise in the image is caused by interference, the length and width of the horizontal noise are random, so a 3*5 convolution template is used to traverse the image. In the 3*5 convolution template selected relative to the middle row, the values ​​of different rows have different effects on the middle row, with a greater impact close to the middle row and a smaller impact far from the middle row, so we give different weight coefficients to different rows to reflect the different effects caused by different positions.

[0023] This embodiment sets the weights adaptively for different weights in different situations. The appearance of horizontal stripes is for uniform scenery and background in a single scene. For complex scenery, the effect of horizontal stripes on imaging is very small, so adaptive weights are needed to handle different situations differently.

[0024] First, find the mean and variance of the pixels in the middle row. If the variance is too large, it means that the scene in this row is complex. At this time, the influence of horizontal stripes is small, and we can assume that there are no horizontal stripes in this row. If the variance is within the threshold TH3, we believe that the scene in this row is uniform and there may be horizontal stripes. At this time, we need to further determine whether there are horizontal stripes in this row. Similarly, we need to find the mean and variance of rows 1, 2, 4, and 5. If the variance is too large, it means that the scene in this row is complex and it is not conducive to removing horizontal stripes, so the weight of this row is 0. If the variance is within the threshold TH3, it means that the scene in this row is uniform and can be used to remove horizontal stripes. At this time, we need to compare from the mean. If the difference between other rows and the middle row is greater than the threshold TH4, it means that there is an edge between the rows. At this time, the weight is 0. If it is within the threshold TH4, it means that there are horizontal stripes.

[0025] Therefore, when the input image is convolved with the 3*5 convolution template, the weight of each row of the input image participating in the convolution operation is calculated based on multiple conditions. The calculation process is: S11. Calculate the pixel mean and variance of each row of the input image, and set the variance weight according to the variance. If the variance of a row is greater than the threshold TH3, the variance weight of the row is 0. If the variance of a row is less than the threshold TH3, the smaller the variance, the more uniform the input image. The default maximum variance weight is 1, and the smaller the variance, the higher the variance weight. The variance weight of the corresponding row is calculated according to the following formula: Variance weight = 1 / (variance + 1).

[0026] The mean weight is set according to the mean difference. If the mean difference between a row and the middle row is greater than or equal to the threshold TH4, the mean weight of the row is 0. If the mean difference between a row and the middle row is less than the threshold TH4, it means that there may be horizontal stripes in the row. If the scene balance transition difference is not large, the edge here may be considered as horizontal stripes. Therefore, when the mean difference is less than the threshold TH4, the mean weight of the corresponding row is calculated according to the following formula: Mean weight = (threshold TH4 - mean difference) / threshold TH4.

[0027] S12. Set weights based on the distance between each row and the middle row. When removing horizontal stripes, the closer to the middle row, the greater the effect on removing horizontal stripes. Therefore, different weights are assigned to different rows. Here, we believe that the weights of the first and fifth rows from the middle row are 1 less than the weights of the second and fourth rows close to the middle row. Specifically, the weight formula for each row participating in the convolution operation is calculated based on the variance weight and the mean weight: Weight 1 = variance weight 1 + mean weight 1; Weight 2 = variance weight 2 + mean weight 2 + 1; Weight 3 = Weight 1 + Weight 2 + Weight 4 + Weight 5; Weight 4 = variance weight 4 + mean weight 4 + 1; Weight 5 = variance weight 5 + mean weight 5; Among them, weight 1, variance weight 1, and mean weight 1 respectively represent the weight, variance weight, and mean weight of the first row of the input image participating in the convolution operation; weight 2, variance weight 2, and mean weight 2 respectively represent the weight, variance weight, and mean weight of the second row of the input image participating in the convolution operation; weight 4, variance weight 4, and mean weight 4 respectively represent the weight, variance weight, and mean weight of the fourth row of the input image participating in the convolution operation; weight 5, variance weight 5, and mean weight 5 respectively represent the weight, variance weight, and mean weight of the fifth row of the input image participating in the convolution operation; and weight 3 represents the weight of the third row, i.e., the middle row, of the input image participating in the convolution operation.

[0028] The above 3*5 template is used to obtain the position where the horizontal stripes first appear in the image. However, in reality, the horizontal stripes in the image may occupy 2 or even 3 rows. This situation needs to be handled separately. After the 3*5 template is judged, if there are no horizontal stripes, continue to traverse with the 3*5 template. If there are horizontal stripes, use the 3*3 template to traverse. The purpose of using the 3*3 convolution template is to distinguish whether the horizontal stripes are continuous. Therefore, different positions of the 3*3 convolution template also need to use different weight values. Based on the situation of horizontal stripes in the previous line, it can be divided into 4 cases, 111, 110, 100, and 101, where 1 is horizontal stripes and 0 is normal. The results obtained by convolution in the first and second cases of the above four cases are small. In summary, the weight of the first row of the 3*3 convolution template is set to 2, the circle of the second row is -3, and the weight of the third row is 1.

[0029] The horizontal stripe noise information determined by the 3*3 convolution template uses the noise information of the previous row of horizontal stripes. If there are horizontal stripes in the previous row, the 3*3 template is used, and if there are no horizontal stripes, the 3*5 template is used. That is to say, the 3*5 template is used to initially detect horizontal stripes, and the 3*3 template is used after the horizontal stripes are found. At this time, the role of the 3*3 template is to continue to determine whether the horizontal stripes found by the 3*5 template are wide horizontal stripes. If the 3*3 template is also determined to be horizontal stripes, it means that this horizontal stripe is a wide horizontal stripe that occupies multiple rows of pixels, so the noise information of the horizontal stripes in this row is the same as that of the previous row of horizontal stripes.

[0030] By traversing the image with the above two templates, we can obtain all the horizontal noise on the image, and superimpose the obtained horizontal noise result with the input image data to finally achieve the effect of removing the horizontal noise. The horizontal noise result obtained in this embodiment is a negative value, and superimposing the negative value on the input image will obtain an image with the horizontal noise removed.

[0031] Horizontal stripe noise appears because there is a difference with adjacent rows. The pixel value of this row may be higher or lower than that of adjacent rows, so the horizontal stripe noise obtained may be negative or positive.

[0032] In this embodiment, threshold TH1 and threshold TH4 are horizontal noise, and the default value is 20; threshold TH2 is a 3*3 judgment threshold, and the default value is 10; threshold TH3 is a variance threshold, and the default value is 10.

[0033] Example 2 This embodiment discloses a non-refrigerated detector horizontal stripe correction device, such as Figure 2 As shown, including: An input image receiving module, configured to receive an input image and divide the input image into pixels; A first convolution module is configured to perform a convolution operation on the input image and a 3*5 convolution template; A first convolution result judgment module is configured to judge the convolution result of the first convolution module. If the convolution result F of a row is less than or equal to the threshold TH1, it is considered that the pixels in the row undergoing the convolution operation are horizontal stripe noise; if the convolution result F of a row is greater than the threshold TH1, it is considered that the pixels in the row undergoing the convolution operation are edges; A second convolution module is configured to perform a convolution operation on the input image and the 3*3 convolution template; The second convolution result judgment module is configured to judge the convolution result of the second convolution module. If the convolution result H of a row is less than or equal to the threshold TH2, the pixel of the row subjected to the convolution operation is considered to be a horizontal stripe noise. If the convolution result H of a row is greater than the threshold TH2, the pixel of the row subjected to the convolution operation is considered to be an edge. The convolution scheduling module is configured to determine the convolution template for convolution operation with the input image, and the determination is based on the determination results of the first convolution result determination module and / or the second convolution result determination module. If the determination result is that no horizontal stripe noise occurs, the 3*5 convolution template is continued to be used to perform convolution operation and convolution result determination with the adjacent last 5 rows of the image. If the determination result is that horizontal stripe noise occurs, the row where the horizontal stripe noise occurs last is used as the starting position, and 3 rows of images are taken from the starting position backward to perform convolution operation with the 3*3 convolution template.

[0034] In this embodiment, the non-cooled detector horizontal stripe correction device also includes a first weight calculation module and a second weight calculation module. The first weight calculation module is used to calculate the weight of each row of the input image convolved with the 3*5 convolution template, and the calculation process is the same as that of Example 1. The second weight calculation module is used to set the weight of each row of the 3*3 convolution template. Specifically, the weight of the first row of the 3*3 convolution template is set to 2, the circle of the second row is -3, and the weight of the third row is 1.

[0035] Example 3 Combination Figure 3As shown, the embodiment of the present disclosure provides a non-cooling detector horizontal stripe correction device 300, including a processor (processor) 304 and a memory (memory) 301. Optionally, the device may also include a communication interface (Communication Interface) 302 and a bus 303. Among them, the processor 304, the communication interface 302, and the memory 301 can communicate with each other through the bus 303. The communication interface 302 can be used for information transmission. The processor 304 can call the logic instructions in the memory 301 to execute the non-cooling detector horizontal stripe correction method of Example 1.

[0036] In addition, the logic instructions in the memory 301 described above 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.

[0037] The memory 301 is a computer-readable storage medium that 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 304 executes the function application and data processing by running the program instructions / modules stored in the memory 301, that is, the non-cooled detector horizontal stripe correction method in Example 1 is implemented.

[0038] The memory 301 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 301 may include a high-speed random access memory and may also include a non-volatile memory.

[0039] Example 4 The 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 non-cooled detector horizontal stripe correction method described in Embodiment 1.

[0040] The computer-readable storage medium mentioned above may be a transient computer-readable storage medium or a non-transitory computer-readable storage medium.

[0041] 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.

[0042] 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.

[0043] 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 correcting horizontal stripes of an uncooled detector, characterized in that: The following steps are involved: S01, receiving an input image, dividing the input image by pixels, taking the first 5 rows of the input image and performing a convolution operation with a 3*5 convolution template, and judging the convolution result F. If the convolution result F of a row is less than or equal to a threshold value TH1, it is considered that the pixel in the row undergoing the convolution operation is a horizontal stripe noise. If the convolution result F of a row is greater than the threshold value TH1, it is considered that the pixel in the row undergoing the convolution operation is an edge. Through this step, the position where the horizontal stripes first appear in the input image is obtained; S02, determine whether horizontal stripe noise occurs in the last convolution operation. If horizontal stripe noise does not occur, continue to use the 3*5 convolution template to perform convolution operation and convolution result judgment with the next 5 adjacent rows of the image. If horizontal stripe noise occurs, take the row where the horizontal stripe noise occurs last as the starting position, take 3 rows of images from the starting position to perform convolution operation with the 3*3 convolution template, and judge the convolution result H. If the convolution result H of a row is less than or equal to the threshold TH2, it is considered that the pixel of the row undergoing convolution operation is horizontal stripe noise. If the convolution result H of a row is greater than the threshold TH2, it is considered that the pixel of the row undergoing convolution operation is an edge. The convolution result F represents the result of the convolution operation between the input image and the 3*5 convolution template, and the convolution result H represents the result of the convolution operation between the input image and the 3*3 convolution template; S03, repeating step S02 to obtain all the horizontal stripe noise on the input image, and superimposing the obtained horizontal stripe noise result with the input image data to obtain an image without the horizontal stripe noise.

2. The method for correcting horizontal stripes of an uncooled detector according to claim 1, characterized in that: When the input image is convolved with the 3*5 convolution template, the weight of each row of the input image participating in the convolution operation is calculated. The calculation process is: S11. Calculate the pixel mean and variance of each row of the input image, and set the variance weight according to the variance. If the variance of a row is greater than the threshold TH3, the variance weight of the row is 0. If the variance of a row is less than the threshold TH3, the variance weight of the row is calculated according to the following formula: Variance weight = 1 / (variance + 1); The mean weight is set according to the mean difference. If the mean difference between a row and the middle row is greater than or equal to the threshold TH4, the mean weight of the row is 0. If the mean difference between a row and the middle row is less than the threshold TH4, the mean weight of the row is calculated according to the following formula: Mean weight = (threshold TH4-mean difference) / threshold TH4; S12. Calculate the weight of each row participating in the convolution operation based on the variance weight and mean weight. Weight 1 = variance weight 1 + mean weight 1; Weight 2 = variance weight 2 + mean weight 2 + 1; Weight 3 = Weight 1 + Weight 2 + Weight 4 + Weight 5; Weight 4 = variance weight 4 + mean weight 4 + 1; Weight 5 = variance weight 5 + mean weight 5; Among them, weight 1, variance weight 1, and mean weight 1 respectively represent the weight, variance weight, and mean weight of the first row of the input image participating in the convolution operation; weight 2, variance weight 2, and mean weight 2 respectively represent the weight, variance weight, and mean weight of the second row of the input image participating in the convolution operation; weight 4, variance weight 4, and mean weight 4 respectively represent the weight, variance weight, and mean weight of the fourth row of the input image participating in the convolution operation; weight 5, variance weight 5, and mean weight 5 respectively represent the weight, variance weight, and mean weight of the fifth row of the input image participating in the convolution operation; and weight 3 represents the weight of the third row, i.e., the middle row, of the input image participating in the convolution operation.

3. The method for correcting horizontal stripes of an uncooled detector according to claim 1, characterized in that: The weight of the first row of the 3*3 convolutional template is 2, the circle in the second row is -3, and the weight of the third row is 1.

4. The method for correcting horizontal stripes of an uncooled detector according to claim 1, characterized in that: The horizontal noise result obtained in step S03 is a negative value.

5. The method for correcting horizontal stripes of an uncooled detector according to claim 1, characterized in that: Threshold TH1 and threshold TH4 are horizontal noise, the default value is 20, threshold TH2 is the 3*3 judgment threshold, the default value is 10, and threshold TH3 is the variance threshold, the default value is 10.

6. A non-cooled detector stripe correction device, characterized in that: include: An input image receiving module is configured to receive an input image and divide the input image into pixels; A first convolution module is configured to perform a convolution operation on the input image and a 3*5 convolution template; A first convolution result judgment module is configured to judge the convolution result of the first convolution module. If the convolution result F of a row is less than or equal to the threshold TH1, it is considered that the pixels in the row undergoing the convolution operation are horizontal stripe noise; if the convolution result F of a row is greater than the threshold TH1, it is considered that the pixels in the row undergoing the convolution operation are edges; A second convolution module is configured to perform a convolution operation on the input image and the 3*3 convolution template; The second convolution result judgment module is configured to judge the convolution result of the second convolution module. If the convolution result H of a row is less than or equal to the threshold TH2, the pixel of the row subjected to the convolution operation is considered to be a horizontal stripe noise. If the convolution result H of a row is greater than the threshold TH2, the pixel of the row subjected to the convolution operation is considered to be an edge. The convolution scheduling module is configured to determine the convolution template for convolution operation with the input image, and the determination is based on the determination results of the first convolution result determination module and / or the second convolution result determination module. If the determination result is that no horizontal stripe noise occurs, the 3*5 convolution template is continued to be used to perform convolution operation and convolution result determination with the adjacent last 5 rows of the image. If the determination result is that horizontal stripe noise occurs, the row where the horizontal stripe noise occurs last is used as the starting position, and 3 rows of images are taken from the starting position backward to perform convolution operation with the 3*3 convolution template.

7. A non-cooled detector stripe correction device, comprising a processor and a memory storing program instructions, characterized in that: The processor is configured to execute the non-cooled detector horizontal stripe correction method according to any one of claims 1 to 5 when running the program instructions.

8. A storage medium storing program instructions, characterized in that: When the program instructions are executed, the method for correcting horizontal stripes of an uncooled detector as described in any one of claims 1 to 5 is executed.

Images