A bimodal image fusion method for near-infrared fluorescence imaging and white light imaging

Abstract

The application provides a dual-mode image fusion method for near-infrared fluorescence imaging and white light imaging, which comprises the following steps: acquiring registered white light images and fluorescence images, wherein the white light images have Y, U and V channels, and the fluorescence images have a Y channel; performing pseudo-color preprocessing on the fluorescence images, mapping the Y channel of the fluorescence images to U and V channels, so that the fluorescence images have U and V channels; fusing the U and V channels of the preprocessed fluorescence images; repeating the above fusion process for each pixel in the white light images and the fluorescence images to obtain U channel fusion images and V channel fusion images; and combining the Y channel of the white light images, the U channel fusion images and the V channel fusion images to obtain a final fusion image. The application can calculate self-adaptive weights at the pixel level during the process of near-infrared fluorescence molecular image surgery navigation to perform optimal fluorescence and white light image fusion, so as to sufficiently retain texture features.

Description

Technical Field

[0001] This invention relates to the field of medical imaging and image processing technology, specifically to a dual-modal image fusion method for near-infrared fluorescence imaging and white light imaging. Background Technology

[0002] Dynamic imaging of the lesion area during surgery provides surgeons with precise surgical guidance. Traditional imaging techniques (such as CT, MRI, PET, SPECT, etc.) struggle to provide clear, real-time lesion localization during surgery. While advancements in camera imaging technology and manufacturing processes have enabled color visible light cameras to meet surgeons' requirements for resolution, image quality, and frame rate during surgery, visible light cameras alone cannot capture spectral information beyond the visible light band. Therefore, with only a visible light camera available, surgeons must rely on visual observation and experience to locate lesions. However, under excitation light, labeling diseased cells with targeted fluorescent probes (such as indocyanine green) generates near-infrared emission light in specific wavelengths. This near-infrared emission light can then be captured and imaged using a near-infrared camera, revealing details that are difficult to distinguish with visible light cameras (such as polyps, tumors, and even blood vessels). This imaging method is called near-infrared fluorescent molecular imaging. Using near-infrared fluorescent molecular imaging to label tumor lesions during surgery assists surgeons in resection, achieving precision medicine. Near-infrared fluorescence molecular imaging offers advantages such as high resolution, no ionizing radiation, and real-time imaging, making it suitable for highlighting lesion areas marked by fluorescent probes during surgery. However, using a near-infrared camera alone cannot provide the appearance of human organs under visible light (including color, shape, and surface texture). Simultaneous imaging using both visible light and near-infrared cameras across multiple spectra allows for capturing both the natural morphology of organs under visible light illumination and accurate localization of fluorescently marked pathological areas, thus revealing rich pathological details.

[0003] Therefore, rapid and accurate image fusion of near-infrared fluorescence imaging and white light imaging can help doctors make better judgments. However, current image fusion methods are based on constant weights, while natural images and fluorescence images have significant differences in different texture features. Using constant weights can easily weaken clinically significant texture features, but to date there is no image fusion method specifically designed for near-infrared fluorescence imaging. Summary of the Invention

[0004] To address the shortcomings of existing technologies, the present invention aims to provide a dual-modal image fusion method for near-infrared fluorescence imaging and white light imaging. This method can calculate adaptive weights at the pixel level and perform optimal fluorescence and white light image fusion during near-infrared fluorescence molecular imaging surgical navigation to fully preserve texture features.

[0005] The present invention achieves the above objectives through the following technical solutions:

[0006] A dual-modal image fusion method for near-infrared fluorescence imaging and white light imaging includes the following steps:

[0007] Step S1: Obtain the registered white light image and fluorescence image, wherein the white light image has three channels: Y, U, and V, and the fluorescence image has a Y channel;

[0008] Step S2: Perform pseudo-color preprocessing on the fluorescence image, mapping the Y channel of the fluorescence image to the U and V channels, so that the fluorescence image has two channels, U and V.

[0009] Step S3: The U channel and V channel of the preprocessed fluorescence image are fused. The fusion process includes:

[0010] For white light and fluorescence images with the same channel, calculate the mean and median of the region of interest centered on the image with a size of 3x3, and use them as the base pixel values. Then calculate the detail pixel values ​​of the white light and fluorescence images respectively; calculate the salient pixel values ​​of the white light and fluorescence images respectively.

[0011] Based on the obtained significant pixel values, the fusion weights of the white light image and the fluorescence image are calculated, and the detail pixel values ​​of the white light image and the fluorescence image are fused based on the fusion weights to obtain the fused detail pixel values.

[0012] The base pixel values ​​of the white light image and the fluorescence image are fused to obtain the fused base pixel values;

[0013] The base pixel value and the detail pixel value are added together to obtain the merged pixel value;

[0014] Step S4: Repeat the fusion process of step S3 for each pixel in the white light image and the fluorescence image to obtain the U-channel fused image and the V-channel fused image.

[0015] Step S5: Combine the Y-channel, U-channel fused image and V-channel fused image of the white light image to obtain the final fused image.

[0016] According to the present invention, a dual-modal image fusion method for near-infrared fluorescence imaging and white light imaging is provided, wherein the pseudo-color preprocessing includes:

[0017] Receive fluorescence grayscale image data, where each pixel has a grayscale value Y, ranging from 0 to 255;

[0018] Based on the preset pseudo-color and the linear interpolation mapping relationship between grayscale value Y and RGB three channels, each grayscale value Y is converted into the corresponding pseudo-color RGB three channel value; wherein, the linear interpolation mapping relationship is determined based on the brightest value 255 of the fluorescent grayscale Y channel corresponding to pseudo-color RGB (0, 255, 0), and the darkest value 0 corresponding to pseudo-color RGB (0, 0, 0).

[0019] Based on the RGB to YUV conversion formula, calculate and output the U and V channel values ​​corresponding to the pseudo-color RGB three-channel values.

[0020] According to the present invention, a dual-modal image fusion method for near-infrared fluorescence imaging and white light imaging is provided. In step S3, the mean value of the region of interest centered on the image and with a size of 3x3 is calculated. The positions of all pixels in the region are (i-1,j-1), (i,j-1), (i+1,j-1), (i-1,j), (i,j), (i+1,j), (i-1,j+1), (i,j+1), (i+1,j+1).

[0021] Calculate the median m of all pixels within the region of interest to obtain the mean μ for white light pixels. β With median m β The mean value μ of fluorescent pixels α With median m α This can be expressed as formula (1):

[0022]

[0023] Where, μ β and μ α These are the base pixel values ​​for white light pixels and fluorescent pixels.

[0024] According to the present invention, a dual-modal image fusion method for near-infrared fluorescence imaging and white light imaging is provided, wherein the detailed pixel values ​​d of the white light image and the fluorescence image are calculated. β and d α This can be expressed as formula (2):

[0025]

[0026] According to the present invention, a dual-modal image fusion method for near-infrared fluorescence imaging and white light imaging is provided, wherein the significant pixel values ​​s of the white light image and the fluorescence image are calculated. β and s α This can be expressed as formula (3):

[0027]

[0028] Where |·| represents the L1 norm.

[0029] According to the dual-modal image fusion method for near-infrared fluorescence imaging and white light imaging provided by the present invention, the fusion weights of the white light image and the fluorescence image are calculated and expressed as formula (4).

[0030]

[0031] According to the present invention, a dual-modal image fusion method for near-infrared fluorescence imaging and white light imaging is provided, which fuses the detail pixel values ​​of white light images and fluorescence images to obtain fused detail pixel values ​​d. f (i,j), represented by formula (5):

[0032] d f (i,j)=w β (i,j)×d β (i,j)+w α (i,j)×d α (i,j) (5)

[0033] According to the present invention, a dual-modal image fusion method for near-infrared fluorescence imaging and white light imaging is provided, wherein the basic pixel values ​​of white light images and fluorescence images are fused to obtain fused basic pixel values ​​μ. f (i,j), represented by formula (6):

[0034]

[0035] According to the dual-modal image fusion method for near-infrared fluorescence imaging and white light imaging provided by the present invention, the fusion base pixel value and the fusion detail pixel value are added together to obtain the fusion pixel value f(i,j), which is expressed as formula (7):

[0036] f(i,j)=μ f (i,j)+d f (i,j) (7)

[0037] According to the present invention, a dual-modal image fusion method for near-infrared fluorescence imaging and white light imaging is provided. After obtaining the final fused image after dual-modal image fusion, a smoothing filtering algorithm is applied to filter the fused image to remove noise and artifacts present in the image. The filtering effect is optimized by adjusting the parameters of the filtering algorithm, thereby obtaining a smoother and higher quality fused image.

[0038] After smoothing and filtering, the fused image undergoes color correction processing according to actual needs or preset color standards to ensure the color accuracy and consistency of the fused image.

[0039] Therefore, compared with the prior art, the method provided by the present invention has the following beneficial effects:

[0040] 1. This invention constructs a targeted image fusion method for fluorescence imaging and white light imaging, providing surgeons with richer fused images during surgery, thereby enabling better surgery and ultimately benefiting patients.

[0041] 2. This invention relates to a near-infrared fluorescence imaging surgical navigation device, which can adaptively adjust the fusion weight according to the texture features of the current fluorescence image and white light image, so that the fused image can retain more texture details and features, making it easier for surgeons to locate lesions and perform surgery.

[0042] 3. This invention makes full use of the multi-channel information of white light images and fluorescence images. White light images have three channels: Y, U, and V, while fluorescence images also have two channels: U and V after pseudo-color preprocessing. This allows the fusion process to more comprehensively consider the color and brightness information of the images, thereby improving the quality of the fused image.

[0043] 4. This invention effectively preserves image detail information during the fusion process by calculating base pixel values ​​and detail pixel values. The base pixel values ​​reflect the overall brightness level of the image, while the detail pixel values ​​contain information about local changes in the image, helping to better preserve image details and textures during the fusion process.

[0044] 5. This invention determines the fusion weight by calculating the salient pixel values ​​of the white light image and the fluorescence image. The salient pixel values ​​reflect the amount of information in important regions or features of the image. By calculating the fusion weight based on the salient pixel values, it can be ensured that the fused image has higher contrast and clarity in important regions or features.

[0045] 6. In the fusion process, this invention optimizes the contrast of the fused image by weighting the basic pixel values ​​and detail pixel values, which helps to better distinguish different regions or features in the fused image and improves the readability and interpretability of the image.

[0046] The present invention will now be described in further detail with reference to the accompanying drawings and specific embodiments. Attached Figure Description

[0047] Figure 1 This is a flowchart of an embodiment of a dual-modal image fusion method for near-infrared fluorescence imaging and white light imaging according to the present invention.

[0048] Figure 2 This is a schematic diagram illustrating the principle of dual-modal image fusion of fluorescence imaging and white light imaging in an embodiment of the present invention.

[0049] Figure 3 This is a schematic diagram illustrating the pixel-level principle of the fusion algorithm in an embodiment of a dual-modal image fusion method for near-infrared fluorescence imaging and white light imaging according to the present invention.

[0050] Figure 4 This is a schematic diagram of a white light image in an embodiment of a dual-modal image fusion method for near-infrared fluorescence imaging and white light imaging according to the present invention.

[0051] Figure 5 This is a schematic diagram of a fluorescence image in an embodiment of a dual-modal image fusion method for near-infrared fluorescence imaging and white light imaging according to the present invention.

[0052] Figure 6 This is a schematic diagram of the final fused image in an embodiment of a dual-modal image fusion method for near-infrared fluorescence imaging and white light imaging according to the present invention. Detailed Implementation

[0053] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this invention. All other embodiments obtained by those skilled in the art based on the embodiments of this invention without creative effort are within the scope of protection of this invention.

[0054] In this document, the term "embodiment" means that a particular feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of this application. The appearance of this phrase in various places throughout the specification does not necessarily refer to the same embodiment, nor is it a separate or alternative embodiment mutually exclusive with other embodiments. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described herein can be combined with other embodiments.

[0055] See Figures 1 to 6 This invention provides a dual-modal image fusion method for near-infrared fluorescence imaging and white light imaging, the method comprising the following steps:

[0056] Step S1: Obtain the registered white light image and fluorescence image, wherein the white light image has three channels: Y, U, and V, and the fluorescence image has a Y channel;

[0057] Step S2: Perform pseudo-color preprocessing on the fluorescence image, mapping the Y channel of the fluorescence image to the U and V channels, so that the fluorescence image has two channels, U and V.

[0058] Step S3: The U channel and V channel of the preprocessed fluorescence image are fused. The fusion process includes:

[0059] For white light and fluorescence images with the same channel, calculate the mean and median of the region of interest centered on the image with a size of 3x3, and use them as the base pixel values. Then calculate the detail pixel values ​​of the white light and fluorescence images respectively; calculate the salient pixel values ​​of the white light and fluorescence images respectively.

[0060] Based on the obtained significant pixel values, the fusion weights of the white light image and the fluorescence image are calculated, and the detail pixel values ​​of the white light image and the fluorescence image are fused based on the fusion weights to obtain the fused detail pixel values.

[0061] The base pixel values ​​of the white light image and the fluorescence image are fused to obtain the fused base pixel values;

[0062] The base pixel value and the detail pixel value are added together to obtain the merged pixel value;

[0063] Step S4: Repeat the fusion process of step S3 for each pixel in the white light image and the fluorescence image to obtain the U-channel fused image and the V-channel fused image.

[0064] Step S5: Combine the Y-channel, U-channel fused image and V-channel fused image of the white light image to obtain the final fused image.

[0065] In step S2 above, the pseudo-color preprocessing specifically includes:

[0066] Receive fluorescence grayscale image data, where each pixel has a grayscale value Y, ranging from 0 to 255;

[0067] Based on the preset pseudo-color and the linear interpolation mapping relationship between grayscale value Y and the RGB three channels, each grayscale value Y is converted into the corresponding pseudo-color RGB three-channel value. The linear interpolation mapping relationship is determined based on the brightest value 255 of the fluorescent grayscale Y channel corresponding to pseudo-color RGB (0, 255, 0), and the darkest value 0 corresponding to pseudo-color RGB (0, 0, 0). According to the RGB to YUV conversion formula, the U and V channel values ​​corresponding to the pseudo-color RGB three-channel values ​​are calculated and output. Therefore, in this embodiment, after determining the RGB three channels of the color to be mapped by the fluorescent grayscale Y channel, taking green pseudo-color as an example, its RGB three channels are set to (0, 255, 0), corresponding to the brightest value 255 of the fluorescent grayscale Y channel. Simultaneously, the pseudo-color RGB three-channel corresponding to the darkest value 0 of the fluorescent grayscale Y channel must be (0, 0, 0). The mapping relationship between the grayscale Y channel from 0 to 255 and the green pseudo-color three-channel from (0, 0, 0) to (0, 255, 0) can be obtained through linear interpolation. Then, based on the RGB to YUV conversion, the correspondence between the RGB three channels and the U and V channels can be calculated, thus obtaining the result after pseudo-color preprocessing.

[0068] In steps S1 and S2 above, it is necessary to acquire registered white light and fluorescence image pairs. The white light image has three channels (Y, U, and V), while the fluorescence image has only one channel (luminance). Before image fusion, the fluorescence image needs to be preprocessed to obtain pseudo-color, i.e., mapping the luminance to the U and V channels. This gives the fluorescence image two channels (U and V). The fusion process of the fluorescence image and the exposure image is as follows: Figure 2 As shown, the fused image retains the Y channel of the white light image, and performs color gamut fusion of the white light image and the fluorescence image in the U channel and V channel respectively.

[0069] In step S3 above, the specific fusion process of the U and V channels is as follows: Figure 3 As shown, for white light pixel β(i,j) and fluorescent pixel α(i,j) at the same position (i,j), the mean μ of the region of interest centered on the image and with a size of 3x3 is first calculated using formula (1). The positions of all pixels in this region are (i-1,j-1), (i,j-1), (i+1,j-1), (i-1,j), (i,j), (i+1,j), (i-1,j+1), (i,j+1), (i+1,j+1).

[0070] Similarly, by calculating the median m of all pixels within the region of interest, we obtain the mean μ for the white light pixels. β With median m β The mean value μ of fluorescent pixels α With median m α This can be expressed as formula (1):

[0071]

[0072] Where, μ β and μ α These are the base pixel values ​​for white light pixels and fluorescent pixels.

[0073] Next, the detail pixel values ​​d of the white light image and the fluorescence image are calculated respectively. β and d α This can be expressed as formula (2):

[0074]

[0075] After obtaining the detailed pixels, the salient pixel values ​​s of the white light image and the fluorescence image are calculated separately. β and s α This can be expressed as formula (3):

[0076]

[0077] Where |·| represents the L1 norm.

[0078] Furthermore, by utilizing the significant pixel values ​​s of white light and fluorescence images... β and s α Calculate the fusion weight w between the white light image and the fluorescence image. β and w α This is represented by formula (4).

[0079]

[0080] After obtaining the fusion weights, the detail pixel values ​​of the white light image and the fluorescence image are fused based on the fusion weights to obtain the fused detail pixel value d. f (i,j), represented by formula (5):

[0081] d f (i,j)=w β (i,j)×d β (i,j)+w α (i,j)×d α (i,j) (5)

[0082] Similarly, the base pixel values ​​of the white light image and the fluorescence image are fused to obtain the fused base pixel value μ. f (i,j), represented by formula (6):

[0083]

[0084] Finally, the base pixel value and the detail pixel value are added together to obtain the merged pixel value f(i,j), which is expressed as formula (7):

[0085] f(i,j)=μ f (i,j)+d f (i,j) (7)

[0086] In step S3 above, a 3x3 sliding window process is performed on the white light image and the fluorescence image in the U channel and V channel respectively. For example, in the U channel, a 3x3 sliding window is performed on the white light image and the fluorescence image simultaneously. Specifically, the basic pixel values ​​of the white light image and the fluorescence image are calculated according to formula (1), the detail pixel values ​​of the white light image and the fluorescence image are calculated according to formula (2), and the salient pixel values ​​of the white light image and the fluorescence image are calculated according to formula (3).

[0087] Based on the obtained significant pixel values, the fusion weights of the white light image and the fluorescence image can be calculated according to formula (4), and the fusion detail pixel values ​​can be obtained according to formula (5). Similarly, the fusion base pixel values ​​can be obtained according to formula (6). Finally, the final fusion pixel values ​​can be obtained according to formula (7).

[0088] Repeating the above operation on all pixel values ​​in the image will yield a U-channel fused image, and repeating the above operation on the V-channel will yield a V-channel fused image.

[0089] Finally, by combining the Y channel of the white light image, the U channel of the fused image, and the V channel of the fused image, the final fused image can be obtained.

[0090] In this embodiment, after obtaining the final fused image after dual-modal image fusion, a smoothing filtering algorithm is applied to filter the fused image to remove noise and artifacts in the image; wherein, by adjusting the parameters of the filtering algorithm, the filtering effect is optimized, thereby obtaining a smoother and higher quality fused image.

[0091] After smoothing and filtering, the fused image undergoes color correction processing according to actual needs or preset color standards to ensure the color accuracy and consistency of the fused image.

[0092] In summary, this embodiment, by specifically constructing an image fusion method for fluorescence imaging and white light imaging, provides surgeons with a more informative fused image during surgery, thereby enabling better surgical procedures and ultimately benefiting patients. This embodiment, targeting near-infrared fluorescence imaging surgical navigation equipment, can adaptively adjust the fusion weights based on the texture features of the current fluorescence and white light images, allowing the fused image to retain more texture details, facilitating surgeons in locating lesions and performing surgery.

[0093] Furthermore, this embodiment makes full use of the multi-channel information of white light images and fluorescence images. White light images have three channels: Y, U, and V, while fluorescence images also have two channels: U and V after pseudo-color preprocessing. This allows the fusion process to more comprehensively consider the color and brightness information of the images, thereby improving the quality of the fused image.

[0094] Furthermore, this embodiment effectively preserves image detail information during the fusion process by calculating base pixel values ​​and detail pixel values. The base pixel values ​​reflect the overall brightness level of the image, while the detail pixel values ​​contain local variation information, helping to better preserve image details and textures during the fusion process.

[0095] Furthermore, this embodiment determines the fusion weight by calculating the salient pixel values ​​of the white light image and the fluorescence image. The salient pixel values ​​reflect the amount of information in important regions or features in the image. By calculating the fusion weight based on the salient pixel values, it can be ensured that the fused image has higher contrast and clarity in important regions or features.

[0096] Furthermore, in this embodiment, by weighting the base pixel values ​​and detail pixel values ​​during the fusion process, the method can optimize the contrast of the fused image, which helps to better distinguish different regions or features in the fused image and improve the readability and interpretability of the image.

[0097] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0098] The above embodiments are merely preferred embodiments of the present invention and should not be construed as limiting the scope of protection of the present invention. Any non-substantial changes and substitutions made by those skilled in the art based on the present invention shall fall within the scope of protection claimed by the present invention.

Claims

1. A dual-modal image fusion method for near-infrared fluorescence imaging and white light imaging, characterized in that, Includes the following steps: Step S1: Obtain the registered white light image and fluorescence image, wherein the white light image has three channels: Y, U, and V, and the fluorescence image has a Y channel; Step S2: Perform pseudo-color preprocessing on the fluorescence image, mapping the Y channel of the fluorescence image to the U and V channels, so that the fluorescence image has two channels, U and V. Step S3: The U channel and V channel of the preprocessed fluorescence image are fused. The fusion process includes: For white light and fluorescence images of the same channel, calculate the mean and median of a 3x3 region of interest centered on the image. Use the calculated mean as the base pixel value to calculate the detail pixel values ​​for both the white light and fluorescence images. Also calculate the salient pixel values ​​for both images. and , expressed as formula (3): Where |·| represents the L1 norm, This represents the average value of white light pixels. The median of white light pixels. This represents the average value of the fluorescent pixels. The median of fluorescent pixels; Based on the obtained significant pixel values, calculate the fusion weights of the white light image and the fluorescence image. and , which can be represented by formula (4) The detail pixel values ​​of the white light image and the fluorescence image are fused based on the fusion weights to obtain the fused detail pixel values; The base pixel values ​​of the white light image and the fluorescence image are fused to obtain the fused base pixel values; The base pixel value and the detail pixel value are added together to obtain the merged pixel value; Step S4: Repeat the fusion process of step S3 for each pixel in the white light image and the fluorescence image to obtain the U-channel fused image and the V-channel fused image. Step S5: Combine the Y-channel, U-channel fused image and V-channel fused image of the white light image to obtain the final fused image.

2. The method according to claim 1, characterized in that: The pseudo-color preprocessing includes: Receive fluorescence grayscale image data, where each pixel has a grayscale value Y, ranging from 0 to 255; Based on the preset pseudo-color and the linear interpolation mapping relationship between grayscale value Y and RGB three channels, each grayscale value Y is converted into the corresponding pseudo-color RGB three channel value; wherein, the linear interpolation mapping relationship is determined based on the brightest value 255 of the fluorescent grayscale Y channel corresponding to pseudo-color RGB (0, 255, 0), and the darkest value 0 corresponding to pseudo-color RGB (0, 0, 0). Based on the RGB to YUV conversion formula, calculate and output the U and V channel values ​​corresponding to the pseudo-color RGB three-channel values.

3. The method according to claim 2, characterized in that: In step S3, the mean of the region of interest, centered on the image and measuring 3x3, is calculated. The positions of all pixels in this region are as follows: Calculate the median of all pixels within the region of interest. This yields the mean value for white light pixels. With median The average value of fluorescent pixels With median , expressed as formula (1): in, and These are the base pixel values ​​for white light pixels and fluorescent pixels.

4. The method according to claim 3, characterized in that: The calculation of detailed pixel values ​​of white light images and fluorescence images and , expressed as formula (2): 。 5. The method according to claim 4, characterized in that: The detail pixel values ​​of the white light image and the fluorescence image are fused to obtain the fused detail pixel values. , expressed as formula (5): 。 6. The method according to claim 5, characterized in that: The base pixel values ​​of the white light image and the fluorescence image are fused to obtain the fused base pixel values. , expressed as formula (6): 。 7. The method according to claim 6, characterized in that: The base pixel value and the detail pixel value are added together to obtain the merged pixel value. , expressed as formula (7): 。 8. The method according to any one of claims 1 to 7, characterized in that, Also execute: After obtaining the final fused image after dual-modal image fusion, a smoothing filtering algorithm is applied to filter the fused image to remove noise and artifacts. The filtering effect is optimized by adjusting the parameters of the filtering algorithm, resulting in a smoother and higher-quality fused image. After smoothing and filtering, the fused image undergoes color correction processing according to actual needs or preset color standards to ensure the color accuracy and consistency of the fused image.

Images