A method, device, equipment and medium for quickly identifying internal dislocation of remote sensing images

Abstract

The present invention discloses a method, device, equipment and medium for quickly identifying the internal dislocation of a remote sensing image, and relates to the technical field of identifying the internal dislocation of a remote sensing image, wherein the method comprises: obtaining a panchromatic image and a multispectral image of the same remote sensing image, finding out the same-name points of the panchromatic image and the multispectral image, calculating the object coordinates corresponding to the same-name points, comparing the position offset of the object coordinates obtained by the same-name points, and identifying whether the remote sensing image is internally dislocated according to the position offset. The present invention obtains the same-name points corresponding to the panchromatic image and the multispectral image by calculating the image-side and image-side geometric relationship in the remote sensing image, and obtains the object coordinates of the same-name image points. By calculating the object coordinate deviation of the same-name image points, it can be determined whether the remote sensing image is internally dislocated. For the object coordinates of the same-name points of the panchromatic image and the multispectral image, when the average deviation exceeds 10 meters, it can be accurately determined and identified, thereby improving the accuracy of the internal dislocation of the remote sensing image.

Description

Technical Field

[0001] The present invention relates to the technical field of internal misalignment recognition of remote sensing images in remote sensing data processing, and in particular to a method, device, equipment and medium for quickly recognizing internal misalignment of remote sensing images. Background Art

[0002] Remote sensing images are real images of the earth's surface obtained through sensors carried by remote sensing platforms. These images are usually taken or recorded by satellites, aircraft, drones or other remote sensing platforms.

[0003] Remote sensing images truly record the different land features on the earth's surface, such as land, water bodies, vegetation, urban areas, etc., as well as their reflection or radiation information in different spectral bands.

[0004] However, remote sensing images may be dislocated, distorted, and deformed during the process of acquisition, transmission, and processing. There are many reasons for the dislocation, distortion, and deformation of remote sensing images, including but not limited to the following aspects: ① The aging of remote sensing sensor hardware facilities such as lenses, detection elements, sampling rates, scanning mirrors, and other components may cause remote sensing images to be dislocated, distorted, and deformed. ② The height, speed, orbital deviation, and attitude changes of the remote sensing platform can also cause remote sensing images to be dislocated, distorted, and deformed. The swing or distortion of the remote sensing platform during movement can also affect the spatial position relationship of the remote sensing image; ③ Factors of the earth itself, such as the earth's rotation, terrain undulations, and earth curvature, can also cause the geometric position relationship of remote sensing images to be distorted. Terrain undulations can cause the elevation position of pixels to change, thereby affecting the geometric relationship of remote sensing; ④ Atmospheric refraction can cause positional offsets in the image, and different projection methods can also affect the geometric characteristics of the image, thereby causing distortion. Therefore, before large-scale remote sensing data processing, it is crucial to identify whether there is serious dislocation, distortion, and deformation inside the remote sensing image and locate the position where the dislocation occurs.

[0005] Existing methods for identifying internal misalignment of remote sensing images mostly use reference geographic information images as benchmark images for comparison. However, after the benchmark image is generated, there are often errors due to the time interval between the generated image and the application time. In addition, the benchmark image is often not accurately corrected, which reduces the recognition rate of internal misalignment of remote sensing images.

[0006] For example: the invention application with application number 202410489196.9 discloses a quantitative quality inspection method, device and electronic equipment for the internal geometric distortion of remote sensing images, which belongs to the technical field of remote sensing image quality inspection. In this method, the edges of the triangulated network in the remote sensing image to be inspected are distinguished by different colors to characterize the distribution of the internal geometric distortion of the remote sensing image to be inspected, so that the user can visually (i.e., in appearance) intuitively and accurately capture the internal geometric distortion of the remote sensing image to be inspected, and also uses the internal geometric distortion mean error of the remote sensing image to be inspected and the gross error rate of the remote sensing image to be inspected as the quality inspection result of the internal geometric distortion of the remote sensing image to be inspected, that is, it can quantitatively describe the internal geometric distortion value of the remote sensing image to be inspected. The application scheme has the problem that it relies too much on the benchmark reference image and cannot be identified based on the relationship between the image and the image, which affects the scope of application and is not easy to accurately identify the internal misalignment of the remote sensing image. Summary of the invention

[0007] In view of the above-mentioned problems, the purpose of the present invention is to provide a method, device, equipment and medium for quickly identifying internal misalignment of remote sensing images, which can utilize the characteristics of the remote sensing image itself and identify internal misalignment of the remote sensing image based on the differences between the image space and the image space, or between the image space and the object space.

[0008] The embodiments of the present invention provide a method, device, equipment and medium for quickly identifying internal dislocation of a remote sensing image.

[0009] Aspect 1: A method for quickly identifying internal dislocation of remote sensing images, comprising:

[0010] S1, obtaining a panchromatic image and a multispectral image of the same remote sensing image, and dividing them into multiple area blocks respectively;

[0011] S2, finding out the image points with the same name in the multispectral image area block corresponding to the image points in the panchromatic image area block, and the image points and the corresponding image points with the same name form a group of points with the same name;

[0012] S3, respectively calculating the object space coordinates corresponding to the image points of the panchromatic image and the image points of the multispectral image with the same name;

[0013] S4, sequentially comparing the position offsets of the object coordinates obtained from the image points of the panchromatic image and the object coordinates obtained from the image points of the same name in the multispectral image;

[0014] S5. Identify whether the remote sensing image is misaligned based on the position offset.

[0015] Optionally, also include:

[0016] S6, obtaining a reference geographic information image corresponding to the remote sensing image as a reference image, and dividing it into a plurality of area blocks;

[0017] S7, finding out the image points with the same name in the reference image area block corresponding to the image points in the panchromatic image area block, and the image points and the corresponding image points with the same name form a group of points with the same name;

[0018] S8, respectively calculating the object space coordinates corresponding to the image points of the panchromatic image and the image points of the same name of the reference image;

[0019] S9, sequentially comparing the position offsets of the object coordinates obtained by the full-color image and the object coordinates obtained by the reference image;

[0020] S10, identifying whether the full-color image is misaligned based on the position offset.

[0021] Optionally, also include:

[0022] S11, obtaining a reference geographic information image corresponding to the multispectral image as a reference image, and dividing it into a plurality of area blocks;

[0023] S12, finding out the image points with the same name in the reference image area block corresponding to the image points in the multispectral image area block, and the image points and the corresponding image points with the same name form a group of points with the same name;

[0024] S13, respectively calculating the object space coordinates corresponding to the image points of the multispectral image and the image points of the same name in the reference image;

[0025] S14, sequentially comparing the position offsets of the object coordinates obtained by the multispectral image and the object coordinates obtained by the reference image;

[0026] S15. Identify whether the multispectral image is misaligned based on the position offset.

[0027] Optionally, finding the image point with the same name corresponding to the image point includes the steps of:

[0028] S21, extracting feature points of panchromatic or multispectral images based on Moravec operator;

[0029] S22, based on the feature points of the extracted image points, using the least squares grayscale matching algorithm to find the image points with the same name corresponding to the target image;

[0030] Among them, the image point and the corresponding image point with the same name form a group of points with the same name.

[0031] Optionally, the step S3 of respectively calculating the object coordinates corresponding to the image points of the panchromatic image and the image points of the multispectral image with the same name includes:

[0032] Calculate the object coordinates corresponding to the image points based on the RPC parameter file of the panchromatic image or multispectral image and the DEM elevation data;

[0033] The formula is:

[0034]

[0035] In the formula, the polynomial The form is:

[0036]

[0037] In the formula, is the rational function coefficient, obtained through the RPC parameter file, and Zn is obtained through the DEM elevation data. , Represents pixel coordinates （r，c） and ground point spatial coordinates (X, Y, Z) Normalized coordinates after coordinate translation and scaling;

[0038] The calculation of the object coordinates corresponding to the image points with the same name in the reference image in S13 includes:

[0039] According to the 6-parameter file of the affine transformation of the reference image, the object space coordinates of the image points of the reference image are calculated. The affine transformation formula is:

[0040] X = ax+by+c

[0041] Y = dx + ey + f

[0042] in X, Y is the object coordinate, x,y is the image coordinate, a, b, c, d, e, f It is the affine transformation 6 parameter file of the reference image, which is directly obtained from the reference image.

[0043] Optionally, identify whether the image is misaligned, including:

[0044] The object coordinate deviations calculated for the same-name points in multiple area blocks are averaged;

[0045] When the average deviation of the object coordinates of the same-name points in the panchromatic image and the multispectral image is greater than the threshold of 10 meters, there is a misalignment inside the remote sensing image;

[0046] When the average deviation of the object coordinates of the same-name points between the panchromatic image and the reference image is greater than the threshold of 20 meters, there is a misalignment inside the panchromatic image;

[0047] When the average deviation of the object coordinates of the same-name points between the multispectral image and the reference image is greater than the threshold of 20 meters, there is misalignment inside the multispectral image.

[0048] A second aspect: A device for quickly identifying internal dislocation of a remote sensing image, comprising:

[0049] Panchromatic image module, which obtains the panchromatic image corresponding to the remote sensing image and divides it into multiple area blocks;

[0050] The multispectral image module obtains the multispectral image corresponding to the remote sensing image and divides it into multiple area blocks;

[0051] The reference image module obtains the reference geographic information image corresponding to the remote sensing image as the reference image and divides it into multiple area blocks;

[0052] The same-name image point search module finds the same-name image points corresponding to the target image based on the panchromatic image or multispectral image points to form a group of same-name points;

[0053] An object coordinate calculation module calculates the position offset of the object coordinates of the image point and the corresponding image point with the same name;

[0054] The misalignment determination module identifies whether there is misalignment inside the remote sensing image based on the position offset.

[0055] A third aspect: An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps of the method provided in the first aspect when executing the program.

[0056] A fourth aspect: A non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the method provided in the first aspect.

[0057] Beneficial effects of the present invention:

[0058] 1. The present invention calculates the geometric relationship between the image space and the image space in the remote sensing image, obtains the same-name points corresponding to the panchromatic image and the multispectral image through a matching algorithm, and then obtains the object coordinates of the same-name image points by using the same-name points and the corresponding RPC parameter file and DEM elevation file. By calculating the object coordinate deviation of the same-name image points, it can be determined whether the remote sensing image has internal misalignment. When the average deviation exceeds 10 meters, it can be accurately determined and identified, thereby improving the accuracy of internal misalignment of the remote sensing image.

[0059] 2. When both the panchromatic image and the multispectral image of the remote sensing image have similar serious misalignment, the present invention can utilize the geometric relationship between the image and object space of the remote sensing image, introduce the image of the reference geographic information coordinates, obtain the same-name points of the panchromatic image and the reference image, and the same-name points of the multispectral image and the reference image through a matching algorithm, and then determine whether the panchromatic image or the multispectral image has internal misalignment by utilizing the deviation of the object space coordinates of the panchromatic image and the multispectral image from the object space coordinates of the reference image. The method is flexible and is conducive to quickly identifying the internal misalignment of the remote sensing image.

[0060] 3. For large-area remote sensing image processing, the present invention can use multi-cluster parallel execution to sequentially compare the image-to-image coordinate relationship and the image-to-object coordinate relationship of the panchromatic images and multispectral images in the area, thereby quickly finding out whether the original remote sensing data of the area has seriously misplaced remote sensing data. BRIEF DESCRIPTION OF THE DRAWINGS

[0061] Figure 1 It is a schematic diagram of the process of the present invention;

[0062] Figure 2 It is the principle flow chart of the method of the present invention;

[0063] Figure 3 It is a flow chart of the method for identifying misalignment between the image side and the image side in a remote sensing image according to the present invention;

[0064] Figure 4 It is a flow chart of the method for identifying the misalignment between the image and object space of a remote sensing image and a reference image of the present invention;

[0065] Figure 5 It is a structural schematic diagram of the device of the present invention;

[0066] Figure 6 It is the panchromatic image and multispectral image of the present invention;

[0067] Figure 7 This is a diagram showing that the panchromatic image and the multispectral image of the present invention are seriously misaligned;

[0068] Figure 8 The full-color image and the reference image of the present invention;

[0069] Fig. 9 The full-color image of the present invention and the reference image are seriously misaligned;

[0070] Fig.10 FIG. 2 is a diagram of an electronic device according to the present invention. DETAILED DESCRIPTION

[0071] Embodiments of the present invention are described in detail below, examples of which are shown in the accompanying drawings, wherein the same or similar symbols throughout represent the same or similar elements or elements with the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present invention, and cannot be understood as limiting the present invention.

[0072] In the actual imaging process of remote sensing satellites, unstable factors such as deformation of photographic materials and lens distortion cause serious internal misalignment of images, which affects the application accuracy of remote sensing images.

[0073] In view of the above problems, the present invention provides a method for quickly identifying internal misalignment of remote sensing images. Figure 1A schematic diagram of a flow chart of a method for rapidly identifying internal misalignment of a remote sensing image provided by an embodiment of the present invention. Figure 2 It is a principle flow chart of the method of the present invention, and the specific example is Figure 3 As shown, the method includes:

[0074] S1. Obtain a panchromatic image and a multispectral image of the same remote sensing image, and divide them into multiple area blocks respectively.

[0075] like Figure 6 As shown, the original data of remote sensing images include two images in TIFF / TIF format, one of which is colorless (also called grayscale image), and the other is color-coded (also called color image). In the application field, grayscale images are called panchromatic images, and color images are called multispectral images.

[0076] First, the panchromatic image and the multispectral image are divided into blocks according to their respective specified sizes. For example, the same specifications are used to divide them into 100 area blocks (10 rows × 10 columns).

[0077] S2. Find out the image points with the same name in the multispectral image area block corresponding to the image points in the panchromatic image area block, and the image points and the corresponding image points with the same name form a group of points with the same name.

[0078] Because panchromatic images only obtain a single band, they are displayed as grayscale images. Panchromatic remote sensing images generally have high spatial resolution, but cannot display the color of objects.

[0079] The multispectral image sensor captures multiple single bands of ground radiation to obtain images containing spectral information of multiple bands. Color images will be obtained by assigning RGB colors to different bands.

[0080] Although there are significant differences in the physical properties of panchromatic and multispectral images, the spatial positions of objects corresponding to panchromatic and multispectral images are less different.

[0081] Based on the above reasons, feature points of 100 area blocks of the panchromatic image can be extracted based on the Moravec operator, and then the same-name image points of the corresponding multispectral image can be found based on the least squares grayscale matching algorithm (the same-name image points are also called corresponding image points, which refer to the image points of any object in different images). The image points and the corresponding same-name image points form a group of same-name points.

[0082] It is also possible to extract feature points from 100 area blocks of the multispectral image based on the Moravec operator, and then find the same-name image points of the corresponding panchromatic image based on the least squares grayscale matching algorithm (the same-name image points are also called corresponding image points, which refer to the image points of any object in different images). The image points and the corresponding same-name image points form a group of same-name points.

[0083] S3. Calculate the object coordinates of the panchromatic image points and the multispectral image points with the same name.

[0084] In addition to the two panchromatic and multispectral images in TIFF / TIF format, the original data of remote sensing images also include RPC parameter files that match the TIFF image (panchromatic image) and the RPC parameter files that match the TIF image (multispectral image). The essence of the RPC (Rational Polynomial Coefficients) parameter file is the rational function correction model RFM (Rational Function Model), which expresses the image point coordinates [r (row number), c (column number)] as the ratio of the polynomial with the corresponding ground point space [X (longitude), Y (latitude), Z (geodetic height)] as the independent variable. The RPC parameter file mainly describes the functional relationship between the row and column coordinates of the pixels of the remote sensing image and the corresponding object coordinates [X (longitude), Y (latitude), Z (geodetic height)]. It describes the conversion relationship from three-dimensional geographic coordinates to two-dimensional satellite image coordinates, which is generally referred to as from object to image, which can be understood as three-dimensional to two-dimensional. This is also called orthographic projection, specifically, given the known ground point space [X (longitude), Y (latitude), Z (geodetic height)], the pixel coordinates [r (row number), c (column number)] are obtained.

[0085] The reverse projection is from the image coordinate system to the geographic coordinate system, that is, from the image space to the object space. It should be noted here that this is still a three-dimensional to two-dimensional relationship, because two-dimensional cannot be directly upgraded to three-dimensional. In order to solve the corresponding geographic coordinates according to the row and column numbers of the remote sensing image, DEM elevation data must be introduced. The matching points of the obtained panchromatic image and multispectral image, that is, the image points with the same name, are calculated in turn according to their corresponding RPC files and DEM elevation data to calculate the corresponding object space coordinates.

[0086] The formula is:

[0087]

[0088] In the formula, the polynomial The form is:

[0089]

[0090] In the formula, is the rational function coefficient, obtained through the RPC parameter file, and Zn is obtained through the DEM elevation data. , Represents pixel coordinates （r，c） and ground point spatial coordinates (X, Y, Z) Normalized coordinates after coordinate translation and scaling.

[0091] S4. Compare the position offset of the object coordinates obtained from the image points of the panchromatic image with the corresponding object coordinates obtained from the image points of the same name in the multispectral image.

[0092] The same-name points of the panchromatic image and the multispectral image obtained above are the image points of the same point on the ground in different images. Ideally, the same-name points of the panchromatic image and the multispectral image have the same corresponding object coordinate values ​​calculated according to the RPC files and DEM elevation data corresponding to their respective images. However, due to the combined influence of various factors in the remote sensing imaging process, the geometric position, shape, size, dimension, orientation and other characteristics of the objects on the original image are often inconsistent with the characteristics of the corresponding ground objects, but they are basically controlled within a reasonable range, generally within 10 meters, and the comparison threshold can be set to 10 meters.

[0093] like Figure 7 As shown in the figure, if the object coordinates calculated from the same-name points have a large position offset, there is a strong reason to suspect that there is a serious misalignment inside the panchromatic image or multispectral image in the original remote sensing image.

[0094] S5. Identify whether the remote sensing image is misaligned based on the position offset.

[0095] The object space coordinate deviations of the same-name points in 100 area blocks of the panchromatic image and the multispectral image are calculated and averaged. The area blocks where the calculated object space coordinate deviation average is greater than 10 meters are defined as areas with serious misalignment.

[0096] The object coordinate deviations calculated for the same-name points in multiple regional blocks in the panchromatic image and the multispectral image are averaged. When the average object coordinate deviation of the same-name points between the panchromatic image and the multispectral image is greater than the threshold of 10 meters, it is determined that there is misalignment inside the remote sensing image.

[0097] The above method calculates the image-to-image geometric relationship in the remote sensing image, that is, firstly, the same-name points corresponding to the panchromatic image and the multispectral image are obtained through a matching algorithm, and then the object-space coordinates of the same-name points are obtained by using the same-name points and the corresponding RPC parameter file and DEM elevation file. By calculating the object-space coordinate deviation of the same-name points, it can be determined whether internal dislocation occurs in the remote sensing image. It is an application of the image-to-image geometric relationship.

[0098] The above method steps are applicable to the case where only one of the panchromatic image and the multispectral image has a serious internal misalignment. If similar offsets occur in both the corresponding panchromatic image and the multispectral image, it is difficult to draw a conclusion by calculating the corresponding object coordinate offset in the above method steps.

[0099] In view of the above situation, if Figure 8As shown, a reference geographic information image with roughly the same area as the original remote sensing image can be introduced as a reference image to assist in judgment, including the following steps: Figure 4 As shown:

[0100] S6, obtaining a reference geographic information image corresponding to the remote sensing image as a reference image, and dividing it into a plurality of area blocks;

[0101] First, calibrate the coordinates of the reference image and the original remote sensing image to form a more accurate correspondence between the reference image and the panchromatic image, and then divide the image into blocks of a specified size, for example, into 100 area blocks (10 rows × 10 columns) corresponding to the panchromatic image.

[0102] S7, finding out the image points with the same name in the reference image area block corresponding to the image points in the panchromatic image area block, and the image points and the corresponding image points with the same name form a group of points with the same name;

[0103] Then, feature points of 100 area blocks of the panchromatic image are extracted based on the Moravec operator, and then the corresponding image points with the same name are matched on the reference image based on the least squares grayscale matching algorithm. The image points of the panchromatic image and the image points with the same name in the reference image constitute a group of points with the same name.

[0104] S8, respectively calculating the object space coordinates corresponding to the image points of the panchromatic image and the image points of the same name of the reference image;

[0105] The corresponding object coordinates of the image points on the panchromatic image are calculated according to the corresponding RPC parameter file and DEM elevation data.

[0106] The reference image contains an affine transformation function that converts pixel points to geographic coordinate points, which can be used to calculate the object coordinates of the same-name image points on the reference image.

[0107] The 6-parameter file of the affine transformation of the reference image is used to calculate the object coordinates of the image points of the reference image. The affine transformation formula is:

[0108] X = ax+by+c

[0109] Y = dx + ey + f

[0110] in X, Y is the object coordinate, x,y is the image coordinate, a, b, c, d, e, f It is the affine transformation 6 parameter file of the reference image, which is directly obtained from the reference image.

[0111] S9, sequentially comparing the position offsets of the object coordinates obtained by the full-color image and the object coordinates obtained by the reference image.

[0112] like Fig. 9As shown in the figure, the object coordinates of the image points in 100 regional blocks of the full-color image are calculated and the object coordinate deviations are obtained from the corresponding reference image to find the average value. The average value is generally within a reasonable range, generally within 20 meters. The threshold is taken as 20 meters. If the calculated object coordinate deviation average value is greater than 20 meters, it means that there are regional blocks with serious misalignment in the full-color image, and then it is processed in different regional blocks.

[0113] S10, identifying whether the full-color image is misaligned based on the position offset.

[0114] The object space coordinate deviations calculated for the same-name points in multiple area blocks between the panchromatic image and the reference image are averaged. When the average object space coordinate deviations of the same-name points between the panchromatic image and the reference image are greater than a threshold of 20 meters, it is determined that there is misalignment inside the panchromatic image.

[0115] This method uses the geometric relationship between the image and object space of remote sensing images, that is, introduces images with reference geographic information coordinates. That is, first, the same-name points between the panchromatic image and the reference image and the same-name points between the panchromatic image and the reference image are obtained through a matching algorithm. Then, by using the deviation between the multispectral object space coordinates and the object space coordinates of the reference image, it can be determined whether the panchromatic image has internal misalignment. This is an application of the geometric relationship between the image and object space.

[0116] Furthermore, multispectral images and reference images can be used for auxiliary judgment, including the following steps:

[0117] S11, obtaining a reference geographic information image corresponding to the multispectral image as a reference image, and dividing it into a plurality of area blocks;

[0118] First, calibrate the coordinates of the reference image and the original remote sensing image to form a more accurate correspondence between the reference image and the multispectral image, and then divide the image into blocks of a specified size, for example, into 100 area blocks (10 rows × 10 columns) corresponding to the multispectral image.

[0119] S12, finding out the image points with the same name in the reference image area block corresponding to the image points in the multispectral image area block, and the image points and the corresponding image points with the same name form a group of points with the same name;

[0120] Then, feature points of 100 regional blocks of the multispectral image are extracted based on the Moravec operator, and then the corresponding image points with the same name are matched on the reference image based on the least squares grayscale matching algorithm. The image points of the multispectral image and the image points with the same name in the reference image constitute a group of points with the same name.

[0121] S13, respectively calculating the object space coordinates corresponding to the image points of the multispectral image and the image points of the same name in the reference image;

[0122] The corresponding object coordinates of the image points on the multispectral image are calculated according to the corresponding RPC parameter file and DEM elevation data.

[0123] The reference image contains an affine transformation function that converts pixel points to geographic coordinate points, which can be used to calculate the object coordinates of the same-name image points on the reference image.

[0124] According to the 6-parameter file of the affine transformation of the reference image, the object space coordinates of the image points of the reference image are calculated. The affine transformation formula is:

[0125] X = ax+by+c

[0126] Y = dx + ey + f

[0127] in X, Y is the object coordinate, x,y is the image coordinate, a, b, c, d, e, f It is the affine transformation 6 parameter file of the reference image, which is directly obtained from the reference image.

[0128] S14, sequentially comparing the position offsets of the object coordinates obtained by the multispectral image and the object coordinates obtained by the reference image;

[0129] The object coordinates of the image points in 100 regional blocks of the multispectral image are calculated and the object coordinate deviations are averaged with those of the corresponding reference image. The threshold is 20 meters. If the calculated average object coordinate deviation is greater than 20 meters, it means that there are severely misaligned regional blocks in the multispectral image, and then the image is processed in blocks.

[0130] S15. Identify whether the multispectral image is misaligned based on the position offset.

[0131] The object coordinate deviations calculated for the same-name points in multiple regional blocks in the multispectral image and the reference image are averaged. When the average object coordinate deviations of the same-name points in the multispectral image and the reference image are greater than a threshold of 20 meters, it is determined that there is misalignment inside the multispectral image.

[0132] This method utilizes the geometric relationship between the image and object space of remote sensing images, that is, introduces images with reference geographic information coordinates. That is, first, the same-name points between the panchromatic image and the reference image, as well as the same-name points between the multispectral image and the reference image are obtained through a matching algorithm. Then, by utilizing the deviation between the object space coordinates of the multispectral image and the object space coordinates of the reference image, it can be determined whether the multispectral image has internal misalignment. This is an application of the geometric relationship between the image and object space.

[0133] The present invention also discloses a device applied to the above method, such as Figure 5 As shown, including:

[0134] Panchromatic image module, which obtains the panchromatic image corresponding to the remote sensing image and divides it into multiple area blocks;

[0135] The multispectral image module obtains the multispectral image corresponding to the remote sensing image and divides it into multiple area blocks;

[0136] The reference image module obtains the reference geographic information image corresponding to the remote sensing image as the reference image and divides it into multiple area blocks;

[0137] The same-name image point search module finds the same-name image points corresponding to the target image based on the panchromatic image or multispectral image points to form a group of same-name points;

[0138] An object coordinate calculation module calculates the position offset of the object coordinates of the image point and the corresponding image point with the same name;

[0139] The misalignment determination module identifies whether there is misalignment inside the remote sensing image based on the position offset.

[0140] When in use, the panchromatic image module obtains the panchromatic image corresponding to the remote sensing image, the multispectral image module obtains the multispectral image corresponding to the remote sensing image, the benchmark image module obtains the reference geographic information image corresponding to the remote sensing image as the benchmark image, and the same-name image point search module finds the same-name image points corresponding to the target image based on the panchromatic image or multispectral image image points to form a group of same-name points; the object coordinates are used to calculate the position offset of the object coordinates of the same-name points, and the misalignment judgment module identifies whether there is misalignment inside the remote sensing image based on the position offset.

[0141] The present invention also provides an electronic device, Fig.10 A schematic diagram of the structure of an electronic device provided by an embodiment of the present invention, such as Fig.10 As shown, the electronic device may include: a processor, a communications interface, a memory, and a communications bus, wherein the processor, the communications interface, and the memory communicate with each other via the communications bus. The processor may call the logic instructions in the memory, for example, to execute the following method:

[0142] S1, obtaining a panchromatic image and a multispectral image of the same remote sensing image, and dividing them into multiple area blocks respectively;

[0143] S2, finding out the image points with the same name in the multispectral image area block corresponding to the image points in the panchromatic image area block, and the image points and the corresponding image points with the same name form a group of points with the same name;

[0144] S3, respectively calculating the object space coordinates corresponding to the image points of the panchromatic image and the image points of the multispectral image with the same name;

[0145] S4, sequentially comparing the position offsets of the object coordinates obtained from the image points of the panchromatic image and the object coordinates obtained from the image points of the same name in the multispectral image;

[0146] S5. Identify whether the remote sensing image is misaligned based on the position offset.

[0147] In addition, the logic instructions in the above-mentioned memory can be implemented in the form of software functional units and can be stored in a computer-readable storage medium when sold or used as an independent product. Based on this understanding, the technical solution of the present invention can be essentially or partly embodied in the form of a software product that contributes to the prior art. The computer software product is stored in a storage medium, including several instructions to enable a computer device (which can be a personal computer, server, or network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk, etc. Various media that can store program codes.

[0148] An embodiment of the present invention further provides a non-transitory computer-readable storage medium on which a computer program is stored. When the computer program is executed by a processor, the method provided in each of the above embodiments is implemented, for example, including:

[0149] S1, obtaining a panchromatic image and a multispectral image of the same remote sensing image, and dividing them into multiple area blocks respectively;

[0150] S2, finding out the image points with the same name in the multispectral image area block corresponding to the image points in the panchromatic image area block, and the image points and the corresponding image points with the same name form a group of points with the same name;

[0151] S3, respectively calculating the object space coordinates corresponding to the image points of the panchromatic image and the image points of the multispectral image with the same name;

[0152] S4, sequentially comparing the position offsets of the object coordinates obtained from the image points of the panchromatic image and the object coordinates obtained from the image points of the same name in the multispectral image;

[0153] S5. Identify whether the remote sensing image is misaligned based on the position offset.

[0154] The device embodiments described above are merely illustrative, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. Ordinary technicians in this field can understand and implement it without paying creative labor.

[0155] Through the description of the above implementation methods, those skilled in the art can clearly understand that each implementation method can be implemented by means of software plus a necessary general hardware platform, and of course, can also be implemented by hardware. Based on this understanding, the above technical solution is essentially or the part that contributes to the prior art can be embodied in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a disk, an optical disk, etc., including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.

[0156] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for quickly identifying internal dislocation of remote sensing images, characterized in that: include: S1, obtaining a panchromatic image and a multispectral image of the same remote sensing image, and dividing them into multiple area blocks respectively; S2, finding out the image points with the same name in the multispectral image area block corresponding to the image points in the panchromatic image area block, and the image points and the corresponding image points with the same name form a group of points with the same name; S3, respectively calculating the object space coordinates corresponding to the image points of the panchromatic image and the image points of the multispectral image with the same name; S4, sequentially comparing the position offsets of the object coordinates obtained from the image points of the panchromatic image and the object coordinates obtained from the image points of the same name in the multispectral image; S5, identifying whether the remote sensing image is misaligned based on the position offset; The method of finding the image points with the same name corresponding to the image points comprises the steps of: S21, extracting feature points of panchromatic or multispectral images based on Moravec operator; S22, based on the feature points of the extracted image points, using the least squares grayscale matching algorithm to find the image points with the same name corresponding to the target image; Among them, the image point and the corresponding image point with the same name form a group of points with the same name.

2. The method according to claim 1, characterized in that: Also includes: S6, obtaining a reference geographic information image corresponding to the remote sensing image as a reference image, and dividing it into a plurality of area blocks; S7, finding out the image points with the same name in the reference image area block corresponding to the image points in the panchromatic image area block, and the image points and the corresponding image points with the same name form a group of points with the same name; S8, respectively calculating the object space coordinates corresponding to the image points of the panchromatic image and the image points of the same name of the reference image; S9, sequentially comparing the position offsets of the object coordinates obtained by the full-color image and the object coordinates obtained by the reference image; S10, identifying whether the full-color image is misaligned based on the position offset.

3. The method according to claim 2, characterized in that Also includes: S11, obtaining a reference geographic information image corresponding to the multispectral image as a reference image, and dividing it into a plurality of area blocks; S12, finding out the image points with the same name in the reference image area block corresponding to the image points in the multispectral image area block, and the image points and the corresponding image points with the same name form a group of points with the same name; S13, respectively calculating the object space coordinates corresponding to the image points of the multispectral image and the image points of the same name in the reference image; S14, sequentially comparing the position offsets of the object coordinates obtained by the multispectral image and the object coordinates obtained by the reference image; S15. Identify whether the multispectral image is misaligned based on the position offset.

4. The method according to claim 3, characterized in that The object space coordinates corresponding to the image points of the panchromatic image and the image points of the multispectral image with the same name are calculated in S3, including: Calculate the object coordinates corresponding to the image points based on the RPC parameter file of the panchromatic image or multispectral image and the DEM elevation data; The formula is: ； ； In the formula, the polynomial The form is: ； In the formula, is the rational function coefficient, obtained through the RPC parameter file, and Zn is obtained through the DEM elevation data. , Represents pixel coordinates （r，c） and ground point spatial coordinates (X, Y, Z) Normalized coordinates after coordinate translation and scaling; The calculation of the object coordinates corresponding to the image points with the same name in the reference image in S13 includes: According to the 6-parameter file of the affine transformation of the reference image, the object space coordinates of the image points of the reference image are calculated. The affine transformation formula is: X = ax+by+c Y = dx + ey + f in X, Y is the object coordinate, x,y is the image coordinate, a, b, c, d, e, f It is the affine transformation 6 parameter file of the reference image, which is directly obtained from the reference image.

5. The method according to claim 3, characterized in that: Identify whether there is misalignment within the image, including: The object coordinate deviations calculated for the same-name points in multiple area blocks are averaged; When the average deviation of the object coordinates of the same-name points in the panchromatic image and the multispectral image is greater than the threshold of 10 meters, there is a misalignment inside the remote sensing image; When the average deviation of the object coordinates of the same-name points between the panchromatic image and the reference image is greater than the threshold of 20 meters, there is a misalignment inside the panchromatic image; When the average deviation of the object coordinates of the same-name points between the multispectral image and the reference image is greater than the threshold of 20 meters, there is misalignment inside the multispectral image.

6. A device based on the method according to any one of claims 1 to 5, characterized in that: include: Panchromatic image module, which obtains the panchromatic image corresponding to the remote sensing image and divides it into multiple area blocks; The multispectral image module obtains the multispectral image corresponding to the remote sensing image and divides it into multiple area blocks; The reference image module obtains the reference geographic information image corresponding to the remote sensing image as the reference image and divides it into multiple area blocks; The same-name image point search module finds the same-name image points corresponding to the target image based on the panchromatic image or multispectral image points to form a group of same-name points; An object coordinate calculation module calculates the position offset of the object coordinates of the image point and the corresponding image point with the same name; The misalignment determination module identifies whether there is misalignment inside the remote sensing image based on the position offset.

7. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the program, the steps of the method according to any one of claims 1 to 5 are implemented.

8. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 5 are implemented.

Images