A panchromatic-multiple spectral remote sensing image fusion method, system and readable storage medium
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- PEARL RIVER HYDRAULIC RES INST OF PEARL RIVER WATER RESOURCES COMMISSION
- Filing Date
- 2026-02-11
- Publication Date
- 2026-06-19
Smart Images
Figure CN122243758A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of remote sensing image processing and image fusion technology, and more specifically, to a panchromatic-multispectral remote sensing image fusion method, system, and readable storage medium. Background Technology
[0002] In remote sensing imaging systems, due to limitations in sensor design and energy distribution mechanisms, it is often difficult to simultaneously acquire remote sensing images with high spatial resolution and high spectral resolution. Panchromatic remote sensing images have high spatial resolution but limited spectral information; multispectral remote sensing images contain multiple spectral bands but have relatively low spatial resolution. Therefore, how to effectively fuse the spatial detail information of panchromatic images into multispectral images is an important technical problem in the field of remote sensing image processing.
[0003] Existing panchromatic-multispectral fusion methods mostly employ luminance component replacement or direct detail injection strategies for fusion. However, these methods typically do not model the illumination non-uniformity in panchromatic images, which can easily lead to spectral distortion and color shifts in shadow or high-contrast regions. Furthermore, some methods are heavily reliant on specific mathematical transformations or fixed processing procedures, limiting their applicability and robustness. Summary of the Invention
[0004] In view of the above problems, the purpose of this invention is to provide a panchromatic-multispectral remote sensing image fusion method, system, and readable storage medium. By performing illumination component suppression processing on the panchromatic image and constructing a proportional modulation model by combining the brightness component of the multispectral image, stable spatial detail information is injected into the multispectral image, thereby improving spatial resolution while maintaining good spectral consistency.
[0005] The first aspect of this invention provides a panchromatic-multispectral remote sensing image fusion method. The method includes:
[0006] S1: Acquire panchromatic remote sensing images with high spatial resolution and multispectral remote sensing images with low spatial resolution;
[0007] S2: Perform spatial resampling on the multispectral remote sensing image to make its spatial resolution consistent with that of the panchromatic remote sensing image;
[0008] S3: Perform statistical transformation processing on the resampled multispectral remote sensing image to extract the brightness component that can characterize the overall brightness information of the multispectral remote sensing image;
[0009] S4: Perform illumination suppression processing based on multi-scale spatial filtering on the panchromatic remote sensing image to obtain an enhanced panchromatic image for characterizing spatial detail information;
[0010] S5: Construct a proportional modulation model based on the proportional relationship between the enhanced panchromatic image and the luminance component;
[0011] S6: Based on the proportional modulation model, the spatial detail information contained in the enhanced panchromatic image is injected into each band of the multispectral remote sensing image in a multiplicative proportional manner to achieve proportional modulation of the multispectral remote sensing image bands.
[0012] S7: Combines the multispectral bands after proportional modulation and outputs a fused high spatial resolution multispectral remote sensing image.
[0013] Preferably, the statistical transformation processing described in S3 includes: principal component analysis, covariance analysis, or luminance information statistics.
[0014] Preferably, the luminance component in the multispectral image that can characterize the overall luminance information of the multispectral remote sensing image is:
[0015]
[0016] in These are the weighting coefficients, which can be obtained through linear regression, statistical estimation, or pre-setting. It is a multispectral remote sensing image with the same spatial resolution as the panchromatic remote sensing image.
[0017] Preferably, step S4 specifically involves: performing illumination suppression processing based on multi-scale spatial filtering on the panchromatic remote sensing image to separate and suppress the illumination component in the panchromatic remote sensing image, obtaining the separated reflection component in the panchromatic remote sensing image, and then obtaining an enhanced panchromatic image representing spatial detail information based on the separated reflection component in the panchromatic remote sensing image.
[0018] Preferably, the separated reflectance components in the panchromatic remote sensing image are represented as follows:
[0019]
[0020]
[0021]
[0022] in, This represents the reflectance component in panchromatic remote sensing images that contains structural and textural information. Represents panchromatic remote sensing imagery. This represents the illumination component in a panchromatic remote sensing image. The standard deviation is Gaussian wrapping function.
[0023] Preferably, the enhanced panchromatic image representing spatial detail information is:
[0024]
[0025] in, This is an enhanced panchromatic image after illumination suppression. Indicates the scale quantity. Represents the Gaussian kernel. Represents convolution. This represents the standard deviation of the Gaussian function at different scales. Indicates weight, Indicates a stable term.
[0026] Preferably, S5 includes:
[0027] Luminance component of overall luminance information of multispectral remote sensing image Extracting the reflection component , making the reflected component Reflectance component in panchromatic remote sensing image Unify to the reflection component domain to complete alignment:
[0028]
[0029] Based on the same-domain reflection component and reflection component Construct an example modulation model:
[0030]
[0031] in, The modulation intensity coefficient, Take the modulation model as an example.
[0032] Preferably, in S6, the spatial detail information contained in the enhanced panchromatic image is injected into each band of the multispectral remote sensing image in a multiplicative manner, as expressed by:
[0033]
[0034] in, For multispectral images The reflection component modulation factor after MSR-Retinex processing for each band.
[0035] A second aspect of the present invention provides a panchromatic-multispectral remote sensing image fusion system, including a memory and a processor. The memory includes a panchromatic-multispectral remote sensing image fusion method program. When the processor executes the panchromatic-multispectral remote sensing image fusion method program, it implements the steps of the panchromatic-multispectral remote sensing image fusion method.
[0036] A third aspect of the present invention provides a computer-readable storage medium comprising a panchromatic-multispectral remote sensing image fusion method program, wherein when the panchromatic-multispectral remote sensing image fusion method program is executed by a processor, the steps of the panchromatic-multispectral remote sensing image fusion method are implemented.
[0037] Compared with existing technologies, the beneficial effects of the technical solution of this invention are as follows: A panchromatic-multispectral remote sensing image fusion method, system, and readable storage medium. This invention effectively reduces the impact of uneven illumination in panchromatic images on the fusion result through illumination suppression processing; it significantly reduces spectral distortion during the fusion process through luminance component modeling and proportional modulation mechanisms; thereby maintaining good spectral consistency while improving image spatial resolution. Furthermore, the method described in this invention requires no training samples during the fusion process, has a clear computational structure, and good interpretability; it has good adaptability to different remote sensing platforms and imaging conditions; and it is suitable for engineering implementation and large-scale remote sensing data processing. Attached Figure Description
[0038] Figure 1 This is a flowchart of a panchromatic-multispectral remote sensing image fusion method as described in Example 1. Detailed Implementation
[0039] To better understand the above-mentioned objectives, features, and advantages of the present invention, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments. It should be noted that, unless otherwise specified, the embodiments and features described in these embodiments can be combined with each other.
[0040] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and therefore the scope of protection of the invention is not limited to the specific embodiments disclosed below.
[0041] Example 1
[0042] like Figure 1 As shown, this embodiment discloses a panchromatic-multispectral remote sensing image fusion method, the method comprising:
[0043] S1: Acquire panchromatic remote sensing images with high spatial resolution and multispectral remote sensing images with low spatial resolution.
[0044] The multispectral remote sensing image in this embodiment is:
[0045]
[0046] in, Number of multispectral bands
[0047] The panchromatic remote sensing image PAN (high resolution) is as follows:
[0048]
[0049] The multispectral remote sensing image upsampled to P resolution is as follows:
[0050]
[0051] In this embodiment, the multispectral remote sensing image includes at least one of the visible light band, multispectral band, or hyperspectral band.
[0052] S2: Perform spatial resampling on the multispectral remote sensing image to make its spatial resolution consistent with that of the panchromatic remote sensing image.
[0053] S3: Perform statistical transformation processing on the resampled multispectral remote sensing image to extract the brightness component that can characterize the overall brightness information of the multispectral remote sensing image.
[0054] In this embodiment, the statistical transformation process may include principal component analysis, covariance analysis, or other linear or nonlinear statistical methods capable of extracting luminance information. In some specific cases, the first principal component may be used as the luminance component.
[0055] In this embodiment, the brightness component in the multispectral image that can characterize the overall brightness information of the multispectral remote sensing image is:
[0056]
[0057] in These are the weighting coefficients, which can be obtained through linear regression, statistical estimation, or pre-setting. It is a multispectral remote sensing image with the same spatial resolution as the panchromatic remote sensing image.
[0058] S4: Perform illumination suppression processing based on multi-scale spatial filtering on the panchromatic remote sensing image to obtain an enhanced panchromatic image for characterizing spatial detail information.
[0059] In this embodiment, the illumination suppression processing described in S4 is performed in the logarithmic domain, and spatial filters with different scale parameters are used to process the panchromatic remote sensing image. The processing results at each scale are fused according to preset weights. The illumination suppression processing can be performed in the spatial domain, frequency domain, or joint domain.
[0060] Specifically, the illumination suppression process includes smoothing the panchromatic remote sensing image at multiple spatial scales to estimate the illumination components at the corresponding scales, and enhancing the spatial structure information in the reflection components by weakening or removing the illumination components.
[0061] It should be noted that the illumination suppression processing described in this embodiment is not limited to a specific algorithm, but refers to a multi-scale processing method that can achieve illumination component suppression and reflection component enhancement.
[0062] As a specific embodiment, the multi-scale spatial filtering includes at least one of Gaussian filtering, mean filtering, guided filtering, or equivalent filtering methods.
[0063] In this embodiment, S4 specifically involves: performing illumination suppression processing based on multi-scale spatial filtering on the panchromatic remote sensing image to separate and suppress the illumination component in the panchromatic remote sensing image, obtaining the separated reflection component in the panchromatic remote sensing image, and then obtaining an enhanced panchromatic image representing spatial detail information based on the separated reflection component in the panchromatic remote sensing image.
[0064] In this embodiment, the enhanced panchromatic image is not used as the final output image, but only as an intermediate processing result for constructing the proportional modulation model.
[0065] In this embodiment, the separated reflectance component in the panchromatic remote sensing image is represented as:
[0066]
[0067]
[0068]
[0069] in, This represents the reflectance component in panchromatic remote sensing images that contains structural and textural information. Represents panchromatic remote sensing imagery. This represents the illumination component in a panchromatic remote sensing image. The standard deviation is The Gaussian surround function, in the multi-scale illumination suppression processing of this embodiment, has a standard deviation of [missing value]. Its contextual function is to adjust the filter scale, therefore This is the scale parameter.
[0070] It should be noted that, according to Retinex theory, panchromatic remote sensing imagery can be represented as:
[0071]
[0072] in, This represents the reflection component, which contains structural and texture information. Indicates the light component
[0073] Taking the logarithm of the panchromatic remote sensing image yields:
[0074]
[0075] This transforms the multiplicative relationship into an additive one, providing a basis for light suppression.
[0076] By estimating the illumination component using a Gaussian surround function, the reflection component can be approximated as:
[0077]
[0078] in, The standard deviation is Gaussian wrapping function.
[0079]
[0080] This is the scale parameter.
[0081] Through this step It actually captured high-frequency edge information in the PAN image.
[0082] In this embodiment, to balance local detail enhancement and global contrast preservation, multi-scale Retinex enhancement is performed on the panchromatic remote sensing image. The enhanced panchromatic image representing spatial detail information is as follows:
[0083]
[0084] in, This is an enhanced panchromatic image after illumination suppression. Indicates the number of scales (3 is commonly used). Represents the Gaussian kernel. Represents convolution. This represents the standard deviation of the Gaussian function at different scales (e.g., 15 / 80 / 250). Indicates weight ( (It can be divided into equal parts) Indicates a stable term, preventing .
[0085] It should be noted that this embodiment uses an enhanced panchromatic image after illumination suppression. As a representation of the structural details of a panchromatic image after illumination suppression, the structural details are spatial features that are robust to illumination changes, and their dynamic range and statistical distribution are suitable as modulation factors in remote sensing image fusion.
[0086] In this embodiment, the structural details refer to the introduction of stabilization terms or constraint mechanisms during the generation process to avoid numerical instability or noise amplification in low-brightness or high-contrast regions.
[0087] In this embodiment, the structural details are configured for proportional modulation or multiplicative injection with the luminance or spectral components in a multispectral remote sensing image.
[0088] S5: Based on the proportional relationship between the enhanced panchromatic image and the luminance component, construct a proportional modulation model.
[0089] In this embodiment, S5 includes:
[0090] Luminance component of overall luminance information of multispectral remote sensing image Extracting the reflection component , making the reflected component Reflectance component in panchromatic remote sensing image Unify to the reflection component domain to complete alignment:
[0091]
[0092] To preserve spectral information while introducing spatial details from panchromatic remote sensing imagery, a method based on the co-domain reflectance component is employed. and reflection component Construct an example modulation model:
[0093]
[0094] in, The modulation intensity coefficient is typically set to [0.3, 0.8]. Take the modulation model as an example.
[0095] It should be noted that the modulation factor in this embodiment is derived from the difference in reflectance components between panchromatic remote sensing images and multispectral remote sensing images, rather than the difference in brightness.
[0096] It should be noted that, in constructing the proportional modulation model in this embodiment, a stabilization term can be introduced to avoid numerical instability caused by excessively small or zero values for the luminance component. The stabilization term is a preset non-zero constant used to prevent abnormal proportional modulation caused by excessively small or zero values for the luminance component.
[0097] S6: Based on the proportional modulation model, the spatial detail information contained in the enhanced panchromatic image is injected into each band of the multispectral remote sensing image in a multiplicative proportional manner, thereby achieving proportional modulation of the multispectral remote sensing image bands.
[0098] In this embodiment, in step S6, the spatial detail information contained in the enhanced panchromatic image is injected into each band of the multispectral remote sensing image in a multiplicative manner, as expressed by:
[0099]
[0100] in, For multispectral images The reflection component modulation factor after MSR-Retinex (multi-scale Retinex) processing of each band.
[0101] S7: Combines the multispectral bands after proportional modulation and outputs a fused high spatial resolution multispectral remote sensing image.
[0102] In summary, this embodiment first performs spatial resampling on the multispectral image and extracts the luminance component through principal component analysis; then, it performs multi-scale illumination suppression processing on the panchromatic remote sensing image to obtain an enhanced panchromatic image; finally, it injects spatial details into each band of the multispectral image through a proportional modulation model, ultimately generating a fused image. This embodiment effectively improves both spatial detail sharpness and spectral consistency in the fused image.
[0103] Example 2
[0104] This embodiment discloses a panchromatic-multispectral remote sensing image fusion system, including a memory and a processor. The memory includes a panchromatic-multispectral remote sensing image fusion method program. When the processor executes the panchromatic-multispectral remote sensing image fusion method program, it implements the steps of a panchromatic-multispectral remote sensing image fusion method as described in Embodiment 1.
[0105] Example 3
[0106] This embodiment discloses a computer-readable storage medium, which includes a panchromatic-multispectral remote sensing image fusion method program. When the panchromatic-multispectral remote sensing image fusion method program is executed by a processor, it implements the steps of a panchromatic-multispectral remote sensing image fusion method as described in Embodiment 1.
[0107] In the several embodiments provided in this application, it should be understood that the disclosed devices and methods can be implemented in other ways. The device embodiments described above are merely illustrative. For example, the division of units is only a logical functional division, and in actual implementation, there may be other division methods, such as: multiple units or components can be combined, or integrated into another system, or some features can be ignored or not executed. In addition, the coupling, direct coupling, or communication connection between the various components shown or discussed can be through some interfaces, and the indirect coupling or communication connection between devices or units can be electrical, mechanical, or other forms.
[0108] The units described above as separate components may or may not be physically separate. The components shown as units may or may not be physical units. They may be located in one place or distributed across multiple network units. Some or all of the units may be selected to achieve the purpose of this embodiment according to actual needs.
[0109] In addition, in the various embodiments of the present invention, each functional unit can be integrated into one processing unit, or each unit can be a separate unit, or two or more units can be integrated into one unit; the integrated unit can be implemented in hardware or in the form of hardware plus software functional units.
[0110] Those skilled in the art will understand that all or part of the steps of the above method embodiments can be implemented by hardware related to program instructions. The aforementioned program can be stored in a computer-readable storage medium. When the program is executed, it performs the steps of the above method embodiments. The aforementioned storage medium includes various media capable of storing program code, such as mobile storage devices, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.
[0111] Alternatively, if the integrated units of this invention are implemented as software functional modules and sold or used as independent products, they can also be stored in a computer-readable storage medium. Based on this understanding, the technical solutions of the embodiments of this invention, or the parts that contribute to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the methods described in the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as mobile storage devices, ROM, RAM, magnetic disks, or optical disks.
Claims
1. A panchromatic-multispectral remote sensing image fusion method, characterized in that, The method includes: S1: Acquire panchromatic remote sensing images with high spatial resolution and multispectral remote sensing images with low spatial resolution; S2: Perform spatial resampling on the multispectral remote sensing image to make its spatial resolution consistent with that of the panchromatic remote sensing image; S3: Perform statistical transformation processing on the resampled multispectral remote sensing image to extract the brightness component that can characterize the overall brightness information of the multispectral remote sensing image; S4: Perform illumination suppression processing based on multi-scale spatial filtering on the panchromatic remote sensing image to obtain an enhanced panchromatic image for characterizing spatial detail information; S5: Construct a proportional modulation model based on the proportional relationship between the enhanced panchromatic image and the luminance component; S6: Based on the proportional modulation model, the spatial detail information contained in the enhanced panchromatic image is injected into each band of the multispectral remote sensing image in a multiplicative proportional manner to achieve proportional modulation of the multispectral remote sensing image bands. S7: Combines the multispectral bands after proportional modulation and outputs a fused high spatial resolution multispectral remote sensing image.
2. The panchromatic-multispectral remote sensing image fusion method according to claim 1, characterized in that, The statistical transformation processing described in S3 includes: principal component analysis, covariance analysis, or luminance information statistics.
3. A panchromatic-multispectral remote sensing image fusion method according to claim 1 or 2, characterized in that, The brightness component in the multispectral image that can characterize the overall brightness information of the multispectral remote sensing image is: in These are the weighting coefficients, which can be obtained through linear regression, statistical estimation, or pre-setting. It is a multispectral remote sensing image with the same spatial resolution as the panchromatic remote sensing image.
4. The panchromatic-multispectral remote sensing image fusion method according to claim 3, characterized in that, Specifically, S4 involves performing illumination suppression processing based on multi-scale spatial filtering on the panchromatic remote sensing image to separate and suppress the illumination component in the panchromatic remote sensing image, obtain the separated reflection component in the panchromatic remote sensing image, and then obtain an enhanced panchromatic image representing spatial detail information based on the separated reflection component in the panchromatic remote sensing image.
5. The panchromatic-multispectral remote sensing image fusion method according to claim 4, characterized in that, The separated reflectance components in the panchromatic remote sensing image are represented as follows: in, This represents the reflectance component in panchromatic remote sensing images that contains structural and textural information. Represents panchromatic remote sensing imagery. This represents the illumination component in a panchromatic remote sensing image. The standard deviation is Gaussian wrapping function.
6. The panchromatic-multispectral remote sensing image fusion method according to claim 5, characterized in that, The enhanced panchromatic image representing spatial detail information is: in, This is an enhanced panchromatic image after illumination suppression. Indicates the scale quantity. Represents the Gaussian kernel. Represents convolution. This represents the standard deviation of the Gaussian function at different scales. Indicates weight, Indicates a stable term.
7. The panchromatic-multispectral remote sensing image fusion method according to claim 6, characterized in that, S5 includes: Luminance component of overall luminance information of multispectral remote sensing image Extracting the reflection component , making the reflected component Reflectance component in panchromatic remote sensing image Unify to the reflection component domain to complete alignment: Based on the same-domain reflection component and reflection component Construct an example modulation model: in, The modulation intensity coefficient, Take the modulation model as an example.
8. The panchromatic-multispectral remote sensing image fusion method according to claim 7, characterized in that, In S6, the spatial detail information contained in the enhanced panchromatic image is injected into each band of the multispectral remote sensing image in a multiplicative manner, as expressed by: in, For multispectral images The reflection component modulation factor after MSR-Retinex processing for each band.
9. A panchromatic-multispectral remote sensing image fusion system, characterized in that, The system includes a memory and a processor. The memory includes a panchromatic-multispectral remote sensing image fusion method program. When the processor executes the panchromatic-multispectral remote sensing image fusion method program, it implements the steps of a panchromatic-multispectral remote sensing image fusion method as described in any one of claims 1 to 6.
10. A computer-readable storage medium, characterized in that, The computer-readable storage medium includes a panchromatic-multispectral remote sensing image fusion method program, which, when executed by a processor, implements the steps of a panchromatic-multispectral remote sensing image fusion method as described in any one of claims 1 to 6.