Ecological and environmental remote sensing data acquisition methods applied to mine ecological restoration

Abstract

This invention relates to the field of image fusion technology, specifically to a method for acquiring ecological environment remote sensing data for mine ecological restoration. The method includes acquiring multispectral and panchromatic images and constructing a color histogram; determining color difference values ​​based on the difference in the number of pixels; dividing the image into sub-regions based on the color difference values; determining the color level discrimination in each sub-region based on the color difference values; determining the discrimination sharpness of each sub-region based on the difference in color level discrimination among all sub-regions in the multispectral image; further determining the contribution of texture details in the sub-regions to image fusion; using the contribution as the fusion weight for the multispectral image, and fusing the multispectral and panchromatic images to obtain a fused image. This invention, through sub-region division, achieves a more flexible and adaptive fusion process, providing a fused image that displays detailed ground features while also offering rich spectral information, fully preserving the detailed vegetation information in the mining area.

Description

Technical Field

[0001] This invention relates to the field of image fusion technology, and specifically to a method for acquiring ecological environment remote sensing data for mine ecological restoration. Background Technology

[0002] While coal mining provides energy, promotes economic development, and supports industrial production, it also disturbs the surface ecological environment of mining areas. The impact of coal mining on the natural environment cannot be ignored. However, with the improvement of overall environmental awareness and technological progress, mine ecological restoration has gradually become a feasible solution. Taking effective environmental governance and ecological restoration measures to reduce the long-term negative impact of mining on the environment is a necessary step to promote the sustainable development of the regional economy.

[0003] The application of remote sensing technology in the ecological restoration of mining areas is a highly effective method. By regularly collecting remote sensing data and extracting, analyzing, and detecting changes in the remote sensing characteristics of surface elements in mines, it is possible to accurately monitor changes in the ecological environment of mining areas and promptly understand the situation regarding vegetation restoration, soil quality, and soil erosion.

[0004] In related technologies, when fusing multispectral and panchromatic images, the panchromatic image is usually directly replaced with the first principal component image of the multispectral image. This preserves the high resolution of the panchromatic image and the color information of the multispectral image in the fused image. However, this approach fails to consider the complex details of different areas of the mine, making it difficult to retain sufficient detail information due to differences in image resolution. This results in the fused image being prone to texture loss. Summary of the Invention

[0005] To address the technical problem in related technologies that fail to consider the complex details of different areas in a mine, resulting in insufficient preservation of detail due to differences in image resolution and easy texture loss in fused images, this invention provides an ecological environment remote sensing data acquisition method for mine ecological restoration. The specific technical solution adopted is as follows:

[0006] This invention proposes a method for acquiring ecological environment remote sensing data for mine ecological restoration, the method comprising:

[0007] Acquire multispectral and panchromatic images of the mining area at different sampling times, and construct histograms of the RGB color channels at different color levels in the multispectral images;

[0008] In the histogram, the color difference value of each pixel is determined based on the difference in the number of pixels of different color channels and different color levels; the regions are divided according to the color difference value of each pixel to obtain different sub-regions;

[0009] In each sub-region, the color gradation discrimination of each sub-region is determined based on the color gradation range and the color difference value of the pixels; the discrimination sharpness of each sub-region is determined based on the difference in color gradation discrimination of all sub-regions in the multispectral image.

[0010] Based on the changes in the sharpness of the same sub-region at all sampling times, the contribution of the texture details of the sub-region to image fusion at the current sampling time is determined; the contribution is used as the fusion weight of the multispectral image, and the multispectral image and panchromatic image are fused to obtain the fused image.

[0011] Further, determining the color difference value of each pixel based on the difference in the number of pixels at different color levels in different color channels of each pixel includes:

[0012] The color level of any pixel in the RGB color channel is determined as the analysis color level; the absolute value of the numerical difference between the analysis color levels of the G channel and the other two channels is calculated and normalized to obtain the color level difference coefficient.

[0013] The number of pixels in each color channel of the histogram is used as the number of pixels analyzed; the absolute value of the difference between the number of pixels analyzed in the G channel and the other two channels is calculated, averaged and normalized to obtain the difference coefficient.

[0014] The product of the color level difference coefficient and the quantity difference coefficient is normalized and used as the color difference value of the pixel.

[0015] Furthermore, the process of dividing the region based on the color difference value of each pixel to obtain different sub-regions includes:

[0016] Based on the region growing algorithm, pixels with color difference values ​​less than a preset color difference threshold are divided into the same region to obtain different sub-regions.

[0017] Further, determining the color gradation discrimination of each sub-region based on the included color gradation range and the color difference value of the pixels includes:

[0018] Calculate the number of color levels of all pixels in all RGB color channels in each sub-region, and normalize it as a color level range coefficient.

[0019] The standard deviation of the color difference values ​​of all pixels in the sub-region is used as the color difference influence coefficient.

[0020] Calculate the product of the color gradation range coefficient and the color difference influence coefficient, and normalize the negative of the product value to obtain the color gradation discrimination.

[0021] Furthermore, determining the sharpness of each sub-region based on the difference in color gradation discrimination among all sub-regions in the multispectral image includes:

[0022] Calculate the mean of the absolute values ​​of the color gradation discrimination differences between any sub-region and every other sub-region, and use this as the discrimination clarity of the corresponding sub-region.

[0023] Furthermore, determining the contribution of the texture details of the sub-region to image fusion at the current sampling time based on the change in the sharpness of the same sub-region at all sampling times includes:

[0024] The first contribution index is determined based on the change in the recognition sharpness at all sampling times;

[0025] Calculate the difference in recognition sharpness between the current sampling time and the previous sampling time, and normalize it as the second contribution index;

[0026] By combining the first contribution index and the second contribution index, the contribution of sub-region texture details to image fusion at the current sampling time is determined.

[0027] Furthermore, based on the numerical change of the recognition sharpness at all sampling times, a first contribution index is determined, including:

[0028] Calculate the numerical range of the recognition sharpness at all sampling times, and normalize the numerical range as the contribution range coefficient;

[0029] Calculate the numerical standard deviation of the recognition sharpness at all sampling times, and use the mean of the numerical standard deviation and the contribution range coefficient as the first contribution index.

[0030] Furthermore, by combining the first contribution index and the second contribution index, the contribution of sub-region texture details to image fusion at the current sampling time is determined, including:

[0031] Calculate the product of the first contribution index and the second contribution index of the sub-region at the current sampling time, and normalize it to obtain the contribution degree.

[0032] Furthermore, the contribution is used as the fusion weight for the multispectral image, and image fusion is performed on the multispectral image and the panchromatic image to obtain a fused image, including:

[0033] Principal component images of multispectral images were determined based on principal component analysis.

[0034] The contribution of the sub-region is used as the fusion weight of the sub-region in the first principal component image, and the image is fused with the panchromatic image to obtain the fused target principal component image. The target principal component image is then subjected to inverse PCA transformation with other pre-preset number of principal component images to obtain the fused image.

[0035] Furthermore, the preset quantity is 3.

[0036] The present invention has the following beneficial effects:

[0037] In this embodiment of the invention, multispectral and panchromatic images are acquired, and then color difference analysis is performed to determine the color difference value. Based on the color difference value, regions are divided to obtain different sub-regions. Sub-region division is performed through color difference representation, thereby enabling adaptive image fusion of different sub-regions. This effectively takes into account the feature differences between vegetation areas and functional areas, avoiding the problem of blurred details caused by mismatch between different regions in unified image fusion. When performing specific fusion analysis on each sub-region, the color level discrimination of each sub-region is determined according to the color level range and the color difference value of the pixels. Based on the difference in color level discrimination of all sub-regions in the multispectral image, the discrimination clarity of each sub-region is determined. Combining the change in discrimination clarity, the contribution of the texture details of the sub-region to image fusion is determined, and the contribution is used as the fusion weight of the multispectral image. The multispectral and panchromatic images are then fused to obtain a fused image. This invention divides images into sub-regions and combines the color changes of the sub-regions to perform regional detail analysis, achieving a more flexible and adaptive fusion process. While displaying the detailed structure of ground features, it can also provide fused images with rich spectral information, fully preserving the detailed information of vegetation in the mining area. Attached Figure Description

[0038] To more clearly illustrate the technical solutions and advantages in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0039] Figure 1 A flowchart illustrating an ecological environment remote sensing data acquisition method for mine ecological restoration, provided in one embodiment of the present invention;

[0040] Figure 2 This is a schematic diagram of a multispectral image provided in one embodiment of the present invention;

[0041] Figure 3 This is a schematic diagram of a panchromatic image provided in one embodiment of the present invention;

[0042] Figure 4 This is a schematic diagram showing the comparison before and after partition fusion according to an embodiment of the present invention. Detailed Implementation

[0043] To further illustrate the technical means and effects adopted by the present invention to achieve its intended purpose, the following, in conjunction with the accompanying drawings and preferred embodiments, details the specific implementation, structure, features, and effects of an ecological environment remote sensing data acquisition method for mine ecological restoration proposed according to the present invention. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. Furthermore, specific features, structures, or characteristics in one or more embodiments can be combined in any suitable form.

[0044] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains.

[0045] The following description, in conjunction with the accompanying drawings, details a specific scheme for an ecological environment remote sensing data acquisition method for mine ecological restoration provided by the present invention.

[0046] Please see Figure 1 The diagram illustrates a flowchart of an ecological environment remote sensing data acquisition method for mine ecological restoration, provided by an embodiment of the present invention. The method includes:

[0047] S101: Acquire multispectral and panchromatic images of the mining area at different sampling times, and construct histograms of the RGB color channels at different color levels in the multispectral images.

[0048] While coal mining provides energy, promotes economic development, and supports industrial production, it also disturbs the surface ecological environment of mining areas. The impact of coal mining on the natural environment cannot be ignored. However, with the improvement of overall environmental awareness and technological progress, mine ecological restoration has gradually become a feasible solution. Taking effective environmental governance and ecological restoration measures to reduce the long-term negative impact of mining on the environment is a necessary step to promote the sustainable development of the regional economy.

[0049] The application of remote sensing technology in the ecological restoration of mining areas is a highly effective method. By regularly collecting remote sensing data and extracting, analyzing, and detecting changes in the remote sensing characteristics of surface elements in mines, it is possible to accurately monitor changes in the ecological environment of mining areas and promptly understand the situation regarding vegetation restoration, soil quality, and soil erosion.

[0050] In this embodiment of the invention, multispectral and panchromatic images of the mining area can be acquired via remote sensing satellites. The sampling date is set to the 15th of each month, with a sampling period of one year, to obtain several multispectral and panchromatic images from the historical period.

[0051] Panchromatic images are single-channel images, where panchromatic refers to the entire visible light band (0.38–0.76 μm). A panchromatic image is a mixed image within this band. Because it is single-band, it is displayed as a grayscale image. Panchromatic remote sensing images generally have high spatial resolution but cannot display the colors of ground features, meaning they have limited spectral information. In practice, panchromatic images need to be fused with multispectral imagery to obtain an image that combines the high resolution of the panchromatic image with the color information of the multispectral image.

[0052] In related technologies, when fusing multispectral and panchromatic images, the panchromatic image is usually directly replaced with the first principal component image of the multispectral image. This preserves the high resolution of the panchromatic image and the color information of the multispectral image in the fused image. However, this approach fails to consider the complex details of vegetation in mining areas, making it difficult to retain sufficient detail due to differences in image resolution. This results in texture loss in the fused image. This invention achieves a more flexible and adaptive fusion process by functionally dividing the image into regions and combining this with color changes in the sub-regions for regional detail analysis, thus fully preserving the detailed vegetation information in the mining area.

[0053] In this embodiment of the invention, a digital elevation model (DEM) and a mathematical model of satellite orbits are used to correct for the effects of atmospheric cloud interference, geometric distortion, and terrain undulations caused by the acquisition of original images. This ensures that the images have consistent spatial radiation and resolution, enabling effective fusion. See also Figure 2 and Figure 3 , Figure 2 This is a schematic diagram of a multispectral image provided in one embodiment of the present invention. Figure 3 This is a schematic diagram of a panchromatic image provided in one embodiment of the present invention.

[0054] The RGB color channels specifically consist of three different color channels: red, green, and blue. Histograms for each of these three color channels can be obtained. The color channel histograms reflect the distribution of pixels in the image under the current color component, which facilitates subsequent analysis.

[0055] S102: In the histogram, the color difference value of each pixel is determined based on the difference in the number of pixels of different color channels and different color levels of each pixel; the region is divided according to the color difference value of each pixel to obtain different sub-regions.

[0056] A complete mining area comprises multiple zones with different uses, such as mining sites, ore dressing plants, tailings ponds, construction areas, and ecological vegetation zones. Due to the influence of mining intensity and duration, the degree of ecological damage and restoration potential vary significantly across different locations within the mining area, which can be categorized into vegetated areas and functionally integrated areas. Therefore, this step utilizes the differences in pixel color information to reflect the boundaries and structural features of features within the mine, identifying sub-regions with different restoration objectives. Subsequently, based on this, the collected multi-source image data is further divided into regions to achieve ecosystem reconstruction and restoration, contributing to the improvement of the overall environmental quality of the mining area.

[0057] Vegetation cover significantly affects the ratio between the RGB channels of multispectral images. Ecologically vegetated areas appear as a dense green in images, while areas such as mines, ore processing plants, tailings ponds, and buildings, being rock or ore, appear gray or brown. That is, in vegetated areas, the green color channel value is significantly higher than the other two colors, and the gray-brown composite color scheme creates a clear difference in color gradation between the vegetated areas and the vegetation areas. In this embodiment of the invention, color difference values ​​are used for differentiation.

[0058] Furthermore, in some embodiments of the present invention, determining the color difference value of each pixel based on the difference in the number of pixels at different color levels in different color channels includes: determining the color level of any pixel in the RGB color channels as the analysis color level; calculating the absolute value of the numerical difference between the analysis color levels of the G channel and the other two channels, and normalizing it to obtain the color level difference coefficient; taking the number of pixels at the analysis color levels in each color channel in the histogram as the analysis quantity; calculating the absolute value of the difference between the analysis quantity of the G channel and the other two channels, and normalizing it to obtain the quantity difference coefficient; and normalizing the product of the color level difference coefficient and the quantity difference coefficient to obtain the color difference value of the pixel.

[0059] Each pixel has an analytical color level in the R, G, and B channels. The absolute value of the numerical difference between the analytical color levels in the G channel and the other two channels is calculated and normalized to obtain the color level difference coefficient.

[0060] In other words, first calculate the absolute value of the numerical difference between the analytical color levels of the G channel and the R channel, then calculate the absolute value of the numerical difference between the analytical color levels of the G channel and the B channel, and normalize the mean of the two absolute values ​​of numerical difference as the color level difference coefficient.

[0061] In one embodiment of the present invention, the normalization process can be specifically, for example, maximum and minimum value normalization. Furthermore, the normalization in subsequent steps can all adopt maximum and minimum value normalization. In other embodiments of the present invention, other normalization methods can be selected according to the specific range of the numerical values, which will not be elaborated further.

[0062] The higher the value of the color gradation difference coefficient, the more prominent the color in the G channel, meaning the more likely the corresponding pixel is to be a pixel in a vegetated area. By analyzing the color gradation differences in the R, G, and B channels, the influence of different forms of vegetation distribution can be avoided, thus accurately analyzing all vegetation-covered areas as a whole.

[0063] In particular, the vegetation coverage in mining areas is relatively high, so the proportion of pixels is relatively large. In this embodiment of the invention, the pixel proportion feature can be introduced for further analysis. The number of pixels in each color channel of the histogram is used as the number of pixels to be analyzed. The absolute value of the difference between the number of pixels in the G channel and the other two channels is calculated, averaged and normalized to obtain the number difference coefficient. The larger the number difference coefficient is, the more it conforms to the vegetation coverage feature.

[0064] In summary, the product of the color gradation difference coefficient and the quantity difference coefficient is normalized and used as the color difference value of the pixel. The color difference value represents the degree to which the color distribution of the pixel conforms to the vegetation cover characteristics.

[0065] Furthermore, in some embodiments of the present invention, the region is divided according to the color difference value of each pixel to obtain different sub-regions, including: based on a region growing algorithm, pixels with color difference values ​​less than a preset color difference threshold are divided into the same region to obtain different sub-regions.

[0066] The region growing algorithm is a well-known algorithm to those skilled in the art, and will not be described in detail here. In this embodiment of the invention, the region growing factor is the color difference value. Pixels with color difference values ​​less than a preset color difference threshold are regarded as the same region to achieve sub-region division.

[0067] The preset color difference threshold is a threshold value for the color difference value when performing region growth. Specifically, it can be, for example, 0.2, which means that adjacent pixels with a color difference value less than 0.2 are divided into the same region to obtain different sub-regions.

[0068] S103: In each sub-region, determine the color gradation discrimination of each sub-region based on the color gradation range and the color difference value of the pixels; determine the discrimination clarity of each sub-region based on the difference in color gradation discrimination of all sub-regions in the multispectral image.

[0069] The color gradation range and color difference value distribution of each sub-region can reflect the color recognition of the sub-region itself and the detailed recognition of the texture within the sub-region. In this embodiment of the invention, color gradation recognition is used to characterize the recognition.

[0070] Based on the included color gradation range and the color difference value of the pixels, the color gradation discrimination of each sub-region is determined, including: calculating the number of color gradations of all pixels in all RGB color channels in each sub-region, normalizing it as the color gradation range coefficient; using the standard deviation of the color difference value of all pixels in the sub-region as the color difference influence coefficient; calculating the product of the color gradation range coefficient and the color difference influence coefficient, normalizing the negative of the product value, and using it as the color gradation discrimination.

[0071] It is understandable that by measuring the color gradation range of a sub-region and the influence of color difference values ​​in each channel of the sub-region, the color characteristics of the current sub-region are reflected. The smaller the color distribution range in the sub-region and the smaller the difference between color channels in the sub-region, the more concentrated the color distribution in the current sub-region is. Consequently, the details in the corresponding sub-region will be blurred due to the concentrated color distribution, which means that the color gradation discrimination value is smaller.

[0072] Since all sub-regions of the entire multispectral image need to be considered in the actual processing, this embodiment of the invention also combines the color gradation discrimination of all sub-regions in the multispectral image to determine the discrimination clarity of each sub-region.

[0073] Furthermore, in some other embodiments of the present invention, determining the sharpness of each sub-region based on the difference in color gradation discrimination of all sub-regions in the multispectral image includes: calculating the mean of the absolute values ​​of the differences in color gradation discrimination between any sub-region and every other sub-region as the sharpness of the corresponding sub-region.

[0074] In this embodiment of the invention, firstly, any sub-region can be selected as the target region, and the absolute value of the color level recognition difference between the target region and each other sub-region is calculated as the recognition difference. The larger the value of the recognition difference, the greater the difference in color distribution concentration between the target region and the corresponding other sub-regions. Then, the mean of all the obtained recognition differences is normalized as the recognition clarity.

[0075] A higher resolution value indicates that the color performance of the target area differs more significantly from that of other sub-regions in the entire multispectral image, and the more attention needs to be paid to the target area to improve its texture details.

[0076] S104: Based on the changes in the sharpness of the same sub-region at all sampling times, determine the contribution of the texture details of the sub-region to image fusion at the current sampling time; use the contribution as the fusion weight of the multispectral image, perform image fusion on the multispectral image and the panchromatic image, and obtain the fused image.

[0077] For different sub-regions, regional ecological restoration can be divided into vegetated areas and functionally integrated areas based on regional environmental conditions and hydrological information. For vegetated areas with extensive vegetation cover, the assessment of ecological restoration effects depends on the spectral characteristics of the vegetation. The regional fusion image needs to better represent the vegetation characteristics to ensure that the growth status of the vegetation and the ecological restoration effect can be determined. Therefore, it is necessary to combine more multispectral image information. For areas such as mining sites, ore processing plants, tailings ponds, and buildings, due to their weak ecological restoration capacity, they will be classified as functionally integrated areas, such as landscape parks, or directly consolidated and maintained in situ. Therefore, the fusion image needs to better represent the detailed features of regional features to identify and detect the boundary information of regional topography and features. Thus, it is necessary to increase the detail information of panchromatic images when fusion images are used.

[0078] Understandably, with the changing seasons, the vegetation status and ground features of different sub-regions within a mine will undergo significant changes. In areas with vigorous vegetation growth, the dense vegetation cover during the revegetation stage will significantly obscure details and textures in the images. Therefore, the proportion of multispectral color information needs to be increased during image fusion. For functional areas with degraded or bare vegetation, the vegetation cover is significantly low. When assessing ecological restoration, it is necessary to analyze more landscape texture information in the images. Therefore, it is necessary to increase the proportion of panchromatic image detail information during the fusion process. The contribution rate represents the proportion of multispectral imagery in the fusion process; the more pronounced the vegetation cover, the greater the need to increase its contribution rate.

[0079] Therefore, in this embodiment of the invention, the contribution of the texture details of the sub-region to image fusion at the current sampling time is determined by combining the numerical changes of the vegetation region in the recognition of emotion. This includes: determining a first contribution index based on the numerical changes of the recognition sharpness at all sampling times; calculating the difference in recognition sharpness between the current sampling time and the previous sampling time, and normalizing it as a second contribution index; and combining the first contribution index and the second contribution index to determine the contribution of the texture details of the sub-region to image fusion at the current sampling time.

[0080] The vegetation area changes color with the four seasons. For example, the leaves turn yellow in winter and the whole area turns yellowish-brown. In spring and summer, the vegetation recovers and the leaves sprout and turn green again. Therefore, based on the changes under the characteristics of time, it is easy to divide the vegetation area and the functional area.

[0081] Furthermore, in some embodiments of the present invention, determining a first contribution index based on the numerical change of the identification sharpness at all sampling times includes: calculating the numerical range of the identification sharpness at all sampling times, normalizing the numerical range as a contribution range coefficient; calculating the numerical standard deviation of the identification sharpness at all sampling times, and taking the mean of the numerical standard deviation and the contribution range coefficient as the first contribution index.

[0082] The larger the numerical range and the larger the numerical standard deviation, the greater the change in the clarity of identification at all sampling times within the period, the more consistent it is with the periodic characteristics of tree growth and recovery, and the more necessary it is to increase the proportion of multispectral color information.

[0083] After analyzing all sampling times, it is also necessary to conduct a specific analysis in conjunction with the real-time characteristics of the current sampling time. In this embodiment of the invention, the difference in recognition clarity between the current sampling time and the previous sampling time is calculated and normalized as a second contribution index. The larger the value of the second contribution index, the more obvious the change at the current sampling time is, and the more it is in the change characteristics under vegetation restoration. Therefore, it is necessary to increase the proportion of multispectral color information to improve the details and texture of the image.

[0084] In summary, by combining the first contribution index and the second contribution index, the contribution of the sub-region texture details to image fusion at the current sampling time is determined, including: calculating the product of the first contribution index and the second contribution index of the sub-region at the current sampling time, and normalizing it as the contribution.

[0085] Since the larger the values ​​of the first and second contribution indices, the greater the need to enhance the proportion of multispectral color information, the contribution degree is obtained by fusion to characterize the fusion ratio of multispectral images, thereby achieving adaptive fusion effect of images in different sub-regions.

[0086] Using the contribution as the fusion weight of the multispectral image, the multispectral image and the panchromatic image are fused to obtain a fused image. This includes: determining the principal component images of the multispectral image based on principal component analysis; using the contribution of the sub-region as the fusion weight of the sub-region in the first principal component image and fusing it with the panchromatic image to obtain the fused target principal component image; and performing inverse PCA transformation on the target principal component image and other pre-preset number of principal component images to obtain the fused image.

[0087] It should be noted that the fusion of multispectral and panchromatic images using principal component analysis is a well-known fusion technique among those skilled in the art. In related techniques, the panchromatic image and the first principal component image are directly replaced, and then inverse PCA transformation is performed to obtain the fused image. The present invention aims to use the contribution of the sub-region as the fusion weight of the sub-region in the first principal component image, so as to achieve adaptive image fusion of the first principal component image and the panchromatic image in different sub-regions, in order to retain more color detail information of the vegetation cover area.

[0088] In this embodiment of the invention, the preset number is specifically 3. That is, during the fusion process, the target principal component image of the first principal component image is first determined. Then, the target principal component image, the second principal component image, and the third principal component image are combined to perform an inverse PCA transformation to obtain the fused image. See [link to documentation]. Figure 4 , Figure 4 This is a schematic diagram showing the comparison before and after partition fusion according to an embodiment of the present invention.

[0089] In this embodiment of the invention, multispectral and panchromatic images are acquired, and then color difference analysis is performed to determine the color difference value. Based on the color difference value, regions are divided to obtain different sub-regions. Sub-region division is performed through color difference representation, thereby enabling adaptive image fusion of different sub-regions. This effectively takes into account the feature differences between vegetation areas and functional areas, avoiding the problem of blurred details caused by mismatch between different regions in unified image fusion. When performing specific fusion analysis on each sub-region, the color level discrimination of each sub-region is determined according to the color level range and the color difference value of the pixels. Based on the difference in color level discrimination of all sub-regions in the multispectral image, the discrimination clarity of each sub-region is determined. Combining the change in discrimination clarity, the contribution of the texture details of the sub-region to image fusion is determined, and the contribution is used as the fusion weight of the multispectral image. The multispectral and panchromatic images are then fused to obtain a fused image. This invention divides images into sub-regions and combines the color changes of the sub-regions to perform regional detail analysis, achieving a more flexible and adaptive fusion process. It fully preserves the detailed information of vegetation in the mining area, and provides a fused image with rich spectral information while displaying the detailed structure of land features.

[0090] It should be noted that the order of the above embodiments of the present invention is merely for descriptive purposes and does not represent the superiority or inferiority of the embodiments. The processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired result. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0091] The various embodiments in this specification are described in a progressive manner. The same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on describing the differences from other embodiments.

Claims

1. An ecological environment remote sensing data collection method applied to mine ecological restoration, characterized in that, The method comprises: acquiring multispectral images and panchromatic images of a mine area at different sampling times, and constructing a histogram of RGB color channels in the multispectral images at different color scales; in the histogram, determining a color difference value of each pixel point according to the number difference of pixel points of each pixel point at different color scales of different color channels; and performing regional division according to the color difference value of each pixel point to obtain different sub-regions; in each sub-region, determining a color scale recognition degree of each sub-region according to the color scale range and the color difference value of the pixel points; and determining the recognition clarity of each sub-region according to the color scale recognition degree difference of all sub-regions in the multispectral images; determining the contribution degree of the texture details of the sub-region to image fusion at the current sampling time according to the recognition clarity change of the same sub-region at all sampling times; and taking the contribution degree as a fusion weight of the multispectral images to perform image fusion on the multispectral images and the panchromatic images, and obtaining a fused image.

2. The ecological environment remote sensing data collection method applied to mine ecological restoration according to claim 1, characterized in that, The method comprises: determining a color difference value of each pixel point according to the number difference of pixel points of each pixel point at different color scales of different color channels; and performing regional division according to the color difference value of each pixel point to obtain different sub-regions. The method comprises: determining a color difference value of each pixel point according to the number difference of pixel points of each pixel point at different color scales of different color channels; and performing regional division according to the color difference value of each pixel point to obtain different sub-regions.

3. The ecological environment remote sensing data collection method for ecological restoration of a mine according to claim 1, characterized in that, The method comprises: determining a color difference value of each pixel point according to the number difference of pixel points of each pixel point at different color scales of different color channels; and performing regional division according to the color difference value of each pixel point to obtain different sub-regions.

4. The ecological environment remote sensing data collection method for ecological restoration of a mine according to claim 1, characterized in that, The method comprises: determining a color difference value of each pixel point according to the number difference of pixel points of each pixel point at different color scales of different color channels; and performing regional division according to the color difference value of each pixel point to obtain different sub-regions. The method comprises: determining a color difference value of each pixel point according to the number difference of pixel points of each pixel point at different color scales of different color channels; and performing regional division according to the color difference value of each pixel point to obtain different sub-regions.

5. The ecological environment remote sensing data collection method for ecological restoration of a mine according to claim 1, characterized in that, The method comprises: determining a color difference value of each pixel point according to the number difference of pixel points of each pixel point at different color scales of different color channels; and performing regional division according to the color difference value of each pixel point to obtain different sub-regions.

6. The ecological environment remote sensing data collection method for ecological restoration of a mine according to claim 1, characterized in that, The method comprises: determining a color difference value of each pixel point according to the number difference of pixel points of each pixel point at different color scales of different color channels; and performing regional division according to the color difference value of each pixel point to obtain different sub-regions. The method comprises: determining a color difference value of each pixel point according to the number difference of pixel points of each pixel point at different color scales of different color channels; and performing regional division according to the color difference value of each pixel point to obtain different sub-regions. The method comprises: determining a color difference value of each pixel point according to the number difference of pixel points of each pixel point at different color scales of different color channels; and performing regional division according to the color difference value of each pixel point to obtain different sub-regions. The method comprises: determining a color difference value of each pixel point according to the number difference of pixel points of each pixel point at different color scales of different color channels; and performing regional division according to the color difference value of each pixel point to obtain different sub-regions. The method comprises: determining a color difference value of each pixel point according to the number difference of pixel points of each pixel point at different color scales of different color channels; and performing regional division according to the color difference value of each pixel point to obtain different sub-regions. The method comprises: determining a color difference value of each pixel point according to the number difference of pixel points of each pixel point at different color scales of different color channels; and performing regional division according to the color difference value of each pixel point to obtain different sub-regions. The method comprises: determining a color difference value of each pixel point according to the number difference of pixel points of each pixel point at different color scales of different color channels; and performing regional division according to the color difference value of each pixel point to obtain different sub-regions.

7. The ecological environment remote sensing data collection method for ecological restoration of a mine according to claim 6, characterized in that, According to the value change of the identification clarity at all sampling time points, a first contribution index is determined, including: calculating the numerical range of the identification clarity at all sampling time points, and normalizing the numerical range as a contribution range coefficient; calculating the numerical standard deviation of the identification clarity at all sampling time points, and taking the average of the numerical standard deviation and the contribution range coefficient as the first contribution index.

8. The ecological environment remote sensing data collection method for ecological restoration of a mine according to claim 6, characterized in that, In combination with the first contribution index and the second contribution index, the contribution degree of the sub-region texture details to the image fusion at the current sampling time point is determined, including: calculating the product value of the first contribution index and the second contribution index of the sub-region at the current sampling time point, and normalizing the product value as the contribution degree.

9. The ecological environment remote sensing data collection method for ecological restoration of a mine according to claim 1, characterized in that, Taking the contribution degree as the fusion weight of the multispectral image, the multispectral image and the panchromatic image are subjected to image fusion to obtain a fused image, including: determining principal component images of the multispectral image based on principal component analysis; taking the contribution degree of the sub-region as the fusion weight of the sub-region in the first principal component image, and performing image fusion with the panchromatic image to obtain a target principal component image after fusion, and performing inverse PCA transformation on the target principal component image and other pre-set number of principal component images to obtain a fused image.

10. The ecological environment remote sensing data collection method for ecological restoration of a mine according to claim 9, characterized in that, The pre-set number is 3.

Images