An infrared and visible light image fusion method for extra-high voltage converter transformer bushing

Abstract

The embodiment of the present application provides a kind of infrared and visible light image fusion method of extra-high voltage converter transformer bushing, belong to image processing field.The method comprises: obtaining infrared light image and visible light image of the same bushing;The infrared light image and the visible light image are preprocessed;The infrared light image and the visible light image after pre-processing are carried out image segmentation;The infrared light image and visible light image after segmentation are carried out refinement processing, to form corresponding weight map;The infrared light image and the visible light image after pre-processing are carried out brightness extraction;The infrared light image and the visible light image after pre-processing are carried out single-scale weighting with corresponding weight map, to form pre-fusion image;The pre-fusion image and the brightness layer after brightness extraction are added to obtain the final fusion image.The method can effectively fuse infrared light and visible light image.

Description

Technical Field

[0001] This invention relates to the field of image processing, and more specifically to a method for fusing infrared and visible light images of ultra-high voltage converter transformer bushings. Background Technology

[0002] Converter transformers are core equipment in DC transmission projects, and their safe operation is crucial for maintaining the normal operation of the high-voltage DC transmission system. Due to the large capacity, high withstand voltage levels, and complex internal electric field distribution of ultra-high-voltage converter transformers, the requirements for insulation strength are extremely high. Bushings, as an important component of the converter transformer, are responsible for fixing the leads and ensuring their insulation from the outside environment; they are also frequently prone to failure. In actual operation, bushings often exhibit localized abnormal heating due to defects such as insulation damage, oil leakage, and poor joint contact, affecting their service life and the safety of power operation. Currently, the operating status of ultra-high-voltage converter transformer bushings is mainly monitored using visible light and infrared images, and inspections are conducted manually. How to achieve intelligent monitoring of the operating status of ultra-high-voltage converter transformer bushings based on visible light and infrared image information is an urgent problem to be solved. Summary of the Invention

[0003] The purpose of this invention is to provide a method for fusing infrared and visible light images of ultra-high voltage converter transformer bushings, which can effectively fuse infrared and visible light images.

[0004] To achieve the above objectives, embodiments of the present invention provide a method for fusing infrared and visible light images of ultra-high voltage converter transformer bushings, the method comprising:

[0005] Acquire infrared and visible light images of the same tube;

[0006] The infrared image and the visible light image are preprocessed;

[0007] Image segmentation is performed on the preprocessed infrared image and the visible light image;

[0008] The segmented infrared and visible light images are refined to form corresponding weight maps;

[0009] Brightness extraction is performed on the preprocessed infrared and visible light images;

[0010] The preprocessed infrared and visible light images are weighted by a single scale with the corresponding weight map to form a pre-fused image;

[0011] The pre-fused image is added to the luminance layer after luminance extraction to obtain the final fused image.

[0012] Optionally, preprocessing the infrared image and the visible light image includes:

[0013] Acquire the infrared image and the visible light image;

[0014] The enhanced infrared image is obtained using formula (1):

[0015] E IR =I IR +2*(I IR -GFLS(I IR )), formula (1)

[0016] Among them, E IR Indicates the enhanced infrared image, I IR This represents the initial infrared image, and GFLS represents the GF-LS model operator.

[0017] The enhanced visible light image is obtained using formula (2):

[0018] E VI =I VI +2*(I VI -GFLS(I VI )), formula (2)

[0019] Among them, E IR I represents the enhanced visible light image. VI This represents the initial visible light image;

[0020] Acquire the enhanced infrared and visible light images;

[0021] The enhanced infrared image is further optimized using formula (3):

[0022]

[0023] Among them, u IR For the optimized infrared image, F() and F -1 () denotes the Fourier transform pair, F(1) denotes the delta function of the fast Fourier transform, δ * This indicates a forward difference operator or a backward difference operator. This indicates that a guided filter is used to guide the gradients of the x and y axes of the enhanced infrared image;

[0024] The enhanced visible light image is further optimized using formula (4):

[0025]

[0026] Among them, u VIFor the optimized visible light image, This indicates that a guided filter is used to guide the gradients of the x and y axes of the enhanced visible light image.

[0027] Optionally, image segmentation of the preprocessed infrared image and the visible light image includes:

[0028] Obtain the optimized infrared and visible light images;

[0029] The infrared image is segmented using formulas (5) and (6):

[0030]

[0031]

[0032] Among them, I th1 (x,y) is the infrared light segmentation image, (a1,b1) are variables in the infrared light image, h(i)1 represents the infrared light image histogram, (x,y) are independent variables, and th1 is the infrared light image threshold.

[0033] The appropriate infrared threshold is obtained according to formula (7):

[0034]

[0035] Wherein, the f cross (th1) is the infrared cross-entropy function, and L1 is the gray level of the infrared image;

[0036] Segmenting visible light images using formulas (8) and (9):

[0037]

[0038]

[0039] Among them, I th2 (x,y) is the visible light segmentation image, (a2,b2) are variables in the visible light image, h(i)2 represents the visible light image histogram, and th2 is the visible light image threshold;

[0040] The appropriate visible light threshold is obtained according to formula (10):

[0041]

[0042] Among them, f cross (th2) is the visible light cross-entropy function, and L2 is the gray level of the visible light image.

[0043] Optionally, the segmented infrared and visible light images can be refined to form corresponding weighted images;

[0044] Acquire the segmented infrared light image;

[0045] The segmented infrared image is processed using the MCET-HHO method, and the infrared image weight map is obtained according to formula (11):

[0046] W IR =MH(I IR ), formula (11)

[0047] Among them, W IR This represents the infrared image weight map, and MH represents the MCET-HHO operator;

[0048] Obtain the infrared image weight map;

[0049] The refined infrared image weight map is obtained according to formula (12):

[0050] R IR =GFLS(W IR ), formula (12)

[0051] Among them, R IR This represents the weighted image of the refined infrared light image, and GFLS represents the GF-LS model operator.

[0052] Optionally, the segmented infrared and visible light images are refined to form corresponding weighted maps, including:

[0053] Obtain the segmented visible light image;

[0054] The segmented visible light image is processed using the MCET-HHO method, and the visible light image weight map is obtained according to formula (13):

[0055] W VI =max(I IR )-W IR , formula (13)

[0056] Among them, W VI Represents the weight map of a visible light image;

[0057] Obtain the visible light image weight map;

[0058] The refined visible light image weight map is obtained according to formula (14):

[0059] R VI =GFLS(W VI ), formula (14)

[0060] Among them, R VI This represents the weighted image of the refined visible light image.

[0061] Optionally, brightness extraction of the preprocessed infrared and visible light images includes:

[0062] Median filtering is applied to the preprocessed infrared and visible light images;

[0063] Substitute the infrared and visible light images after median filtering into formula (15) to obtain the difference image between the infrared and visible light images:

[0064] F n =|F med -F smooth |, formula (15)

[0065] Among them, F n For the difference image, F med F represents the image after median filtering. smooth This represents the image after processing using the GF-LS model;

[0066] Substitute the difference image between the infrared image and the visible light image into formula (16) to obtain the enhanced infrared brightness layer and the visible light brightness layer:

[0067] S n =F n +2*(F n -GFLS(F n )), formula (16)

[0068] Among them, S n This indicates the enhanced brightness layer;

[0069] Substitute the enhanced infrared brightness layer and visible light brightness layer into formula (17) to obtain the final brightness layer:

[0070] S f =max{S n}, formula (17)

[0071] Among them, S f This represents the final brightness layer.

[0072] Optionally, the preprocessed infrared and visible light images are weighted at a single scale with their corresponding weight maps to form a pre-fused image, including:

[0073] Acquire the preprocessed infrared and visible light images;

[0074] Obtain the refined infrared image weight map and visible light image weight map;

[0075] Obtain the pre-fused image according to formula (18):

[0076] P f =R IR *u IR +R VI *u VI Formula (18)

[0077] Among them, P f This represents the pre-fused image.

[0078] Optionally, adding the pre-fused image to the luminance layer after luminance extraction to obtain the final fused image includes:

[0079] Acquire the pre-fused image;

[0080] Obtain the final brightness layer;

[0081] The final fused image is obtained according to formula (19):

[0082] F = P f +S f , formula (19)

[0083] Where F represents the final fused image.

[0084] The present invention provides a method for fusing infrared and visible light images of ultra-high voltage converter transformer bushings. This method acquires infrared and visible light images of the same bushing, then preprocesses these images to make them usable for subsequent operations. The segmented infrared and visible light images are refined to form corresponding weight maps. Brightness extraction can be performed on the preprocessed infrared and visible light images. A pre-fused image is then formed by single-scale weighting of the preprocessed infrared and visible light images with the corresponding weight maps. Finally, the pre-fused image is added to the brightness layer to obtain the final fused image.

[0085] Other features and advantages of the embodiments of the present invention will be described in detail in the following detailed description section. Attached Figure Description

[0086] The accompanying drawings are provided to further illustrate embodiments of the present invention and form part of the specification. They are used together with the following detailed description to explain the embodiments of the present invention, but do not constitute a limitation thereof. In the drawings:

[0087] Figure 1This is a flowchart of a method for fusing infrared and visible light images of an ultra-high voltage converter transformer bushing according to an embodiment of the present invention;

[0088] Figure 2 This is a flowchart of the preprocessing method for fusing infrared and visible light images of an ultra-high voltage converter transformer bushing according to an embodiment of the present invention.

[0089] Figure 3 This is a flowchart of the segmentation of infrared and visible light images in a method for fusing infrared and visible light images of an ultra-high voltage converter transformer bushing according to an embodiment of the present invention.

[0090] Figure 4 This is a flowchart illustrating the formation of a weighted graph using an infrared and visible light image fusion method for an ultra-high voltage converter transformer bushing, according to an embodiment of the present invention.

[0091] Figure 5 This is a flowchart illustrating the formation of a weighted graph using an infrared and visible light image fusion method for an ultra-high voltage converter transformer bushing, according to an embodiment of the present invention.

[0092] Figure 6 This is a flowchart of a brightness extraction method for infrared and visible light image fusion of ultra-high voltage converter transformer bushings according to an embodiment of the present invention;

[0093] Figure 7 This is a flowchart illustrating the formation of a pre-fused image using an infrared and visible light image fusion method for an ultra-high voltage converter transformer bushing according to an embodiment of the present invention.

[0094] Figure 8 This is a flowchart illustrating the formation of the final fused image using an infrared and visible light image fusion method for an ultra-high voltage converter transformer bushing according to an embodiment of the present invention. Detailed Implementation

[0095] The specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. It should be understood that the specific embodiments described herein are for illustration and explanation only and are not intended to limit the scope of the present invention.

[0096] Figure 1 This is a flowchart illustrating a method for fusing infrared and visible light images of an ultra-high voltage converter transformer bushing according to an embodiment of the present invention. In this invention, the process of fusing infrared and visible light images may include:

[0097] In step S1, infrared and visible light images of the same tube are acquired.

[0098] In step S2, the infrared and visible light images are preprocessed.

[0099] In step S3, image segmentation is performed on the preprocessed infrared and visible light images.

[0100] In step S4, the segmented infrared and visible light images are refined to form corresponding weighted maps.

[0101] In step S5, brightness is extracted from the preprocessed infrared and visible light images.

[0102] In step S6, the preprocessed infrared and visible light images are weighted by a single scale with the corresponding weight map to form a pre-fused image.

[0103] In step S7, the pre-fused image is added to the brightness layer after brightness extraction to obtain the final fused image.

[0104] In this invention, after acquiring infrared and visible light images from the same tube, these images can be preprocessed to meet subsequent processing requirements. After preprocessing, image segmentation can be performed to optimize the infrared and visible light images. Following segmentation, the segmented images can be refined to smooth their edges, creating a corresponding weight map. Brightness extraction can be performed on the preprocessed images to address brightness loss during fusion. A pre-fused image is formed by single-scale weighting of the preprocessed images and their corresponding weight maps. Finally, the pre-fused image is added to the extracted brightness layer to obtain the final fused image.

[0105] In one embodiment of the present invention, such as Figure 2 As shown, the preprocessing steps for infrared and visible light images may include:

[0106] In step S8, infrared light images and visible light images are acquired.

[0107] In step S9, the enhanced infrared image is obtained using formula (1).

[0108] E IR =I IR +2*(I IR-GFLS(I IR )), formula (1)

[0109] Among them, E IR Indicates the enhanced infrared image, I IR This represents the initial infrared image, and GFLS represents the GF-LS model operator.

[0110] In step S10, the enhanced visible light image is obtained using formula (2).

[0111] E VI =I VI +2*(I VI -GFLS(I VI )), formula (2)

[0112] Among them, E IR I represents the enhanced visible light image. VI This represents the initial visible light image.

[0113] In step S11, the enhanced infrared light image and visible light image are acquired.

[0114] In step S12, the enhanced infrared image is further optimized using formula (3):

[0115]

[0116] Among them, u IR For the optimized infrared image, F() and F -1 () denotes the Fourier transform pair, F(1) denotes the delta function of the fast Fourier transform, δ * This indicates a forward difference operator or a backward difference operator. This indicates that a guided filter is used to guide the gradients of the x-axis and y-axis of the enhanced infrared image.

[0117] In step S13, the enhanced visible light image is further optimized using formula (4):

[0118]

[0119] Among them, u VI For the optimized visible light image, This indicates that a guided filter is used to guide the gradients of the x and y axes of the enhanced visible light image.

[0120] In this invention, when preprocessing infrared and visible light images, it is necessary to acquire both infrared and visible light images. The acquired infrared image can then be enhanced using formula (1) to recover more details. The acquired visible light image can then be enhanced using formula (2). After enhancing the infrared and visible light images, the enhanced infrared image can be further optimized using formula (3) to ensure smooth edges. The enhanced visible light image can then be further optimized using formula (4) to ensure smooth edges.

[0121] In one embodiment of the present invention, such as Figure 3 As shown, the process of image segmentation for preprocessed infrared and visible light images may include:

[0122] In step S14, the optimized infrared light image and visible light image are acquired.

[0123] In step S15, the infrared image is segmented using formulas (5) and (6):

[0124]

[0125]

[0126] Among them, I th1 (x,y) represents the infrared light segmentation image, (a1,b1) represents the variables in the infrared light image, h(i)1 represents the infrared light image histogram, (x,y) represents the independent variable, and th1 represents the infrared light image threshold.

[0127] In step S16, a suitable infrared light threshold is obtained according to formula (7):

[0128]

[0129] Among them, f cross (th1) is the infrared cross-entropy function, and L1 is the gray level of the infrared image.

[0130] In step S17, the visible light image is segmented using formulas (8) and (9):

[0131]

[0132]

[0133] Among them, I th2(x,y) represents the visible light segmentation image, (a2,b2) represents the variables in the visible light image, h(i)2 represents the visible light image histogram, and th2 is the visible light image threshold.

[0134] In step S18, a suitable visible light threshold is obtained according to formula (10):

[0135]

[0136] Among them, f cross (th2) is the visible light cross-entropy function, and L2 is the gray level of the visible light image.

[0137] In this invention, image segmentation is a crucial stage in image processing. Compared to the classic two-level thresholding, the multi-level thresholding method can use more thresholds to represent different features in the image, thus making it more efficient in image segmentation. Before acquiring and segmenting the optimized infrared and visible light images, this invention can crop the infrared and visible light images so that their sizes correspond when finally matched. After acquiring the optimized infrared and visible light images, the infrared image can be segmented using formulas (5) and (6) to make it more accurate. When segmenting the infrared image, the most suitable infrared threshold for segmenting the infrared image can be obtained using formula (7). The optimized visible light image can be segmented using formulas (8) and (9). When segmenting the visible light image, the suitable visible light threshold for segmenting the visible light image can be obtained using formula (10).

[0138] In one embodiment of the present invention, such as Figure 4 As shown, the process of refining the segmented infrared and visible light images can include:

[0139] In step S19, the segmented infrared light image is acquired.

[0140] In step S20, the segmented infrared image is processed using the MCET-HHO method, and the infrared image weight map is obtained according to formula (11):

[0141] W IR =MH(I IR ), formula (11)

[0142] Among them, W IR This represents the infrared image weight map, and MH represents the MCET-HHO operator.

[0143] In step S21, the infrared light image weight map is obtained.

[0144] In step S22, the refined infrared image weight map is obtained according to formula (12):

[0145] R IR =GFLS(W IR ), formula (12)

[0146] Among them, R IR This represents the weighted image of the refined infrared light image, and GFLS represents the GF-LS model operator.

[0147] In this invention, when refining the segmented infrared light image, the segmented infrared light image can be obtained, and then processed using the MCET-HHO method to avoid the large amount of noise caused by the coarse weight map generated during the segmentation of the infrared light image, which would result in a large number of artifacts in the final fused image. Therefore, after obtaining the infrared light image weight map according to formula (11), the infrared light image weight map can be refined using formula (12) to keep the edges of the infrared light image smooth and avoid generating a large number of artifacts in the final fused image.

[0148] In one embodiment of the present invention, such as Figure 5 As shown, the process of thinning the segmented visible light image may include:

[0149] In step S23, the segmented visible light image is acquired.

[0150] In step S24, the segmented visible light image is processed by the MCET-HHO method, and the visible light image weight map is obtained according to formula (13).

[0151] W VI =max(I IR )-W IR , formula (13)

[0152] Among them, W VI Represents the weight map of a visible light image;

[0153] In step S25, a visible light image weight map is obtained.

[0154] In step S26, the refined visible light image weight map is obtained according to formula (14):

[0155] R VI =GFLS(W VI ), formula (14)

[0156] Among them, R VI This represents the weighted image of the refined visible light image.

[0157] In this invention, when refining the segmented visible light image, the segmented visible light image can be obtained, and then processed using the MCET-HHO method to avoid the large amount of noise caused by the coarse weight map generated during the segmentation of the visible light image, which would result in a large number of artifacts in the final fused image. Therefore, after obtaining the visible light image weight map according to formula (13), the visible light image weight map can be refined using formula (14) to keep the edges of the visible light image smooth and avoid generating a large number of artifacts in the final fused image.

[0158] In one embodiment of the present invention, such as Figure 6 As shown, the process for extracting brightness from preprocessed infrared and visible light images can include:

[0159] In step S27, median filtering is performed on the preprocessed infrared and visible light images.

[0160] In step S28, the infrared image and the visible image after median filtering are substituted into formula (15) to obtain the difference image between the infrared image and the visible image:

[0161] F n =|F med -F smooth |, formula (15)

[0162] Among them, F n For the difference image, F med F represents the image after median filtering. smooth This represents the image after passing through the GF-LS model.

[0163] In step S29, the difference between the infrared light image and the visible light image is substituted into formula (16) to obtain the enhanced infrared light brightness layer and the visible light brightness layer.

[0164] S n =F n +2*(F n -GFLS(F n )), formula (16)

[0165] Among them, S n This indicates the enhanced brightness layer;

[0166] In step S30, the enhanced infrared brightness layer and visible light brightness layer are substituted into formula (17) to obtain the final brightness layer:

[0167] S f =max{S n}, formula (17)

[0168] Among them, S f This represents the final brightness layer.

[0169] In this invention, some information from the source image is lost during image fusion. This brightness extraction method can solve the problem of information loss. Median filtering is applied to the preprocessed infrared and visible light images. Then, the difference image can be obtained according to formula (15). The difference image can be the difference between the infrared and visible light images after median filtering and the infrared and visible light images after passing through the GF-LS model. Then, the enhanced infrared brightness layer and visible light brightness layer can be obtained through formula (16). Finally, the final brightness layer can be obtained through formula (17).

[0170] In one embodiment of the present invention, such as Figure 7 As shown, the process of performing single-scale weighting of the preprocessed infrared and visible light images with their corresponding weight maps to form a pre-fused image may include:

[0171] In step S31, the preprocessed infrared light image and visible light image are acquired.

[0172] In step S32, the refined infrared image weight map and visible light image weight map are obtained.

[0173] In step S33, the pre-fused image is obtained according to formula (18):

[0174] P f =R IR *u IR +R VI *u VI Formula (18)

[0175] Among them, P f This represents the pre-fused image.

[0176] In this invention, the single-scale fusion rule can better reduce information loss. Therefore, the preprocessed infrared light image and visible light image can be multiplied with the corresponding weight map to generate a pre-fused image. When generating the pre-fused image, the preprocessed infrared light image and visible light image can be obtained, and then the refined infrared light image weight map and visible light image weight map can be obtained. Thus, the pre-fused image can be obtained according to formula (18), which can effectively fuse the infrared light image and visible light image.

[0177] In one embodiment of the present invention, such as Figure 8 As shown, combining the pre-fused image with the brightness layer after brightness extraction to obtain the final fused image can include:

[0178] In step S34, a pre-fused image is obtained.

[0179] In step S35, the final brightness layer is obtained.

[0180] In step S36, the final fused image is obtained according to formula (19):

[0181] F = P f +S f , formula (19)

[0182] Where F represents the final fused image.

[0183] In this invention, after obtaining the pre-fused image and the brightness layer, the pre-fused image and the final brightness layer can be added together to obtain the final fused image. This final fused image can reduce artifacts and brightness loss, thereby enabling better fusion of infrared and visible light images to obtain a fused image containing rich feature information.

[0184] The present invention provides a method for fusing infrared and visible light images of ultra-high voltage converter transformer bushings. This method acquires infrared and visible light images of the same bushing, then preprocesses these images to make them usable for subsequent operations. The segmented infrared and visible light images are refined to form corresponding weight maps. Brightness extraction can be performed on the preprocessed infrared and visible light images. A pre-fused image is then formed by single-scale weighting of the preprocessed infrared and visible light images with the corresponding weight maps. Finally, the pre-fused image is added to the brightness layer to obtain the final fused image.

[0185] It should also be noted that the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element.

[0186] The above are merely embodiments of this application and are not intended to limit the scope of this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of the claims of this application.

Claims

1. A method for fusing infrared and visible light images of ultra-high voltage converter transformer bushings, characterized in that, The method includes: Acquire infrared and visible light images of the same tube; The infrared image and the visible light image are preprocessed; Image segmentation is performed on the preprocessed infrared image and the visible light image; The segmented infrared and visible light images are then refined to form corresponding weighted images; Brightness extraction is performed on the preprocessed infrared and visible light images; The preprocessed infrared and visible light images are weighted by a single scale with the corresponding weight map to form a pre-fused image; The pre-fused image is combined with the luminance layer after luminance extraction to obtain the final fused image; Preprocessing of the infrared image and the visible light image includes: Acquire the infrared image and the visible light image; The enhanced infrared image is obtained using formula (1): , Official (1) in, This represents the enhanced infrared image. This represents the initial infrared image. Represents the operators of the GF-LS model; The enhanced visible light image is obtained using formula (2): , Official (2) in, This represents the enhanced visible light image. This represents the initial visible light image; Acquire the enhanced infrared and visible light images; The enhanced infrared image is further optimized using formula (3): , Official (3) in, For the optimized infrared image, and Indicates a Fourier transform pair. The delta function represents the Fast Fourier Transform. This indicates a forward difference operator or a backward difference operator. This indicates the use of a guided filter for enhancing the infrared image. shaft and Guided filtering is performed using the axis gradient. The enhanced visible light image is further optimized using formula (4): , Official (4) in, For the optimized visible light image, This indicates the use of a guided filter for enhancing the visible light image. shaft and Guided filtering is performed using the axis gradient.

2. The method according to claim 1, characterized in that, Image segmentation of the preprocessed infrared and visible light images includes: Obtain the optimized infrared and visible light images; The infrared image is segmented using formulas (5) and (6): , Official (5) , Official (6) in, Segmenting the image using infrared light, For variables in infrared light images, Represents the histogram of an infrared light image. As the independent variable, The threshold for infrared light images; The appropriate infrared threshold is obtained according to formula (7): , Official (7) Among them, the The infrared cross-entropy function is... The grayscale levels of the infrared image; Segmenting visible light images using formulas (8) and (9): , Official (8) , Official (9) in, Segmenting the image for visible light, For variables in a visible light image, This represents the histogram of a visible light image. Threshold for visible light images; The appropriate visible light threshold is obtained according to formula (10): , Official (10) in, The visible light cross-entropy function, The grayscale level of the visible light image.

3. The method according to claim 1, characterized in that, The segmented infrared and visible light images are then refined to form corresponding weighted images; Acquire the segmented infrared light image; The segmented infrared image is processed using the MCET-HHO method, and the infrared image weight map is obtained according to formula (11): , Official (11) in, Represents the weight map of infrared light images. This represents the MCET-HHO operator; Obtain the infrared image weight map; The refined infrared image weight map is obtained according to formula (12): , Official (12) in, This represents the weighted image of the refined infrared light image. This represents the operator of the GF-LS model.

4. The method according to claim 3, characterized in that, The segmented infrared and visible light images are then refined to form corresponding weighted maps, including: Obtain the segmented visible light image; The segmented visible light image is processed using the MCET-HHO method, and the visible light image weight map is obtained according to formula (13): , Official (13) in, Represents the weight map of a visible light image; Obtain the visible light image weight map; The refined visible light image weight map is obtained according to formula (14): Official (14) in, This represents the weighted image of the refined visible light image.

5. The method according to claim 4, characterized in that, Brightness extraction of the preprocessed infrared and visible light images includes: Median filtering is applied to the preprocessed infrared and visible light images; Substitute the infrared and visible light images after median filtering into formula (15) to obtain the difference image between the infrared and visible light images: Official (15) in, For difference images, This represents the image after median filtering. This represents the image after processing using the GF-LS model; Substitute the difference image between the infrared image and the visible light image into formula (16) to obtain the enhanced infrared brightness layer and the visible light brightness layer: , Official (16) in, This indicates the enhanced brightness layer; Substitute the enhanced infrared brightness layer and visible brightness layer into formula (17) to obtain the final brightness layer: Official (17) in, This indicates the final brightness layer.

6. The method according to claim 5, characterized in that, The preprocessed infrared and visible light images are weighted at a single scale with their corresponding weight maps to form a pre-fused image, including: Acquire the preprocessed infrared and visible light images; Obtain the refined infrared image weight map and visible light image weight map; Obtain the pre-fused image according to formula (18): Official (18) in, This represents the pre-fused image.

7. The method according to claim 6, characterized in that, The final fused image is obtained by adding the pre-fused image to the luminance layer after luminance extraction, including: Acquire the pre-fused image; Obtain the final brightness layer; The final fused image is obtained according to formula (19): Official (19) in, This represents the final merged image.

Images