Infrared image processing method and device for realizing night vision function

Abstract

The present invention discloses an infrared image processing method and device for achieving night vision function, which relates to the technical field of image processing; acquires active infrared images and passive infrared images; performs image registration based on a preset overlapping field of view; utilizes a bilinear interpolation method to perform resolution alignment; performs denoising on the active infrared image using an improved non-local mean filtering algorithm; determines a suspected heat source area based on the passive infrared image; enhances the active infrared image based on the suspected heat source area to obtain a target fused image. The active infrared image can provide environmental terrain information, and the passive infrared image can provide heat source information of trapped persons; the heat source information is added to the active infrared image to obtain a fused image, so that search and rescue personnel can see clear terrain features through the fused image and can more easily find trapped persons, thereby freeing search and rescue personnel from the dark environment at night and greatly improving search and rescue efficiency.

Description

Technical Field

[0001] The present invention relates to the technical field of image processing, and in particular to an infrared image processing method and device for realizing night vision function. Background Art

[0002] In the wild, especially in unfamiliar environments, visibility is extremely low at night, making it easy for people to lose their sense of direction and become lost. The risk of getting lost is particularly high in complex terrain, such as forests and mountainous areas. When outdoors at night, poor visibility and uneven terrain can easily lead to accidents such as falls and injuries. Certain natural disasters, such as earthquakes, mudslides, and flash floods, can occur at night, trapping or leaving people missing. In such cases, swift nighttime search and rescue operations are necessary to save lives.

[0003] Time is the most precious resource in nighttime search and rescue efforts. As darkness deepens, the risks faced by those in distress increase, such as hypothermia, dehydration, and getting lost. Therefore, search and rescue teams must race against time to locate and rescue those in distress as quickly as possible. However, in the wilderness at night, light is almost completely absent. Limited visibility and complex conditions often slow search and rescue efforts compared to daytime operations, further exacerbating the urgency of time.

[0004] Existing technology can detect trapped people in hidden corners through infrared thermal imaging. However, search and rescue teams must grope their way through darkness, relying on limited light sources like flashlights and headlamps to illuminate the way ahead, while remaining vigilant to potential terrain obstacles and unknown dangers. This limited vision restricts rescuers' movements, significantly reducing search and rescue efficiency and significantly increasing the risks faced by trapped people. Summary of the Invention

[0005] The purpose of the present invention is to solve the problem of low search and rescue efficiency mentioned in the above background technology, and to propose an infrared image processing method and device for realizing night vision function.

[0006] A first aspect of the present invention provides an infrared image processing method for achieving night vision function, which is applied to search and rescue imaging equipment; the method comprises:

[0007] Acquire a multi-source image in the target direction; the multi-source image includes a first active infrared image and a first passive infrared image;

[0008] Performing regional cropping on the multi-source image according to a preset overlapping field of view to obtain a second active infrared image and a second passive infrared image;

[0009] performing bilinear interpolation processing on the second passive infrared image to obtain a third passive infrared image;

[0010] Denoising the second active infrared image using an improved non-local means filtering algorithm to obtain a third active infrared image;

[0011] obtaining a plurality of suspicious heat source areas according to the third passive infrared image;

[0012] The third active infrared image is enhanced according to the multiple suspicious heat source areas to obtain a target fusion image.

[0013] Optionally, the search and rescue imaging device is equipped with a near-infrared camera, a thermal radiation camera, and L1 and L5 antennas; the near-infrared camera and the thermal radiation camera are installed on the search and rescue imaging device at a fixed position and angle; the near-infrared camera is used to obtain a first active infrared image; the thermal radiation camera is used to obtain a first passive infrared image;

[0014] Before acquiring the multi-source image in the target direction, the method further includes:

[0015] Acquiring a distress signal from a trapped person, and determining first positioning information of the trapped person based on the distress signal;

[0016] Through the L1 and L5 antennas, the signal data of the satellite L1 and L5 frequency bands are obtained to determine the second positioning information of the search and rescue personnel;

[0017] A target direction of the search is determined according to the first positioning information and the second positioning information.

[0018] Optionally, performing denoising on the second active infrared image by using an improved non-local means filtering algorithm to obtain a third active infrared image includes:

[0019] Step 1: Initialize control parameters, entropy comparison value, optimal image and number of recording values;

[0020] Step 2: increasing the control parameter by a first preset step size;

[0021] Step three: perform non-local mean filtering on the second active infrared image according to the control parameters to obtain a temporary image, and calculate the entropy value of the temporary image:

[0022]

[0023] Among them, H is the entropy value to be sought; p i is the probability of a pixel with gray level i in the temporary image;

[0024] Step 4: if the entropy value is greater than the entropy comparison value, assign the entropy value to the entropy comparison value and assign the temporary image to the optimal image; otherwise, the number of times recorded is increased by 1;

[0025] Step 5: If the recorded number of times is less than the preset number of changes, return to step 2; otherwise, use the optimal image as the third active infrared image.

[0026] Optionally, obtaining a plurality of heat source areas according to the third passive infrared image includes:

[0027] Segmenting the third passive infrared image according to a first preset threshold to obtain a plurality of heat source foregrounds;

[0028] determining an area characteristic value of a target heat source foreground according to the number of pixels of the target heat source foreground; the target heat source foreground is any one of a plurality of heat source foregrounds;

[0029] If the area characteristic value is greater than a second preset threshold, the target heat source prospect is regarded as a suspicious heat source area.

[0030] Optionally, enhancing the third active infrared image according to the multiple suspicious heat source areas to obtain a target fused image includes:

[0031] Calculate the pixel values ​​of the target fused image in the multiple suspicious heat source areas based on the pixel values ​​of the third passive infrared image and the third active infrared image in the multiple suspicious heat source areas:

[0032]

[0033] Wherein, TFI(x) is the grayscale value of pixel x in the target fused image; AI(x) is the grayscale value of pixel x in the third active infrared image; PI(x) is the grayscale value of pixel x in the third passive infrared image; pixel x is constrained within multiple suspicious heat source areas; w1 and w2 are weight coefficients.

[0034] A second aspect of the present invention provides an infrared image processing device for achieving night vision, which is applied to search and rescue imaging equipment; the device comprises:

[0035] A multi-source image acquisition module is used to acquire a multi-source image in the target direction; the multi-source image includes a first active infrared image and a first passive infrared image;

[0036] An image registration module is used to perform regional cropping on the multi-source images according to a preset overlapping field of view to obtain a second active infrared image and a second passive infrared image;

[0037] a pixel alignment module, configured to perform bilinear interpolation processing on the second passive infrared image to obtain a third passive infrared image;

[0038] a filtering and denoising module, configured to perform denoising processing on the second active infrared image by using an improved non-local means filtering algorithm to obtain a third active infrared image;

[0039] a heat source information extraction module, configured to obtain a plurality of suspicious heat source areas based on the third passive infrared image;

[0040] The image fusion module is used to enhance the third active infrared image according to multiple suspicious heat source areas to obtain a target fusion image.

[0041] Optionally, the search and rescue imaging device is equipped with a near-infrared camera, a thermal radiation camera, and L1 and L5 antennas; the near-infrared camera and the thermal radiation camera are installed on the search and rescue imaging device at a fixed position and angle; the near-infrared camera is used to obtain a first active infrared image; the thermal radiation camera is used to obtain a first passive infrared image;

[0042] The device further comprises:

[0043] A distress signal acquisition module, configured to acquire a distress signal from a trapped person and determine first positioning information of the trapped person based on the distress signal;

[0044] Satellite signal acquisition module, used to obtain satellite L1 and L5 frequency band signal data through L1 and L5 antennas to determine the second positioning information of the search and rescue personnel;

[0045] The target direction determination module is used to determine the target direction of the search based on the first positioning information and the second positioning information.

[0046] Optionally, the filtering and denoising module includes:

[0047] Initialization module, used to initialize control parameters, entropy comparison value, optimal image and number of recording values;

[0048] An increasing module, configured to increase the control parameter by a first preset step size;

[0049] A denoising effect calculation module is used to perform non-local mean filtering on the second active infrared image according to the control parameters to obtain a temporary image, and calculate the entropy value of the temporary image:

[0050]

[0051] Among them, H is the entropy value to be sought; p i is the probability of a pixel with gray level i in the temporary image;

[0052] a priority maintaining module, configured to assign the entropy value to the entropy comparison value and assign the temporary image to the optimal image if the entropy value is greater than the entropy comparison value; otherwise, increment the number of times recorded by 1;

[0053] The output judgment module is used to return to the increment module if the number of recorded values ​​is less than the preset number of changes; otherwise, the optimal image is used as the third active infrared image.

[0054] Optionally, the heat source information extraction module includes:

[0055] a threshold segmentation module, configured to segment the third passive infrared image according to a first preset threshold value to obtain a plurality of heat source foregrounds;

[0056] a noise removal module, configured to determine an area characteristic value of a target heat source foreground according to the number of pixels of the target heat source foreground; the target heat source foreground is any one of a plurality of heat source foregrounds;

[0057] The heat source area determination module is used to treat the target heat source prospect as a suspicious heat source area if the area characteristic value is greater than a second preset threshold.

[0058] Optionally, the image fusion module includes:

[0059] A heat source information adding module is used to calculate the pixel values ​​of the target fused image in the multiple suspicious heat source areas based on the pixel values ​​of the third passive infrared image and the third active infrared image in the multiple suspicious heat source areas:

[0060]

[0061] Wherein, TFI(x) is the grayscale value of pixel x in the target fused image; AI(x) is the grayscale value of pixel x in the third active infrared image; PI(x) is the grayscale value of pixel x in the third passive infrared image; pixel x is constrained within multiple suspicious heat source areas; w1 and w2 are weight coefficients.

[0062] Beneficial effects of the present invention:

[0063] The present invention proposes an infrared image processing method for realizing night vision function, which is applied to search and rescue imaging equipment; the method comprises: acquiring a multi-source image in the target direction; the multi-source image comprises a first active infrared image and a first passive infrared image; performing regional cropping on the multi-source image according to a preset overlapping field of view to obtain a second active infrared image and a second passive infrared image; performing bilinear interpolation processing on the second passive infrared image to obtain a third passive infrared image; performing denoising processing on the second active infrared image by using an improved non-local mean filtering algorithm to obtain a third active infrared image; obtaining multiple suspicious heat source areas based on the third passive infrared image; and enhancing the third active infrared image based on the multiple suspicious heat source areas to obtain a target fusion image.

[0064] By acquiring active and passive infrared images in the target direction, comprehensive visual information and heat source information can be obtained. Among them, the active infrared image can provide environmental terrain information, and the passive infrared image can provide heat source information of the trapped person. The passive infrared heat source information is added to the active infrared image to obtain a fused image, so that search and rescue personnel can see clear terrain features through the fused image and can also more easily find trapped people, so that search and rescue personnel are not restricted by the dark environment at night, greatly improving the search and rescue efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0065] The present invention will be further described below with reference to the accompanying drawings.

[0066] Figure 1 A flowchart of an infrared image processing method for achieving night vision function is provided for an embodiment of the present invention;

[0067] Figure 2 An improved non-local mean filtering flow chart is provided for an embodiment of the present invention;

[0068] Figure 3 The present invention provides a structural diagram of an infrared image processing device for realizing night vision function. DETAILED DESCRIPTION

[0069] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making any creative efforts shall fall within the scope of protection of the present invention.

[0070] The embodiment of the present invention provides an infrared image processing method for realizing night vision function, which is applied to search and rescue imaging equipment. Figure 1 , Figure 1This is a flowchart of an infrared image processing method for achieving night vision function provided by an embodiment of the present invention. The method includes the following steps:

[0071] S101, acquiring multi-source images in a target direction.

[0072] S102 , performing regional cropping on the multi-source image according to a preset overlapping field of view to obtain a second active infrared image and a second passive infrared image.

[0073] S103 , performing bilinear interpolation processing on the second passive infrared image to obtain a third passive infrared image.

[0074] S104 , performing denoising processing on the second active infrared image by using an improved non-local means filtering algorithm to obtain a third active infrared image.

[0075] S105 , obtaining a plurality of suspicious heat source areas according to the third passive infrared image.

[0076] S106 , enhancing the third active infrared image according to the multiple suspicious heat source areas to obtain a target fusion image.

[0077] The multi-source image includes a first active infrared image and a first passive infrared image.

[0078] An infrared image processing method for realizing night vision function provided by an embodiment of the present invention can obtain comprehensive visual information and heat source information by acquiring active and passive infrared images in the target direction, wherein the active infrared image can provide environmental terrain information, and the passive infrared image can provide heat source information of trapped persons; the passive infrared heat source information is added to the active infrared image to obtain a fused image, so that search and rescue personnel can see clear terrain features and more easily find trapped persons through the fused image, thereby freeing search and rescue personnel from the dark environment at night and greatly improving search and rescue efficiency.

[0079] In one embodiment, the search and rescue imaging device is equipped with a near-infrared camera, a thermal radiation camera, and L1 and L5 antennas; the near-infrared camera and the thermal radiation camera are installed on the search and rescue imaging device at a fixed position and angle; the near-infrared camera is used to obtain a first active infrared image; and the thermal radiation camera is used to obtain a first passive infrared image.

[0080] Before step S101, the infrared image processing method for implementing a night vision function provided by an embodiment of the present invention further includes:

[0081] Step 1: Obtain a distress signal from a trapped person, and determine the first positioning information of the trapped person based on the distress signal.

[0082] Step 2: Obtain satellite L1 and L5 frequency band signal data through the L1 and L5 antennas to determine the second positioning information of the search and rescue personnel.

[0083] Step three: determine the target direction of the search based on the first positioning information and the second positioning information.

[0084] In one implementation, when the search and rescue imaging equipment is installed, the near-infrared camera and the thermal radiation camera are fixed at a preset angle and position, which can keep the overlapping area of ​​the field of view of multiple cameras unchanged. This makes it easier to crop and align the image area in subsequent processing, reducing alignment time and improving image fusion efficiency.

[0085] In one implementation, the optimal search direction or path can be quickly calculated by combining the location information of the trapped person and the rescue personnel. This not only shortens the search and rescue time, but also reduces the risk of rescue personnel getting lost in complex environments, reduces ineffective searches, and improves search and rescue efficiency.

[0086] In one embodiment, step S104 includes:

[0087] Step 1: Initialize the control parameters, entropy comparison value, optimal image and number of recording values.

[0088] Step 2: Increase the control parameter by a first preset step size.

[0089] Step 3: Perform non-local mean filtering on the second active infrared image according to the control parameters to obtain a temporary image, and calculate the entropy value of the temporary image:

[0090]

[0091] Among them, H is the entropy value to be sought; p i is the probability of a pixel with gray level i in the temporary image.

[0092] Step 4: If the entropy value is greater than the entropy comparison value, the entropy value is assigned to the entropy comparison value and the temporary image is assigned to the optimal image; otherwise, the number of times recorded is increased by 1.

[0093] Step 5: If the recorded number of times is less than the preset number of changes, return to step 2; otherwise, use the optimal image as the third active infrared image.

[0094] Among them, p i is the probability of a pixel with gray level i in the temporary image, which is characterized by the frequency of gray level i appearing in the temporary image:

[0095]

[0096] Wherein, num(TI, i) represents the number of pixels in the temporary image TI whose grayscale value is equal to i; N is the total number of pixels in the temporary image TI.

[0097] In one implementation, performing non-local mean filtering on the second active infrared image according to the control parameter to obtain a temporary image includes:

[0098]

[0099] Where I1(x) is the pixel value of pixel x in the temporary image; I0(y) is the pixel value of pixel y in the second active infrared image; Ω is the local window in the second active infrared image; w(x,y) is the similarity weighting coefficient between pixel x and pixel y; P(x) and P(y) are the pixel intensity vectors of the local windows of pixels x and y in the second active infrared image, respectively; ‖P(x)-P(y)‖ 2 represents the square of the Euclidean distance between P(x) and P(y); σ is the control parameter; C(x,y) is the normalization coefficient.

[0100] A larger control parameter increases the noise reduction intensity, resulting in a more blurred image. A smaller control parameter decreases the noise reduction intensity, preserving more edge information but also retaining more noise points. Therefore, the size of the control parameter should vary depending on the noise level in different images. This method uses entropy as a metric for image quality. By comparing it with a preset entropy comparison value, the temporary image with the higher entropy value is selected as the optimal image. This method can automatically select the optimal image processing result, improving processing efficiency and consistency.

[0101] In one implementation, the first preset step size can be set to 0.01; and the number of changes can be set to 7. Figure 2 , Figure 2 This is a flowchart of an improved non-local means filtering method provided by an embodiment of the present invention. Here, σ is a control parameter, r is the number of recorded times, E is the entropy comparison value, OPI is the optimal image, TI is the temporary image, f(σ, SAI) represents the temporary image obtained based on the control parameter σ and the second active infrared image SAI, H is the entropy value, and g(TI) represents the entropy value calculated for the temporary image TI.

[0102] In one embodiment, step S105 includes:

[0103] Step 1: Segment the third passive infrared image according to a first preset threshold to obtain multiple heat source foregrounds.

[0104] Step 2: Determine the area characteristic value of the target heat source foreground according to the number of pixels of the target heat source foreground.

[0105] Step three: If the area characteristic value is greater than the second preset threshold, the target heat source prospect is regarded as a suspicious heat source area.

[0106] The target heat source prospect is any one of the multiple heat source prospects.

[0107] In one implementation, the third passive infrared image is segmented using a first preset threshold to quickly locate possible heat source areas, reducing processing complexity and computational effort. Suspicious heat source areas are further screened based on the number of pixels in the heat source foreground (i.e., area eigenvalues), removing noise and ensuring the accuracy and reliability of the detection results.

[0108] In one embodiment, step S106 includes:

[0109] Calculate the pixel values ​​of the target fused image in the multiple suspicious heat source areas based on the pixel values ​​of the third passive infrared image and the third active infrared image in the multiple suspicious heat source areas:

[0110]

[0111] Where TFI(x) is the grayscale value of pixel x in the target fusion image; AI(x) is the grayscale value of pixel x in the third active infrared image; PI(x) is the grayscale value of pixel x in the third passive infrared image; pixel x is constrained within multiple suspicious heat source regions; w1 and w2 are weight coefficients, w1 + w2 = 1.

[0112] In one implementation, active infrared images have high resolution and contrast, clearly displaying environmental details. Passive infrared images, on the other hand, capture the target's own infrared radiation, providing superior detection capabilities for hidden targets. By fusing these two different types of infrared images, the target fusion image combines the advantages of both. By performing pixel value fusion calculations only within the suspected heat source area, the background noise in the passive infrared image can be prevented from interfering with the environmental information in the active infrared image, while also enhancing the visibility of the heat source location. This allows rescuers to see the details of the surrounding terrain, unimpeded by visual lags and improving search and rescue efficiency. It also makes it easier to distinguish whether the heat source outline is a person, enabling the detection of trapped individuals in concealed locations.

[0113] In one implementation, w1 can be 0.6; w2 can be 0.4.

[0114] The embodiment of the present invention provides an infrared image processing device for realizing night vision function, which is applied to search and rescue imaging equipment. Figure 3 , Figure 3 This is a structural diagram of an infrared image processing device for implementing night vision function provided by an embodiment of the present invention. The device includes:

[0115] The multi-source image acquisition module is used to acquire multi-source images in the target direction.

[0116] The image registration module is used to perform regional cropping on the multi-source images according to a preset overlapping field of view to obtain a second active infrared image and a second passive infrared image.

[0117] The pixel alignment module is used to perform bilinear interpolation processing on the second passive infrared image to obtain a third passive infrared image.

[0118] The filtering and denoising module is used to perform denoising processing on the second active infrared image by using an improved non-local mean filtering algorithm to obtain a third active infrared image.

[0119] The heat source information extraction module is used to obtain multiple suspicious heat source areas based on the third passive infrared image.

[0120] The image fusion module is used to enhance the third active infrared image according to multiple suspicious heat source areas to obtain a target fusion image.

[0121] The multi-source image includes a first active infrared image and a first passive infrared image.

[0122] An infrared image processing device for realizing night vision function provided by an embodiment of the present invention can obtain comprehensive visual information and heat source information by acquiring active and passive infrared images in the target direction, wherein the active infrared image can provide environmental terrain information, and the passive infrared image can provide heat source information of trapped persons; the passive infrared heat source information is added to the active infrared image to obtain a fused image, so that search and rescue personnel can see clear terrain features and more easily find trapped persons through the fused image, thereby freeing search and rescue personnel from the dark environment at night and greatly improving search and rescue efficiency.

[0123] In one embodiment, the search and rescue imaging device is equipped with a near-infrared camera, a thermal radiation camera, and L1 and L5 antennas; the near-infrared camera and the thermal radiation camera are installed on the search and rescue imaging device at a fixed position and angle; the near-infrared camera is used to obtain a first active infrared image; and the thermal radiation camera is used to obtain a first passive infrared image.

[0124] An infrared image processing device for implementing a night vision function provided by an embodiment of the present invention further includes:

[0125] The distress signal acquisition module is used to acquire the distress signal of the trapped person and determine the first positioning information of the trapped person according to the distress signal.

[0126] The satellite signal acquisition module is used to obtain the signal data of the satellite L1 and L5 frequency bands through the L1 and L5 antennas to determine the second positioning information of the search and rescue personnel.

[0127] The target direction determination module is used to determine the target direction of the search based on the first positioning information and the second positioning information.

[0128] In one embodiment, the filtering and denoising module includes:

[0129] The initialization module is used to initialize the control parameters, entropy comparison value, optimal image and number of recording values.

[0130] The increasing module is used to increase the control parameter by a first preset step size.

[0131] The denoising effect calculation module is used to perform non-local mean filtering on the second active infrared image according to the control parameters to obtain a temporary image and calculate the entropy value of the temporary image:

[0132]

[0133] Among them, H is the entropy value to be sought; p i is the probability of a pixel with gray level i in the temporary image.

[0134] The optimal image preserving module is used to assign the entropy value to the entropy comparison value and the temporary image to the optimal image if the entropy value is greater than the entropy comparison value; otherwise, the number of recorded values ​​is increased by 1.

[0135] The output judgment module is used to return to the increment module if the number of recorded times is less than the preset number of changes; otherwise, the optimal image is used as the third active infrared image.

[0136] In one embodiment, the heat source information extraction module includes:

[0137] The threshold segmentation module is used to segment the third passive infrared image according to the first preset threshold to obtain multiple heat source foregrounds.

[0138] The noise removal module is used to determine the area characteristic value of the target heat source foreground according to the number of pixels of the target heat source foreground.

[0139] The heat source area determination module is used to treat the target heat source prospect as a suspicious heat source area if the area characteristic value is greater than a second preset threshold.

[0140] The target heat source prospect is any one of the multiple heat source prospects.

[0141] In one embodiment, the image fusion module includes:

[0142] The heat source information adding module is used to calculate the pixel values ​​of the target fusion image in multiple suspicious heat source areas based on the pixel values ​​of the third passive infrared image and the third active infrared image in multiple suspicious heat source areas:

[0143]

[0144] Where TFI(x) is the grayscale value of pixel x in the target fusion image; AI(x) is the grayscale value of pixel x in the third active infrared image; PI(x) is the grayscale value of pixel x in the third passive infrared image; pixel x is constrained within multiple suspicious heat source regions; w1 and w2 are weight coefficients.

[0145] The above is a detailed description of an embodiment of the present invention. However, the content described is only a preferred embodiment of the present invention and should not be considered to limit the scope of the present invention. All equivalent changes and improvements made within the scope of the present invention should still fall within the scope of the patent coverage of the present invention.

Claims

1. An infrared image processing method for realizing night vision function, characterized in that: Applied to search and rescue imaging equipment; the method comprises: Acquire a multi-source image in the target direction; the multi-source image includes a first active infrared image and a first passive infrared image; Performing regional cropping on the multi-source image according to a preset overlapping field of view to obtain a second active infrared image and a second passive infrared image; performing bilinear interpolation processing on the second passive infrared image to obtain a third passive infrared image; Denoising the second active infrared image using an improved non-local means filtering algorithm to obtain a third active infrared image; obtaining a plurality of suspicious heat source areas according to the third passive infrared image; enhancing the third active infrared image according to the multiple suspected heat source areas to obtain a target fusion image; The step of performing denoising on the second active infrared image by using an improved non-local means filtering algorithm to obtain a third active infrared image includes: Step 1: Initialize control parameters, entropy comparison value, optimal image and number of recording values; Step 2: increasing the control parameter by a first preset step size; Step three: perform non-local mean filtering on the second active infrared image according to the control parameters to obtain a temporary image, and calculate the entropy value of the temporary image: Among them, H is the entropy value to be sought; p i is the probability of a pixel with gray level i in the temporary image; Step 4: if the entropy value is greater than the entropy comparison value, assign the entropy value to the entropy comparison value and assign the temporary image to the optimal image; otherwise, the number of times recorded is increased by 1; Step 5: If the recorded number of times is less than the preset number of changes, return to step 2; otherwise, use the optimal image as the third active infrared image.

2. The infrared image processing method for realizing night vision function according to claim 1, characterized in that: The search and rescue imaging device is equipped with a near-infrared camera, a thermal radiation camera, and L1 and L5 antennas; the near-infrared camera and the thermal radiation camera are installed on the search and rescue imaging device at a fixed position and angle; the near-infrared camera is used to obtain a first active infrared image; The thermal radiation camera is used to obtain a first passive infrared image; Before acquiring the multi-source image in the target direction, the method further includes: Acquiring a distress signal from a trapped person, and determining first positioning information of the trapped person based on the distress signal; Through the L1 and L5 antennas, the signal data of the satellite L1 and L5 frequency bands are obtained to determine the second positioning information of the search and rescue personnel; A target direction of the search is determined according to the first positioning information and the second positioning information.

3. The infrared image processing method for realizing night vision function according to claim 1, characterized in that: The obtaining of a plurality of heat source areas according to the third passive infrared image comprises: Segmenting the third passive infrared image according to a first preset threshold to obtain a plurality of heat source foregrounds; determining an area characteristic value of a target heat source foreground according to the number of pixels of the target heat source foreground; the target heat source foreground is any one of a plurality of heat source foregrounds; If the area characteristic value is greater than a second preset threshold, the target heat source prospect is regarded as a suspicious heat source area.

4. The infrared image processing method for realizing night vision function according to claim 1, characterized in that: The step of enhancing the third active infrared image according to the plurality of suspicious heat source areas to obtain a target fused image includes: Calculate the pixel values ​​of the target fused image in the multiple suspicious heat source areas based on the pixel values ​​of the third passive infrared image and the third active infrared image in the multiple suspicious heat source areas: TFI(x)=w1*AI(x)+w2*PI(x) Wherein, TFI(x) is the grayscale value of pixel x in the target fused image; AI(x) is the grayscale value of pixel x in the third active infrared image; PI(x) is the grayscale value of pixel x in the third passive infrared image; pixel x is constrained within multiple suspicious heat source areas; w1 and w2 are weight coefficients.

5. An infrared image processing device for realizing night vision function, characterized in that: Applicable to search and rescue imaging equipment; the device comprises: A multi-source image acquisition module is used to acquire a multi-source image in the target direction; the multi-source image includes a first active infrared image and a first passive infrared image; An image registration module is used to perform regional cropping on the multi-source images according to a preset overlapping field of view to obtain a second active infrared image and a second passive infrared image; a pixel alignment module, configured to perform bilinear interpolation processing on the second passive infrared image to obtain a third passive infrared image; a filtering and denoising module, configured to perform denoising processing on the second active infrared image by using an improved non-local means filtering algorithm to obtain a third active infrared image; a heat source information extraction module, configured to obtain a plurality of suspicious heat source areas based on the third passive infrared image; an image fusion module, configured to enhance the third active infrared image according to the plurality of suspicious heat source areas to obtain a target fusion image; The filtering and denoising module includes: Initialization module, used to initialize control parameters, entropy comparison value, optimal image and number of recording values; An increasing module, configured to increase the control parameter by a first preset step size; A denoising effect calculation module is used to perform non-local mean filtering on the second active infrared image according to the control parameters to obtain a temporary image, and calculate the entropy value of the temporary image: Among them, H is the entropy value to be sought; p i is the probability of a pixel with gray level i in the temporary image; a priority maintaining module, configured to assign the entropy value to the entropy comparison value and assign the temporary image to the optimal image if the entropy value is greater than the entropy comparison value; otherwise, increment the number of times recorded by 1; The output judgment module is used to return to the increment module if the number of recorded values ​​is less than the preset number of changes; otherwise, the optimal image is used as the third active infrared image.

6. The infrared image processing device for realizing night vision function according to claim 5, characterized in that: The search and rescue imaging device is equipped with a near-infrared camera, a thermal radiation camera, and L1 and L5 antennas; the near-infrared camera and the thermal radiation camera are installed on the search and rescue imaging device at a fixed position and angle; the near-infrared camera is used to obtain a first active infrared image; The thermal radiation camera is used to obtain a first passive infrared image; The device further comprises: A distress signal acquisition module, configured to acquire a distress signal from a trapped person and determine first positioning information of the trapped person based on the distress signal; Satellite signal acquisition module, used to obtain satellite L1 and L5 frequency band signal data through L1 and L5 antennas to determine the second positioning information of the search and rescue personnel; The target direction determination module is used to determine the target direction of the search based on the first positioning information and the second positioning information.

7. The infrared image processing device for realizing night vision function according to claim 5, characterized in that: The heat source information extraction module includes: a threshold segmentation module, configured to segment the third passive infrared image according to a first preset threshold value to obtain a plurality of heat source foregrounds; a noise removal module, configured to determine an area characteristic value of a target heat source foreground according to the number of pixels of the target heat source foreground; the target heat source foreground is any one of a plurality of heat source foregrounds; The heat source area determination module is used to treat the target heat source prospect as a suspicious heat source area if the area characteristic value is greater than a second preset threshold.

8. The infrared image processing device for realizing night vision function according to claim 5, characterized in that: The image fusion module includes: A heat source information adding module is used to calculate the pixel values ​​of the target fused image in the multiple suspicious heat source areas based on the pixel values ​​of the third passive infrared image and the third active infrared image in the multiple suspicious heat source areas: TFI(x)=w1*AI(x)+w2*PI(x) Wherein, TFI(x) is the grayscale value of pixel x in the target fused image; AI(x) is the grayscale value of pixel x in the third active infrared image; PI(x) is the grayscale value of pixel x in the third passive infrared image; pixel x is constrained within multiple suspicious heat source areas; w1 and w2 are weight coefficients.

Images