Image dimming method applied to infrared tracking and collimating system

Abstract

The application discloses an image dimming method applied to an infrared tracking and sighting system, relates to the field of adaptive image dimming technology applied to the infrared tracking and sighting system, and comprises the following steps: a, performing median filtering on a 14-bit original image I0 collected by an infrared camera and then performing quantile truncation to obtain I1; b, normalizing I1 into an 8-bit image I2 and obtaining a binary template image I B by using a triangle thresholding method; c, dividing I1 into a high gray scale area and a low gray scale area according to I B ; d, performing improved histogram equalization on the high gray scale area and then performing normalization; and e, performing normalization on the low gray scale area, and synthesizing pixel values of the high and low gray scale areas into an 8-bit image I e after dimming. The application has the beneficial effects that the calculation method of the application can reach high efficiency in completing processing within 1ms after acceleration, and can reach balanced and good effects in highlighting image details, enhancing small and weak targets and weakening target fire halos, and the striking efficiency of the tracking and sighting system is improved by about 30%.

Description

Technical Field

[0001] This invention relates to the field of adaptive image dimming applied to infrared tracking systems, specifically an image dimming method applied to infrared tracking systems. Background Technology

[0002] In real-time infrared tracking systems, to achieve more stable tracking and detection under various scenarios, it is necessary to perform dimming processing on the images output by the infrared camera. This mainly includes the following three methods:

[0003] One approach is to apply a clahe-like algorithm to the original 14-bit image to adjust the data distribution. While this can suppress the halo effect after the target object catches fire, it is not effective for detecting small targets.

[0004] The second method is to use supervised learning in deep learning to train the dimming mapping, but the biggest problem with this method is that the image after dimming as required is quite rare.

[0005] Third, the grayscale values ​​of the image are piecewise linearly mapped. This method can perform well under certain parameters but cannot adapt to different parameter conditions (such as different focal lengths of the camera).

[0006] To summarize and analyze the above methods, we first need to consider that they maintain a beneficial effect on tracking and detection under various conditions; secondly, they need to complete the dimming in a short time; and finally, they should not be too restricted by manually set parameters. Taking these factors into account, we designed an image dimming method for infrared tracking systems that meets these requirements. Summary of the Invention

[0007] The purpose of this invention is to provide an image dimming method for use in infrared tracking systems, addressing the problems encountered. To achieve this purpose, the present invention provides the following technical solution: an image dimming method for use in infrared tracking systems, comprising the following steps:

[0008] Step a: Perform median filtering on the 14-bit original image I0 captured by the infrared camera and then truncate it to I1 by quantiles;

[0009] Step b: Normalize I1 to an 8-bit image I2 and use triangulation thresholding to obtain a binary template image I. B ;

[0010] Step c, according to I B I1 is divided into high grayscale area and low grayscale area;

[0011] Step d: Perform improved histogram equalization on the high grayscale areas and then normalize them.

[0012] Step e: Normalize the low grayscale areas and combine the pixel values ​​of the high and low grayscale areas into an 8-bit image I after dimming. e .

[0013] As a further aspect of the present invention: In step a: the 14-bit original image I0 captured by the infrared camera is subjected to median filtering and then truncated to quantiles to I1. Let the image after median filtering be I1. median Let quantile be α, gray level be denoted as g, minimum gray level in the image be m, X be the width of the image, Y be the height of the image, and maximum gray level be M. Then:

[0014] Cutoff threshold

[0015] The truncated image I1(x,y) = min{N,I median (x,y)}.

[0016] As a further aspect of the present invention: in step b, I1 is normalized to an 8-bit image I2 and thresholded using the triangular method to obtain a binary template image I. B The process is as follows:

[0017] The normalization process is as follows

[0018] Let the triangular thresholding operator be τ and the threshold be T, then T = τ(I2);

[0019] Binary template image I B For: I B (x,y)=1, ifI2(x,y)≥T;

[0020] I B (x,y)=0, ifI2(x,y)<T.

[0021] As a further embodiment of the present invention: in step c, according to I B Divide I1 into high-grayscale and low-grayscale regions: Let the high-grayscale region be H and the low-grayscale region be L, then H = {(x,y)|I B (x,y)=1}L={(x,y)|I B (x,y)=0}.

[0022] As a further aspect of the present invention: In step d, the high grayscale area is subjected to improved histogram equalization followed by normalization: Let k be the grayscale level after equalization, t k n is the quotient of the product of the frequency of the gray level and the height and width of the image. i It is the number of pixels with gray level i in the image. The improved histogram equalization is:

[0023]

[0024] The values ​​of each pixel in H are mapped as described above (i→t), and then normalized in the high grayscale region. Given 8-bit data, let the ordered set of values ​​be V(H).

[0025] As a further aspect of the present invention: in step e, the low grayscale area is normalized, and the pixel values ​​of the high and low grayscale areas are combined into an 8-bit image I after dimming. e The process is as follows:

[0026] For low grayscale regions, the normalized mapping is: Let V(L) be the ordered set of values, and let I be the process of combining high and low grayscale regions. e (x,y)={V(L)|(x,y)∈L}∪{V(H)|(x,y)∈H}.

[0027] Compared with the prior art, the beneficial effects of the present invention are: the calculation method of the present invention achieves high efficiency by completing the processing within 1ms after acceleration, and achieves a balanced and good effect in highlighting image details, enhancing weak targets, and reducing the halo of fire on the target, which is reflected in a significant improvement of the strike efficiency of about 30% in the tracking and aiming system. Attached Figure Description

[0028] Figure 1 This is a module diagram of an adaptive image dimming method applied to an infrared tracking system; Detailed Implementation

[0029] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0030] In the description of this invention, it should be noted that the terms "center," "upper," "lower," "left," "right," "vertical," "horizontal," "inner," and "outer," etc., indicating orientation or positional relationships, are based on the orientation or positional relationships shown in the accompanying drawings and are only for the convenience of describing this invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of this invention. Furthermore, the terms "first," "second," and "third" are used for descriptive purposes only and should not be construed as indicating or implying relative importance. In the description of this invention, it should be noted that unless otherwise explicitly specified and limited, the terms "installed," "connected," "linked," and "set up" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal communication of two components. Those skilled in the art can understand the specific meaning of the above terms in this invention based on the specific circumstances. The following describes embodiments of the invention based on its overall structure.

[0031] like Figure 1 As shown, the adaptive image dimming method for infrared tracking systems disclosed in this invention includes the following steps:

[0032] Step a: Perform median filtering on the 14-bit original image I0 acquired by the infrared camera, and then truncate it to I1 using quantiles. Let the image after median filtering be I1. median Let quantile be α, gray level be denoted as g, minimum gray level in the image be m, X be the width of the image, Y be the height of the image, and maximum gray level be M. Then:

[0033] Cutoff threshold

[0034] The truncated image I1(x,y) = min{N,I median (x,y)}.

[0035] Step b: Normalize I1 to an 8-bit image I2 and threshold it using the triangular method to obtain a binary template image I. B The process is as follows:

[0036] The normalization process is as follows

[0037] Let the triangular thresholding operator be τ and the threshold be T, then T = τ(I2);

[0038] Binary template image I B For: I B (x,y)=1, ifI2(x,y)≥T;

[0039] IB (x,y)=0, ifI2(x,y)<T.

[0040] Step c: According to I B Divide I1 into high-grayscale and low-grayscale regions: Let the high-grayscale region be H and the low-grayscale region be L, then H = {(x,y)|I B (x,y)=1}L={(x,y)|I B (x,y)=0}.

[0041] Step d: Perform improved histogram equalization on the high grayscale areas, followed by normalization: Let k be the grayscale level after equalization, t k n is the quotient of the product of the frequency of the gray level and the height and width of the image. i It is the number of pixels with gray level i in the image. The improved histogram equalization is:

[0042]

[0043] The values ​​of each pixel in H are mapped as described above (i→t), and then normalized in the high grayscale region. Given 8-bit data, let the ordered set of values ​​be V(H).

[0044] Step e: Normalize the low grayscale areas and combine the pixel values ​​of the high and low grayscale areas into an 8-bit image I after dimming. e The process is as follows:

[0045] For low grayscale regions, the normalized mapping is: Let V(L) be the ordered set of values, and let I be the process of combining high and low grayscale regions. e (x,y)={V(L)|(x,y)∈L}∪{V(H)|(x,y)∈H}.

[0046] Example 1:

[0047] In this embodiment, the parameters are an image bit depth of 14 bits and an image size of 640*512. Empirical parameters used in this system application only involve truncating high grayscale values ​​to avoid image flicker; through repeated experiments, α was determined to be 0.99. It should be noted that this method is also applicable to images with other bit depths, and α can be adjusted accordingly.

[0048] Step a: Perform median filtering on the 14-bit original image I0 acquired by the infrared camera, and then truncate it to I1 using quantiles. Let the image after median filtering be I1. median Let g be the gray level, m be the minimum gray level in the image, X be the width of the image, Y be the height of the image, and M be the maximum gray level.

[0049] Cutoff threshold

[0050] The truncated image I1(x,y) = min{N,I median (x,y)}.

[0051] Step b: Normalize I1 to an 8-bit image I2 and threshold it using the triangular method to obtain a binary template image I. B The process is as follows:

[0052] The normalization process is as follows

[0053] Let the triangular thresholding operator be τ, and the threshold be T, then T = τ(I²).

[0054] Binary template image I B For: I B (x,y)=1, ifI2(x,y)≥T;

[0055] I B (x,y)=0, ifI2(x,y)<T.

[0056] Step c: According to I B Divide I1 into high-grayscale and low-grayscale regions: Let the high-grayscale region be H and the low-grayscale region be L, then

[0057] H={(x,y)|I B (x,y)=1}L={(x,y)|I B (x,y)=0}.

[0058] Step d: Perform improved histogram equalization on the high grayscale areas, followed by normalization: Let k be the grayscale level after equalization, t k The quotient n is the product of the frequency of the gray level and the height and width of the image. i , which is the number of pixels with gray level i in the image. The improved histogram equalization is:

[0059]

[0060] The values ​​of each pixel in H are mapped as described above (i→t), and then normalized in the high grayscale region.

[0061] Given 8-bit data, let the ordered set of values ​​be V(H).

[0062] Step e: Normalize the low grayscale areas and combine the pixel values ​​of the high and low grayscale areas into an 8-bit image I after dimming. e The process is as follows:

[0063] For low grayscale regions, the normalized mapping is: Let V(L) be the ordered set of values, and let I be the process of combining high and low grayscale regions. e(x,y)={V(L)|(x,y)∈L}∪{V(H)|(x,y)∈H},

[0064] After adjusting the brightness, the image is then sent to the detection and tracking programs.

[0065] The above description is merely a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. An image dimming method applied to an infrared tracking system, characterized in that, Includes the following steps: Step a: Perform median filtering on the 14-bit original image I0 captured by the infrared camera and then truncate it to I1 by quantiles; Step b: Normalize I1 to an 8-bit image I2 and use triangulation thresholding to obtain a binary template image I. B ; Step c, according to I B I1 is divided into high grayscale area and low grayscale area; Step d: Perform improved histogram equalization on the high grayscale areas and then normalize them. In step d, the high grayscale area is subjected to improved histogram equalization followed by normalization as follows: Let k be the gray level after equalization, t k n is the quotient of the product of the frequency of the gray level and the height and width of the image. i It is the number of pixels with gray level i in the image. The improved histogram equalization is: Where X is the width of the image and Y is the height of the image; that is, the value of each pixel in H is mapped as i→t as described above, and then normalized in the high grayscale region. Given 8-bit data, let the ordered set of values ​​be V(H); Step e: Normalize the low grayscale areas and combine the pixel values ​​of the high and low grayscale areas into an 8-bit image I after dimming. e ; In step e, the low grayscale area is normalized, and the pixel values ​​of the high and low grayscale areas are combined into an 8-bit image I after dimming. e The process is as follows: For low grayscale regions, the normalized mapping is: Let the ordered set of values ​​be V(L), and the process of combining high and low grayscale regions is as follows: I e (x,y)={V(L)|(x,y)∈L}∪{V(H)|(x,y)∈H}。 2. The image dimming method for an infrared tracking system according to claim 1, characterized in that, In step a: the 14-bit original image I0 captured by the infrared camera is subjected to median filtering and then truncated to quantiles to I1. Let the image after median filtering be I1. median Let quantile be α, gray level be denoted as g, minimum gray level in the image be m, X be the width of the image, Y be the height of the image, and maximum gray level be M. Then: Cutoff threshold The truncated image I1(x,y) = min{N,I median (x,y)}.

3. The image dimming method for an infrared tracking system according to claim 1, characterized in that, In step b, I1 is normalized to an 8-bit image I2 and thresholded using the triangulation method to obtain a binary template image I. B The process is as follows: The normalization process is as follows: Let the triangular thresholding operator be τ and the threshold be T, then T = τ(I2); Binary template image I B For: I B (x,y)=1, ifI2(x,y)≥T; I B (x,y)=0,ifI2(x,y)<T。 4. The image dimming method for an infrared tracking system according to claim 1, characterized in that, In step c, according to I B Divide I1 into high grayscale and low grayscale regions: Let the high grayscale region be H and the low grayscale region be L, then: H={(x,y)|I B (x,y)=1}; L={(x,y)|I B (x,y)=0}。

Images