A multi-source ecological remote sensing data fusion method and system

Abstract

The application relates to the technical field of image processing, in particular to a multi-source ecological remote sensing data fusion method and system. The method comprises the following steps: collecting multi-band images and panchromatic images; obtaining a closed operation image after edge detection and closed operation of the images, and dividing the images into a plurality of partitions based on the closed operation image; respectively performing interpolation on different layer images of the panchromatic images, and correcting the interpolation based on the gradient deflection direction to obtain interpolation correction values; obtaining principal component images of the panchromatic images after the interpolation is completed, performing histogram matching on the principal component images and the panchromatic images, and adjusting the matching values based on the gray scale change and the quantity change to obtain matching adjustment values; and comparing the partition images after the matching with the multi-band images and the panchromatic images to obtain fusion values of edge pixel points, and obtaining a fusion image based on the fusion values. The application improves the definition and color performance of the fusion image.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of image processing, in particular to a multi-source ecological remote sensing data fusion method and system. BACKGROUND

[0002] The process of urbanization and global warming has led to an increasing number of ecological damage events, affecting the atmosphere, vegetation, soil and other aspects. In addition to focusing on economic development, the state is gradually increasing its requirements for ecological environment. Monitoring of the ecological environment can effectively determine the effectiveness of measures, and remote sensing images can observe the ecological environment from a global perspective. Panchromatic band images are grayscale images, but have high spatial resolution. Multi-band images can obtain color images, but have lower spatial resolution. Therefore, in ecological monitoring, single-band images and multi-band images are often fused to retain the advantages of both. The principal components of the multi-spectral image are extracted by PCA transformation, the grayscale distribution histograms of the panchromatic image and the principal components of the multi-spectral image are matched, and the principal components are replaced with the panchromatic image. The multi-spectral image is reconstructed by inverse PCA transformation, and the reconstructed image contains both the detail information of the panchromatic image and the color information. However, due to the difference in resolution between the two images, the multi-spectral image shows weak details and textures of the ecological environment, resulting in loss of image details during the resampling process and histogram matching process, which further leads to poor quality of the fused image and deviation in the subsequent analysis of the ecological environment by the staff. SUMMARY

[0003] To solve the technical problem of poor quality of the fused image caused by matching loss of details, the present application provides a multi-source ecological remote sensing data fusion method and system, and the technical solution adopted is as follows:

[0004] In a first aspect, the present application provides a multi-source ecological remote sensing data fusion method, which comprises the following steps:

[0005] Using a spectrometer to collect multi-band images and panchromatic images in an ecological region;

[0006] Using an edge detection algorithm on the multi-band images and the panchromatic images, taking the intersection and performing a closing operation to obtain a closing operation image; setting a preset window in the closing operation image, and dividing the ecological region to obtain a plurality of partitions based on the difference in the increase of edge pixel points after the closing operation in adjacent windows and merging the windows;

[0007] For the panchromatic image, interpolate the matrix of each partition in each layer image; set the direction of the interpolation point, obtain the interpolation bias according to the difference between the pixel value in each direction and the interpolation of the interpolation point; obtain the interpolation correction value based on the interpolation correction of the adjacent gray value difference in the interpolation bias direction and the interpolation bias; after interpolating all images of the panchromatic image based on the interpolation correction value, perform PCA transformation to obtain the principal component image; based on the principal component image, histogram matching is performed on the panchromatic image to obtain the matching value of each gray value; based on the change of the number of gray levels after matching and the change of the number of pixel points of each gray level, the blur degree of each partition is calculated; based on the continuous range of the gray level in the histogram, the matching adjustment value of each gray value is obtained based on the gray level range and the blur degree of the principal component image and the panchromatic image.

[0008] Based on the matching adjustment value, the fusion image of each partition is obtained, and based on the gray value difference between the fusion image and the multi-band image and the panchromatic image, the weight of the partition where each edge pixel point is located is obtained; based on the weight, the fusion value is obtained by weighting the plurality of matching adjustment values of the edge pixel point; the fusion value is replaced by the edge pixel point to obtain the final fusion image.

[0009] In the above scheme, the present application aims at the problem that when the existing PCA transformation is used for multi-band image and panchromatic image fusion, the difference in resolution of the images leads to the loss of details of the fusion image, resulting in insufficient quality of the fusion image. The multi-band image and the panchromatic image are analyzed by partition, and the interpolation process and the adaptive adjustment of the histogram matching process are performed respectively, so that each partition can retain the ecological details collected by the panchromatic image, and the splicing of each partition is completed, which is beneficial to improve the clarity and color performance of the fusion image.

[0010] In one embodiment, the method for obtaining the closed operation image by taking the intersection of the edge detection algorithm of the multi-band image and the panchromatic image and performing closed operation is as follows:

[0011] The edge intersection pixel points of the multi-band image and the panchromatic image are obtained by using the edge detection algorithm, all the edge intersection pixel points constitute an edge intersection image, and the edge intersection image is closed operated to obtain a closed operation edge intersection image, which is denoted as a closed operation image.

[0012] In one embodiment, the method for obtaining a plurality of partitions by partitioning the ecological region based on the difference in the increase of the closed operation edge pixel points in the adjacent window and the window merging is as follows:

[0013] m10 represents the number of pixels in the first window, m20 represents the number of pixels in the second window, m1 represents the number of edge pixels added to the first window, m2 represents the number of edge pixels added to the second window, norm() represents the linear normalization function, and P represents the merged value.

[0014] When the merge value is greater than the preset threshold, the windows are merged; otherwise, the windows are not merged. After merging all windows, several partitions are obtained.

[0015] In one embodiment, the method for obtaining the interpolation bias based on the range of the difference between the grayscale value of a pixel and the interpolation point in each direction is as follows:

[0016] c ′ c represents the interpolation at the interpolation point. i,j Let represent the grayscale value of the j-th pixel in the i-th direction, max represents the maximum value function, min represents the minimum value function, norm[] represents the linear normalization function, and α represents the interpolation bias.

[0017] In one embodiment, the method for obtaining the interpolation correction value based on the difference between adjacent gray values ​​in the interpolation bias direction and the interpolation bias is as follows:

[0018] The direction corresponding to the interpolation bias is denoted as the deviation direction. The gray value difference between adjacent pixels along the deviation direction is calculated, and the maximum value of the gray value difference is used as the interpolation adjustment value. The expression for the interpolation correction value is:

[0019] c″=c ′ +(c1-c ′ )×α,c″ are the interpolation correction values ​​at the interpolation points; c ′ c1 is the interpolation value at the interpolation point; c1 is the interpolation adjustment value in the deviation direction; and α is the interpolation bias.

[0020] In one embodiment, the blurriness of each partition is positively correlated with the change in gray level and positively correlated with the change in the number of pixels at the same gray level.

[0021] In one embodiment, the matching adjustment value is positively correlated with the blur level of the principal component image, the matching value of the panchromatic image, the maximum difference in the grayscale concentration range of the panchromatic image, and the maximum difference in the grayscale concentration range of the principal component image.

[0022] In one embodiment, the method for obtaining the weight of the partition where each edge pixel is located based on the difference in grayscale values ​​between the fused image and the multi-band image and panchromatic image is as follows:

[0023] Select an edge pixel as the target point, and denote the pixels on its normal as the normal points; the weight expression is:

[0024] h s denotes the gray value of the target point in the s-th partition of the fused image, h a denotes the gray value of the a-th normal point in the fused image, h s1 denotes the gray value of the target point in the s-th partition of the multi-band image, h 1,a denotes the gray value of the a-th normal point in the multi-band image, h s2 denotes the gray value of the target point in the s-th partition of the panchromatic image, h 1,a denotes the gray value of the a-th normal point in the panchromatic image, A denotes the number of normal points in the normal line, exp[] denotes the exponential function with the natural constant as the base, K s denotes the weight of the target point in the s-th partition in the fused image, the gray value of the target point in each partition is the matching adjustment value thereof in each partition.

[0025] In an embodiment, the method for obtaining the fusion value by weighting the plurality of matching adjustment values of the edge pixel point based on the weight is as follows:

[0026] For the plurality of gray values of the target point, let the weight of each partition be the ratio of the weight of each partition to the sum of the weights of all partitions, and let the ratio be multiplied by each gray value and accumulated, and the obtained value is the fusion value of the target point.

[0027] On the other hand, the embodiment of the present application also provides a multi-source ecological remote sensing data fusion system, which comprises a memory, a processor, and a computer program stored in the memory and running on the processor, and the processor implements the steps of the multi-source ecological remote sensing data fusion method according to any one of the above embodiments when executing the computer program.

[0028] The present application has the following beneficial effects:

[0029] The present application is aimed at the problem that the details of the fused image are lost due to the resolution difference of the images when the multi-band image and the panchromatic image are fused by PCA transformation, and the quality of the fused image is insufficient. The present application performs partition analysis on the multi-band image and the panchromatic image, and respectively performs adaptive adjustment of the interpolation process and the histogram matching process, so that each partition can retain the ecological details collected by the panchromatic image, and the splicing of each partition is completed, which is beneficial to improving the definition and color performance of the fused image. BRIEF DESCRIPTION OF DRAWINGS

[0030] In order to more clearly illustrate the technical solutions and advantages of the embodiments of the present application or the prior art, the drawings needed to be used in the description of the embodiments or the prior art will be briefly introduced. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor.

[0031] Figure 1 A multi-source ecological remote sensing data fusion method flow chart provided by an embodiment of the present application;

[0032] Figure 2 A band edge map;

[0033] Figure 3 A panchromatic edge map;

[0034] Figure 4 A closing operation image;

[0035] Figure 5 A gray level histogram of a corresponding partition of a multi-band image;

[0036] Figure 6 A gray level histogram of a corresponding partition of a panchromatic image. DETAILED DESCRIPTION

[0037] In order to further illustrate the technical means and effects adopted by the present application to achieve the predetermined purpose of the application, the specific implementation, structure, features and effects of the multi-source ecological remote sensing data fusion method and system according to the present application are described in detail as follows. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. In addition, the specific features, structures or characteristics in one or more embodiments can be combined in any suitable form.

[0038] 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 the present application belongs.

[0039] A multi-source ecological remote sensing data fusion method and system embodiment:

[0040] The specific scheme of the multi-source ecological remote sensing data fusion method and system provided by the present application will be described in detail below with reference to the drawings.

[0041] Please refer to Figure 1 , which shows a multi-source ecological remote sensing data fusion method flow chart provided by an embodiment of the present application. The method comprises the following steps:

[0042] Step S001, multi-band images and panchromatic images are collected.

[0043] The multi-band image and the panchromatic image of the required ecological region are acquired by MODIS (Moderate Resolution Imaging Spectroradiometer), and the multi-band image and the panchromatic image are geometrically corrected to ensure the spatial position consistency. The spatial resolution of the multi-band image is relatively low, but the multi-band image contains color information, and the spatial resolution of the panchromatic image is high, but the panchromatic image does not contain color information.

[0044] At this point, the panchromatic image and the multi-band image are acquired.

[0045] In step S002, the closed operation image is acquired by performing edge detection on the image and closed operation, and the image is divided into a plurality of partitions based on the closed operation image.

[0046] Since the size of the multi-band image and the panchromatic image is inconsistent, the multi-band image needs to be resampled to reach the size of the panchromatic image before fusion. However, in this process, the multi-band image is already blurred, and interpolation based on the blurred image cannot increase the detail presentation, but rather weakens the gradient change of some edge pixel points. Then, in the histogram matching process, since it is a global matching, the process adjusts the gray scale of all pixel points as a whole, which may cause local overexposure / underexposure. Therefore, the ecological region of the collected data is first divided, and the ecological region is the content collected by the panchromatic image and the multi-band image.

[0047] One ecological region may contain various ecological environments, such as water, vegetation, and buildings, and the main color of each environment is different. Therefore, the region can be divided based on color distribution combined with edge detection. In addition, for a region with complex details and textures, the region will be divided into multiple regions by color division because the color performance is diverse and the edges are also multiple. However, the overall structure of this region is relatively similar, and therefore the region can be regarded as a whole.

[0048] Specifically, edge detection is performed on the multi-band image and the panchromatic image respectively to obtain a band edge image and a panchromatic edge image, which are represented by Figure 2 and Figure 3 respectively. In this embodiment, the canny edge detection algorithm is used. The common edges represent the ecological region that can be divided in the multi-band image and the panchromatic image. Therefore, the intersection pixel points of the edges of the multi-band image and the panchromatic image are acquired, and all the intersection pixel points constitute an edge intersection image. In order to ensure that the region surrounded by the edge is closed, closed operation is performed on the edge intersection image to obtain a closed operation edge intersection image, which is denoted as a closed operation image, as shown in Figure 4 .

[0049] For the closed operation edge intersection image, the region with a high density of edge pixel points will increase the pixel points, as shown in the lower right corner of Figure 4 .

[0050] A window is set in the closed operation image, and each window is set with each pixel point as the center. In this embodiment, the window size is 15*15. If the window size of a certain pixel point is insufficient, the existing size is used.

[0051] For two adjacent windows, whether to merge the windows is determined based on the number difference of the edge pixel points of the two windows. The windows are adjacent in the center pixel points. In this embodiment, the expression of window merging is:

[0052] m10 represents the number of pixel points in the first window, m20 represents the number of pixel points in the second window, m1 represents the number of edge pixel points added in the first window, m2 represents the number of edge pixel points added in the second window, norm() represents a linear normalization function, and P represents a merging value.

[0053] When the merging value is greater than a threshold Q, the two windows are merged, and the edge pixel points of the merged window are removed. In this embodiment, the threshold Q is 0.5. In a region with complex details, a large number of pixel points are added due to local edge density during the closing operation, but these added edges may belong to the internal details of the same structure. At this time, the difference is large (P is high), indicating that the two windows belong to the same complex whole and need to be merged to preserve the structural integrity and avoid excessive segmentation. If only the edge intersection is relied on, the dense region will be divided into many meaningless small regions. After merging, the integrity of the ecological unit can be restored. Therefore, the edge pixel points are removed to eliminate the interference of local details on the region division. After merging, the sub-regions divided by the dense edges will be as a whole, which is more consistent with the actual ecological structure.

[0054] All windows are merged by the above-mentioned manner until the windows are no longer merged. The final all windows are used as several partitions of the ecological region.

[0055] Thus, the multiple partitions of the ecological region are obtained.

[0056] In step S003, different layer images of the panchromatic image are respectively interpolated, and interpolation correction values are obtained by correcting the interpolation based on the gradient direction. After the interpolation is completed, the principal component image of the panchromatic image is obtained. The principal component image and the panchromatic image are histogram matched, and matching adjustment values are obtained by adjusting the matching values based on the gray scale change and the number change.

[0057] In the interpolation process, since the linear interpolation does not consider the numerical transformation of the neighborhood, the gradient of the adjacent pixel points of the gray scale mutation is weakened, for example (2 2 8 16), the linear interpolation of which reduces the gray scale change, for example, the linear interpolation between 2 and 8 is 5; at this time, 5, 8, 16 reduces the gray scale transformation, and the interpolation value should be more biased to the side with larger numerical transformation, therefore, the method adjusts the interpolation process by referring to the numerical change of the neighborhood of the interpolation.

[0058] Firstly, for each partition of the panchromatic image, each partition has three matrices, respectively corresponding to R matrix, G matrix, and B matrix, wherein the value in each matrix is the R value, G value and B value of the partition under the corresponding layer. The R value, G value and B value of the non-partition are recorded as 0.

[0059] For each matrix corresponding to each partition, the rows and columns of each matrix of each partition are expanded by two times through the bilinear interpolation. Since the interpolated value of the bilinear interpolation is in the middle of the four existing points in the matrix, in the original bilinear interpolation, the interpolation is the average of the adjacent four values in the matrix, but using the average to expand the matrix will weaken the gradient of the R value change of the pixel, therefore, the interpolated value should be adjusted by referring to the gradient change.

[0060] Taking the interpolation point as the initial point, the four directions above, below, left and right of the interpolation point are obtained, for example: taking the interpolation point as the initial point, a ray is emitted to the right of the initial point as the right direction. Since the greater the gray scale change is, the more significant the edge details are, in order not to reduce the gray scale change, the interpolation should be corrected in the direction where the edge details are more significant. Therefore, the interpolation bias of the interpolation point is determined according to the gradient change in each direction.

[0061] Preferably, in the embodiment, the difference between the gray scale value of each direction of the pixel point and the interpolation of the interpolation point is calculated; the interpolation bias of the interpolation point is obtained according to the range of the difference value in each direction. The expression of the interpolation bias is:

[0062] c ′ represents the interpolation of the interpolation point, c i,j represents the gray scale value of the jth pixel point in the ith direction, max represents the maximum value function, min represents the minimum value function, norm[] represents the linear normalization function, and a represents the interpolation bias. represents the gray scale change in any direction, the edge details bias can be obtained by subtracting the minimum value from the maximum value, and then the interpolation bias of the interpolation point is obtained.

[0063] The greater the interpolation bias of the interpolation point, the greater the degree of correction of the interpolation to the edge detail significant direction. The corresponding direction of the interpolation bias is recorded as a deviation direction, the gray value difference of adjacent pixels in the deviation direction is calculated, the maximum value of the gray value difference is taken as an interpolation adjustment value, and the interpolation is adjusted based on the interpolation adjustment value and the interpolation bias to obtain an interpolation correction value. The expression of the interpolation correction value is:

[0064] c" = c ′ + (c1-c ′ ) x a, c" is the interpolation correction value of the interpolation point; c ′ is the interpolation of the interpolation point; c1 is the interpolation adjustment value of the deviation direction, and a is the interpolation bias.

[0065] Based on the above method, the interpolation correction value is obtained by adjusting all interpolations. For any partition, the enlarged R matrix, G matrix, and B matrix are obtained.

[0066] After completing the interpolation of the panchromatic image, the resolution of the panchromatic image is the same as that of the multi-band image, and the pixel points have a one-to-one correspondence. The principal components are obtained by performing PCA transformation on the matrix of any partition of the multi-band image, and the matrix of the partition is projected into the principal component space to obtain the PC1 matrix, that is, the principal component image corresponding to the partition.

[0067] Further, the histogram is matched; for any partition, the gray mean and variance of the principal component image are obtained, the gray mean and variance of the panchromatic image are obtained, and the gray value of the panchromatic image is matched and calculated based on the gray mean and variance of the principal component image and the panchromatic image to obtain the matching value of the panchromatic image, and the expression is:

[0068] μ1 is the gray mean of the principal component image, σ1 is the variance of the principal component image; μ2 is the gray mean of the panchromatic image, σ2 is the variance of the panchromatic image, q represents any gray value of the panchromatic image, and q ′ represents the matching value of the gray value q.

[0069] In the original process, the histogram of the entire image is matched. Since the panchromatic image has uneven illumination, global matching will cause all ecological environment regions to follow the statistical distribution of the principal component, which will cause overexposure or underexposure in the shadow area. Therefore, partition is adopted, each region has different color tones, and global over-adjustment can be avoided. However, since the principal component image comes from the multi-band image with low resolution, some regions are relatively blurred, for example Figure 5 is the gray histogram of a partition of the multi-band image, Figure 6is the gray level histogram of the corresponding partition in the panchromatic image. Since the gray level range of this region in the multi-band image is smaller, and the gray level range of the panchromatic image is larger, after the histogram matching, the image will become dark and the texture will be blurred. Therefore, the blur degree of the principal component image of any partition is calculated first.

[0070] If the gray level changes more between different images, and the number of pixels of each gray level changes more, it means that after the histogram matching, the image will become blurred, i.e. the blur degree is larger. Therefore, the blur degree of the principal component image is positively correlated with the gray level change, and is positively correlated with the number of pixels of the same gray level.

[0071] It should be noted that the positive correlation means that when one variable increases, the other variable also increases, and the two variables change in the same direction. When one variable changes from large to small or from small to large, the other variable also changes from large to small or from small to large. The specific relationship is determined by actual application, and the present application does not make special limitations.

[0072] Preferably, in the present embodiment, the expression of the blur degree of each partition of the principal component image is:

[0073] m10 represents the number of non-zero gray levels in the gray level histogram of the principal component image, m20 represents the number of non-zero gray levels in the gray level histogram of the panchromatic image, m1 q represents the number of pixels of the gray level q in the gray level histogram of the principal component image, m2 q represents the number of pixels of the gray level q in the gray level histogram of the panchromatic image, norm() represents a linear normalization function, and E represents the blur degree of each partition of the principal component image.

[0074] The higher the blur degree is, the more serious the detail blur caused by the low resolution of the multi-band image and the PCA transformation is. Therefore, the details in the panchromatic image should be retained more when the histogram is matched. The matching value of any gray value of the panchromatic image is adjusted. The gray level range of any partition of the principal component image and the panchromatic image is obtained. The gray level range is a continuous gray interval with a number greater than 3 in the gray level histogram. For example, the gray levels 1-100 all have pixels, and the gray level 101 does not have pixels. Therefore, 1-100 is considered as a gray level range. The maximum value and the minimum value of the gray level range are obtained. The larger the difference between the maximum value and the minimum value is, the more uneven the histogram distribution is. The smaller the difference between the maximum value and the minimum value is, the more uniform the histogram distribution is. For each partition, based on the matching value of the panchromatic image, the blur degree of the principal component image, and the difference between the maximum difference of the gray level range of the panchromatic image and the maximum difference of the gray level range of the principal component image, the corresponding matching adjustment value of the panchromatic image is obtained.

[0075] The matching adjustment value is positively correlated with the blurring degree of the principal component image, the matching value of the panchromatic image, the maximum difference of the gray value concentration range of the panchromatic image, and the difference of the maximum difference of the gray value concentration range of the principal component image.

[0076] Preferably, in the embodiment, the expression of the matching adjustment value is:

[0077] q 1,max is the maximum value of the gray value concentration range of the principal component image, q 1,min is the minimum value of the gray value concentration range of the principal component image, q 2,max is the maximum value of the gray value concentration range of the panchromatic image, q 2,min is the minimum value of the gray value concentration range of the panchromatic image, E is the blurring degree of each partition of the principal component image, and q ′ is the matching value of the gray value q, and q" is the matching adjustment value of the gray value q.

[0078] Up to now, the matching adjustment value of each pixel point of each partition is obtained.

[0079] In step S004, the matching partition image is compared with the multi-band image and the panchromatic image to obtain the fusion value of the edge pixel point, and based on this, the fusion image is obtained.

[0080] The matching adjustment value of each gray value of each partition is obtained through the above steps, and the histogram matching of the partition is completed. After that, the adjusted image of each partition is obtained through the above operation for all partitions. The fusion image of all partitions is obtained by using the inverse PCA transformation. The gray value in each fusion image is the matching adjustment value obtained by calculation. Since each partition is transformed inversely, the edge pixel point exists in multiple partitions, that is, the edge pixel point has different gray values in different partitions, so it is necessary to calculate a gray value from all the gray values of each edge pixel point. The gray value needs to meet the different ecological region colors and the clear division of the overall image. The gray value that is more consistent with the panchromatic image and the multi-band image is given a larger weight.

[0081] The partition where the edge is located is recorded as the adjacent partition of the edge. For any pixel point, the gray values of the pixel point in all adjacent partitions form a fusion sequence, and the sequence length is the number of adjacent partitions.

[0082] A normal line is drawn at any pixel point in the edge, the normal line intersects with the adjacent partition of the edge, and the gradient change on the normal line reflects the division clarity of the pixel point. In all gray values, the higher the division clarity, the larger the weight.

[0083] An edge pixel point is selected as a target point, and a pixel point on the normal line of the target point is selected as a normal point; if the difference between the gradient on the normal line after fusion and the gradient on the normal line before fusion is larger, it indicates that the fusion has a negative effect on the edge definition, and the weight is smaller.

[0084] Therefore, the weight of the target point is obtained based on the difference between the gray value of the target point and the gray value of the normal point in the fused image, the panchromatic image and the multi-band image.

[0085] Preferably, the expression of the weight is:

[0086] h s h represents the gray value of the target point in the s th partition of the fused image, h a h represents the gray value of the a th normal point in the fused image, h s1 h represents the gray value of the target point in the s th partition of the multi-band image, h 1,a h represents the gray value of the a th normal point in the multi-band image, h s2 h represents the gray value of the target point in the s th partition of the panchromatic image, h 1,a h represents the gray value of the a th normal point in the panchromatic image, A represents the number of normal points in the normal line, exp[] represents an exponential function with a natural constant as the base, K s h represents the weight of the target point in the s th partition of the fused image.

[0087] The gray value of the target point in each partition is weighted by the weight of the target point in each partition to obtain the fusion value of the target point. Preferably, in the method of the embodiment, for multiple gray values of the target point, the weight of each partition is the ratio of the weight of each partition to the sum of the weights of all partitions, and the ratio is multiplied by each gray value and accumulated, and the obtained value is the fusion value of the target point.

[0088] All edge pixel points in the fused image are replaced by the fusion value to obtain a final fused image, and the remote sensing data fusion is completed.

[0089] Based on the same inventive concept as the above method, the embodiment of the present application also provides a multi-source ecological remote sensing data fusion system, which comprises a memory, a processor and a computer program stored in the memory and running on the processor, and the processor implements the steps of any one of the above multi-source ecological remote sensing data fusion methods when executing the computer program.

[0090] It should be noted that the above-mentioned embodiments are only used to illustrate the technical solutions of the present application, but not limit them; although the present application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that they can still modify the technical solutions recorded in the foregoing embodiments, or make equivalent replacements for some of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the scope of the technical solutions of the embodiments of the present application, and should be included in the protection scope of the present application.

[0091] Each of the embodiments in the specification is described in a progressive manner, and the same or similar parts between the embodiments can be referred to each other, and each embodiment mainly explains the difference from other embodiments.

Claims

1. A multi-source ecological remote sensing data fusion method, characterized in that, The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The matching adjustment value is used to obtain the fusion image of each partition, and the weight of the partition where each edge pixel point is located is obtained based on the gray value difference between the fusion image and the multi-band image and the panchromatic image The fusion value is obtained by weighting the multiple matching adjustment values of the edge pixel point based on the weight, and the final fusion image is obtained by replacing the edge pixel point with the fusion value. The method comprises the following steps: , represents the number of pixels within the first window, represents the number of pixels within the second window, represents the number of edge pixels added to the first window, represents the number of edge pixels added to the second window, represents a linear normalization function, represents a merge value; The method comprises the following steps: The method comprises the following steps: , denotes an interpolation of interpolation points, denotes a gray value of the jth pixel point in the ith direction, denotes a maximum function, denotes a minimum function, denotes a linear normalization function, denotes an interpolation bias; The method comprises the following steps: The method comprises the following steps: , is an interpolation correction value for the interpolation point; is an interpolation for the interpolation point; is an interpolation adjustment value for the deviation direction, is an interpolation deviation; The method comprises the following steps: The method comprises the following steps: , Gs represents the gray value of the target point in the s-th partition of the fused image, Ga represents the gray value of the a-th normal point in the fused image, Gs represents the gray value of the target point in the s-th partition of the multi-band image, Ga represents the gray value of the a-th normal point in the multi-band image, Gs represents the gray value of the target point in the s-th partition of the panchromatic image, Ga represents the gray value of the a-th normal point in the panchromatic image, N represents the number of normal points in the normal line, E represents the exponential function with the natural constant as the base, Ws represents the weight of the target point in the s-th partition of the fused image, and the gray value of the target point in each partition is the matching adjustment value thereof in each partition.

2. The multi-source ecological remote sensing data fusion method of claim 1, wherein, The method comprises the following steps: The method comprises the following steps:

3. The multi-source ecological remote sensing data fusion method of claim 1, wherein, The method comprises the following steps:

4. The multi-source ecological remote sensing data fusion method of claim 1, wherein, The method comprises the following steps:

5. The multi-source ecological remote sensing data fusion method of claim 1, wherein, The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the following steps: The method comprises the For multiple gray values of the target point, the weight of each partition is the ratio of the weight of the partition to the sum of the weights of all partitions, and the obtained value is the fusion value of the target point after the ratio is multiplied by each gray value and accumulated.

6. A multi-source ecological remote sensing data fusion system comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that, The processor implements the steps of the multi-source ecological remote sensing data fusion method according to any one of claims 1-5 when executing the computer program.

Images