An image mosaic method based on latitude and longitude grid division constraint

Abstract

This invention discloses an image mosaicking method based on latitude and longitude grid segmentation constraints, belonging to the field of surveying and remote sensing technology. The method includes: acquiring mosaicking range lines, standard map sheet data, and standard map sheet data containing map sheet row and column boundaries and inner map frame inflection points; generating a minimum rectangular range area for the mosaicking region using the mosaicking range lines; filtering image data to be mosaicked from the standard map sheet data; arranging the map sheet row and column boundaries according to row and column boundaries respectively; calculating the median value of adjacent boundaries; and writing the four extreme values ​​of the minimum rectangular range area for the mosaicking region into the sorting endpoints of the median values ​​of adjacent boundaries to obtain row and column segmentation data; generating intermediate segmentation lines based on the row and column segmentation data; and forming small rectangles by the intersection of the intermediate segmentation lines; converting the inner map frame inflection points within the small rectangles into image point coordinates to generate quarter points. This invention can stably constrain a single image source in the overlapping area under the condition of latitude and longitude grid standard map sheet image expansion and overlap.

Description

Technical Field

[0001] This invention relates to the field of surveying and remote sensing technology, specifically to an image mosaicking method based on latitude and longitude grid constraints. Background Technology

[0002] Latitude and longitude grid standard sheet image mosaic is widely used in surveying and mapping production, remote sensing image mapping and geographic information data updating. Organizing images according to unified sheet division rules is conducive to batch retrieval, result management and cross-sheet edge processing. Furthermore, the rectangular image range is easy to crop, load and stitch in image processing software.

[0003] Currently, in the process of batch mosaicking of standard latitude and longitude grid images, there are often situations where the inner map frame is not completely parallel as the latitude and longitude grid position changes, and the image results are stored in an outer rectangular range. If the overlapping areas are directly selected, covered or merged based on the image coverage range, it is easy to cause the edge boundary to deviate from the inner map frame constraint and to disturb the image results that have been edged.

[0004] Secondly, when the data source for overlapping areas switches between multiple standard image frames, the correspondence between the image area and the image frame source is easily weakened. This makes it difficult to determine which image frame a certain local area actually comes from when conducting subsequent results inspection, problem localization, or data backtracking, thus affecting the consistency and traceability of the data source of batch mosaic results. Summary of the Invention

[0005] To address the aforementioned technical problems, this invention provides an image mosaicking method based on latitude and longitude grid segmentation constraints.

[0006] An image mosaicking method based on latitude and longitude grid segmentation constraints includes: Collect mosaic range lines, standard map sheet data, and standard map sheet data including map sheet row and column boundaries and inner map frame inflection points. Generate the minimum rectangular range of the mosaic area through the mosaic range lines, and filter the image data to be mosaicked from the standard map sheet data. Arrange the map sheet row and column boundaries according to the row boundary and column boundary respectively, calculate the median value of adjacent boundaries, and write the four extreme values ​​of the smallest rectangular range of the mosaic region into the sorting endpoints of the median values ​​of adjacent boundaries to obtain the row and column segmentation data. Based on the row and column segmentation data, an intermediate dividing line is generated, and the intersection of the intermediate dividing lines forms a small rectangle. The inflection points of the inner outline within the small rectangle are converted into image point coordinates to generate quarter points. The small rectangle is divided by perpendicular lines drawn from the quarter points to the four sides. The divided regions are then matched with the source data of the image to be mosaicked to generate source results. Single-source image data is extracted from the image data to be mosaicked according to the source results, and the single-source image data is connected to generate a non-overlapping mosaic image.

[0007] Compared with the prior art, the beneficial effects of the present invention are as follows: This invention, through the division of labor constraints between standard latitude and longitude grid image data and standard map sheet data, ensures that the overlapping of the outer rectangle caused by the four boundaries of the image no longer directly determines the data selection boundary of the mosaic area, and that the map sheet row and column boundaries, row and column segmentation values, and quarter points jointly maintain the correspondence between the segmentation boundary and the inflection point of the inner map frame, thereby reducing the disturbance of the overlapping area to the already joined image results; Furthermore, this invention also constrains the correspondence between the segmented regions and the image data to be mosaicked by the source results, so that the single-source image data maintains a single image source when connected into non-overlapping mosaic images, thereby enhancing the consistency of data source of image regions in batch mosaic results and facilitating retrospective verification of non-overlapping mosaic images according to the source results.

[0008] In summary, this invention can stably constrain a single image source in the overlapping area under the condition of overlapping of standard latitude and longitude grid images, reduce disturbance of the already joined results, and improve the consistency and traceability of the data source of the mosaic results. Attached Figure Description

[0009] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in this invention. For those skilled in the art, other drawings can be obtained based on these drawings.

[0010] Figure 1 The flowchart illustrates an image mosaicking method based on latitude and longitude grid segmentation constraints provided by this invention. Detailed Implementation

[0011] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0012] Please see Figure 1 As shown, this embodiment discloses an image mosaicking method based on latitude and longitude grid segmentation constraints, the method comprising: S11: Collect mosaic range lines, standard map sheet data, and standard map sheet data including map sheet row and column boundaries and inner map frame inflection points. Generate the minimum rectangular range of the mosaic area through the mosaic range lines, and filter the image data to be mosaicked from the standard map sheet data.

[0013] In one specific embodiment, the mosaic range line is read from the work area boundary line, result clipping line, or task range line in the surveying and mapping production system. Standard sheet image data is orthophoto data produced according to the standard sheet division rules of latitude and longitude grid. Standard sheet image data includes the image's boundaries and image resolution values. Standard map sheet data is a combination of map sheet data generated according to the standard map sheet division rules of latitude and longitude grid. Standard map sheet data includes the map sheet row and column boundaries and the inflection points of the inner map frame.

[0014] It should be noted that the row and column boundaries of the map sheet are derived from the standard map sheet data of the latitude and longitude grid. Among them, the row boundary corresponds to the latitudinal map sheet boundary, and the column boundary corresponds to the longitudinal map sheet boundary. Because meridians converge with latitude, the inner map boundaries of standard map sheets are not always parallel to each other in the plane projection coordinate system. The inflection points of the inner map boundaries are used to characterize the intersection of the map boundaries of adjacent latitude and longitude grid map sheets. The image boundaries of standard sheet image data are the rectangular image range formed by expanding the standard sheet image. When multiple standard sheet image data are mosaicked, overlapping areas will occur. In this embodiment, the overlapping areas will be divided into single-source areas by row and column segmentation data and quarter points.

[0015] Specifically, the steps for filtering the image data to be mosaicked are as follows: S111: Extract the range boundary based on the tessellation range line, and determine the four extreme values ​​of the range surface according to the range boundary to generate the minimum rectangular range surface of the tessellation region.

[0016] In a specific embodiment, the horizontal and vertical coordinates of each boundary point in the mosaic range line are read in the Cartesian coordinate system. The minimum value among all horizontal coordinates is taken as the horizontal minimum extreme value of the range surface, the maximum value among all horizontal coordinates is taken as the horizontal maximum extreme value of the range surface, the minimum value among all vertical coordinates is taken as the vertical minimum extreme value of the range surface, and the maximum value among all vertical coordinates is taken as the vertical maximum extreme value of the range surface. The smallest rectangular range of the tessellation region is generated by the horizontal minimum extreme value, the horizontal maximum extreme value, the vertical minimum extreme value, and the vertical maximum extreme value of the range surface. The boundary points of the tessellation range line are represented as follows: ; in, Indicates the first in the tessellation range line A boundary point, Indicates the first The horizontal coordinates of the boundary points in a Cartesian coordinate system Indicates the first The vertical coordinates of the boundary points in a Cartesian coordinate system Indicates the boundary point number, , This indicates the number of boundary points within the tessellation range.

[0017] The four extreme values ​​of the range surface include: ; in, This represents the minimum extreme value in the lateral direction of the range surface. This represents the maximum extreme value in the lateral direction of the range surface. The vertical minimum extreme value of the range surface. This represents the maximum vertical extremum of the range surface.

[0018] By taking the extreme values ​​of the coordinates of all boundary points of the tessellation range line, the extreme values ​​of the four boundaries of the range surface are obtained, which are expressed as: ; ; ; ; in, This represents the function that takes the minimum value. This represents the function that takes the maximum value. The x-coordinates of all boundary points within the tessellation range are given. The coordinates are the longitudinal coordinates of all boundary points within the tessellation range.

[0019] The logic for generating the minimum rectangular range of the tessellation region based on the extreme values ​​of the range surface is as follows: , , and As four rectangular corner points, connect the four rectangular corner points in a clockwise or counterclockwise direction to generate the minimum rectangular range of the tessellation region.

[0020] S112: Extract the extreme values ​​of the four boundaries of the range surface from the minimum rectangular range surface of the mosaic region, and extract the four boundaries of the image from the standard sectional image data; In one specific embodiment, the image boundaries are read from each standard frame image data; The boundaries of the image include the minimum horizontal value, the maximum horizontal value, the minimum vertical value, and the maximum vertical value. It should be noted that the image's boundaries are the rectangular coverage area after the standard latitude and longitude grid image is expanded outwards, and are not equivalent to the inner map boundary in the standard map sheet data; No. The image boundaries of a standard frame image data sheet include: ; in, For the first The minimum horizontal dimension of a standard sheet image data set. For the first The maximum horizontal value of a standard-size image data sheet. For the first The minimum vertical dimension of a standard sheet image data set. For the first The maximum vertical value of a standard sheet image data set. Image number indicating standard frame image data. , This indicates the number of images in the standard sheet image data.

[0021] S113: Calculate the overlapping interval of the extreme values ​​of the four boundaries of the range surface and the four boundaries of the image to obtain the overlapping range of the image.

[0022] In one specific embodiment, the extreme values ​​of the range plane are calculated by performing horizontal and vertical interval overlap calculations on the image ranges of each standard frame image data to obtain the first... The image overlap range corresponding to the standard frame image data; No. The image overlap range corresponding to the standard-size image data includes: ; in, For the first The minimum horizontal value of the image overlap range corresponding to each standard-size image data sheet. For the first The maximum horizontal value of the image overlap range corresponding to each standard-size image data sheet. For the first The minimum vertical value of the image overlap range corresponding to each standard-size image data sheet. For the first The maximum vertical value of the image overlap range corresponding to each standard frame image data.

[0023] The method for calculating the image overlap range is as follows: ; ; ; ; in, , , and Originating from the extreme values ​​of the four boundaries of the range surface, , , and Source: The image boundaries of a standard frame image data.

[0024] when ,and When the condition is met, the image overlap range is not empty; when any condition is not met, the image overlap range is empty.

[0025] S114: Filter standard frame image data whose overlapping range is not empty to obtain the image data to be mosaicked.

[0026] In one specific embodiment, the image overlap range is determined frame by frame for all standard frame-by-frame image data. Standard frame-by-frame image data with non-empty image overlap ranges are retained as image data to be mosaicked, while standard frame-by-frame image data with empty image overlap ranges are excluded from the current mosaicking task. The image data to be mosaicked retains the image's boundaries and resolution values. The image's boundaries are used for subsequent source matching, and the image resolution values ​​are used for subsequent conversion of the inner contour inflection points to image point coordinates.

[0027] S12: Arrange the row and column boundaries of the map sheet according to the row and column boundaries respectively, calculate the median value of adjacent boundaries, and write the four extreme values ​​of the minimum rectangular range of the mosaic region into the sorting endpoints of the median values ​​of adjacent boundaries to obtain the row and column segmentation data; It should be noted that: because the inner boundary of the standard latitude and longitude grid map sheet is not completely consistent with the expanded rectangular image range, the row and column boundaries of the map sheet are not replaced by the image's four boundaries; When standard latitude and longitude grid images are mosaicked in a planar projection coordinate system, the outer rectangular image ranges will overlap. This step involves obtaining the intermediate value of adjacent boundaries from the map sheet row and column boundaries in the standard latitude and longitude grid map sheet data, and then adding the four extreme values ​​of the minimum rectangular range of the mosaic region to the sorting endpoints so that the subsequently generated intermediate dividing line covers the minimum rectangular range of the mosaic region.

[0028] Specifically, the steps to obtain row-column partitioned data are as follows: S121: Divide the row and column boundaries from the map sheet, and arrange the row and column boundaries according to their values ​​to generate boundary arrangement data.

[0029] In a specific embodiment, the map sheet row and column boundaries covering the area corresponding to the image data to be mosaicked are read from the standard map sheet data. The map sheet boundaries belonging to the row direction are divided into row boundaries, and the map sheet boundaries belonging to the column direction are divided into column boundaries. Based on row boundaries, data is arranged in ascending order of boundary values, and column boundaries are also arranged in ascending order of boundary values, generating boundary-arranged data.

[0030] Row boundaries in boundary-arranged data are represented as follows: ; in, This represents the row-direction boundaries after arranging the boundary values. This represents the value of the longitudinal boundary corresponding to the row boundary in a Cartesian coordinate system. This represents the number of row boundaries.

[0031] Column boundaries in boundary-arranged data are represented as follows: ; in, This is the column-to-boundary arrangement based on the boundary values. This represents the value of the transverse boundary corresponding to the column boundary in a Cartesian coordinate system. This represents the number of column-to-boundary boundaries.

[0032] S122: Calculate the median value of adjacent boundaries within the boundary arrangement data to obtain the median value of adjacent boundaries; Specifically, the steps to obtain the intermediate value of adjacent boundaries are as follows: S122.1: Read the adjacent row boundaries within the boundary arrangement data, calculate the mean of the row direction values ​​of the adjacent row boundaries, and obtain the row direction median value; The row-to-row median value is represented as: ; in, For the first Each row moves towards the middle value, This represents the mean value calculated from the row direction values ​​of two adjacent rows towards the boundary. Number the middle value of the row. Arrange the data within the boundary. Each row is directed towards the boundary. Arrange the data within the boundary. Each row is directed towards the boundary. .

[0033] S122.2: Read the adjacent column boundaries within the boundary arrangement data, calculate the mean of the column values ​​of the adjacent column boundaries, and obtain the column median value; The column median is represented as: ; in, For the first Each column is directed towards the middle value. This represents the mean value calculated from the column values ​​taken from the boundaries of two adjacent columns. Number the column median. Arrange the data within the boundary. Each column is directed towards the boundary. Arrange the data within the boundary. Each column is directed towards the boundary. .

[0034] S122.3: Arrange the row-to-row and column-to-row median values ​​according to the order of the data at the boundaries to obtain the median values ​​of adjacent boundaries; In one specific embodiment, all row intermediate values ​​are arranged according to the row boundary order, all column intermediate values ​​are arranged according to the column boundary order, and an intermediate value containing the row intermediate value and the column intermediate value adjacent to the intermediate value is constructed.

[0035] S123: Write the four extreme values ​​of the minimum rectangular range of the mosaic region into the sorting endpoints of the intermediate values ​​of the adjacent boundaries to obtain the row and column segmentation data; In a specific embodiment, the minimum and maximum vertical extreme values ​​of the range plane are written to the sorting endpoints of the row midpoint to obtain the row segmentation value; Write the horizontal minimum and maximum extreme values ​​of the range plane into the sorting endpoints of the column median values ​​to obtain the column split values, and construct row and column split data containing the row split values ​​and column split values; The row-direction segmentation value is represented as: ; in, The vertical minimum extreme value of the range surface. This represents the maximum extreme value in the longitudinal direction of the range surface. This is the midpoint of the row.

[0036] Column-oriented partition values ​​are represented as: ; in, This represents the minimum extreme value in the lateral direction of the range surface. This represents the maximum extreme value in the lateral direction of the range surface. This is the median value for a column.

[0037] Row and column segmentation data are constructed based on both row and column segmentation values; It should be noted that the row and column splitting data is subsequently used to generate intermediate dividing lines, which in turn form small rectangles covering the mosaic area.

[0038] S13: Generate intermediate dividing lines based on row and column segmentation data, and form small rectangles by the intersection of the intermediate dividing lines. Convert the inflection points of the inner map outline within the small rectangles into image point coordinates to generate quarter points. Specifically, the steps to generate the quarter points are as follows: S131: By splitting data by rows and columns, generate row-to-middle dividing lines and column-to-middle dividing lines.

[0039] In one specific embodiment, each vertical segmentation value in the row segmentation value is read, and the corresponding row middle segmentation line is generated in a Cartesian coordinate system; each horizontal segmentation value in the column segmentation value is read, and the corresponding column middle segmentation line is generated in a Cartesian coordinate system.

[0040] The line divider is represented as follows: ; in, Indicates the first Each row-direction partition value, Indicates the row-direction splitting value number. , Indicates the number of row-wise split values. It represents the vertical coordinate in a Cartesian coordinate system.

[0041] The column-to-middle dividing line is represented as: ; in, Indicates the first Each column-oriented partition value, Indicates the column-directed split value number. , Indicates the number of column-directed partition values. This represents the horizontal coordinate in a Cartesian coordinate system.

[0042] S132: The intersection of the row-to-middle dividing line and the column-to-middle dividing line generates a small rectangle.

[0043] In one specific embodiment, the intersection of the dividing lines from two adjacent rows to the middle and the dividing lines from two adjacent columns to the middle is used to obtain a small rectangle; Perform this process on all adjacent rows toward the center dividing line and all adjacent columns toward the center dividing line to generate a small rectangle covering the minimum rectangle range of the mosaic region. No. OK The smaller rectangle in the column is denoted as the first. The nth small rectangle, the nth The boundaries of the small rectangle include: ; in, For the first The horizontal value of the left boundary of each small rectangle is taken. For the first The horizontal value is taken from the right boundary of each small rectangle. For the first The lower boundary of each small rectangle is vertically oriented. For the first The vertical value of the upper boundary of each small rectangle is taken. , , OK The row and column positions of the small rectangles.

[0044] S133: Filter the inflection points of the inner outline that fall within the small rectangle to obtain the inflection points of the inner outline of the small rectangle.

[0045] In one specific embodiment, the inner map outline inflection point is read from the standard map sheet data. The inner map outline inflection point is the intersection point of the inner map outlines of adjacent map sheets in the latitude and longitude grid standard map sheet.

[0046] Read the horizontal and vertical coordinates of each inner outline inflection point in the Cartesian coordinate system, compare the coordinates of the inner outline inflection points with the four boundaries of the small rectangle, filter out the inner outline inflection points that fall within the small rectangle, and obtain the inner outline inflection points of the small rectangle.

[0047] No. The inflection point of the inner contour is represented as: ; in, Indicates the first Inflection point of the inner outline, For the first The horizontal coordinates of the inflection points of the inner contour in a Cartesian coordinate system. For the first The vertical coordinates of the inflection points of the inner contour in a Cartesian coordinate system. Indicates the inflection point number of the inner outline.

[0048] When the inflection point of the inner contour satisfies And satisfy At that time, the first The inner contour inflection point is used as the first The inflection point of the inner outline of the inner rectangle of the inner rectangle.

[0049] It should be noted that when an inflection point of the inner map outline falls near the boundary of a small rectangle due to coordinate accuracy errors, the inflection point filtering tolerance value is set according to the coordinate accuracy of the survey data. The inflection point filtering tolerance value is determined based on the coordinate accuracy of the map sheet and the data. When an inflection point of the inner map outline simultaneously meets the boundary inclusion condition of adjacent small rectangles, the inflection point of the inner map outline is written into the small rectangle corresponding to the intersection of the latitude and longitude grid where the inflection point of the inner map outline is located.

[0050] S134: Determine the origin of the image coordinates and the direction of the rows and columns of the image based on the four boundaries of the image data to be mosaicked, and convert the inflection points of the inner outline of the small rectangle into image point coordinates according to the image resolution value to generate the quarter points.

[0051] In one specific embodiment, the image data to be mosaicked is read, covering the inflection points of the inner outline of the small rectangle, and the image coordinate origin and image row and column directions are determined from the image boundary range of the image data to be mosaicked.

[0052] If the top left corner of the image data to be mosaicked is used as the origin of the image point coordinates, and the column direction increases from left to right and the row direction increases from top to bottom, then the inflection points of the map outline within the small rectangle are converted into image point coordinates based on the image boundary range and image resolution value, thus generating the quarter points.

[0053] Image resolution values ​​are expressed as: ; in, The image resolution value is the image data to be mosaicked, which is read from the standard frame image data;

[0054] Converting the inflection points of the inner outline of the small rectangle into image point coordinates, it is represented as follows: ; ; in, List the coordinates of the image points of the quarter points. The image point row coordinates of the quarter point. Here are the horizontal coordinates of the inflection points of the inner outline of the small rectangle. Here are the vertical coordinates of the inflection points of the inner outline of the small rectangle. The minimum horizontal value of the image data to be mosaicked, covering the inflection points of the map outline within the small rectangle. The maximum vertical value of the image data to be mosaicked, covering the inflection points of the map outline within the small rectangle.

[0055] The quarter point is represented as: ; in, It is a quarter point.

[0056] It should be noted that when the image data to be mosaicked uses the lower left corner as the origin of the image point coordinates, and the row direction increases from bottom to top, the image point row coordinates are changed from... Calculated; The image coordinate origin and image row and column directions are determined by the image boundaries and raster storage direction of the image data to be mosaicked. The quarter points retain the planar coordinates and image point coordinates corresponding to the inflection points of the map outline within the small rectangle. The planar coordinates participate in the subsequent vertical segmentation of the small rectangle, and the image point coordinates participate in the subsequent single-source image data cropping.

[0057] S14: Divide the small rectangle by perpendicular lines from the quarter points to the four sides, and match the source of the divided region with the image data to be mosaicked to generate source results. Extract single-source image data from the image data to be mosaicked according to the source results, and connect the single-source image data to generate a non-overlapping mosaic image.

[0058] Specifically, the steps to generate the source results are as follows: S141: Read the quarter point and the four boundaries of the small rectangle, and generate perpendicular lines from the quarter point to the four boundaries respectively to obtain the four perpendicular lines.

[0059] In one specific embodiment, the left, right, bottom, and top boundaries of the small rectangle are read, and the planar coordinates corresponding to the quarter points are read. Let the planar coordinates corresponding to the quarter points be: ; in, The horizontal coordinates of the plane corresponding to the quarter points. The vertical coordinates of the plane corresponding to the quarter points; The perpendicular lines from the quarter point to the left and right boundaries of the small rectangle are located at: ; The perpendicular lines from the quarter point to the lower and upper boundaries of the small rectangle are located at: ; in, This represents the perpendicular line passing through the quarter point and perpendicular to the lower and upper boundaries of the small rectangle. This represents the perpendicular line that passes through the quarter point and is perpendicular to the left and right boundaries of the small rectangle.

[0060] The four perpendicular lines are obtained by intersecting the two lines mentioned above within the small rectangle.

[0061] S142: Divide the small rectangles by four perpendicular lines to obtain the segmented region.

[0062] In one specific embodiment, the four perpendicular lines divide the small rectangle into four segmented regions; Let the horizontal value of the left boundary of the small rectangle be [value]. The horizontal value of the right boundary is The vertical value of the lower boundary is The vertical value of the upper boundary is The four divided regions are defined by the following boundaries: The first segmented region is composed of , , and limited; The second segmented region is composed of , , and limited; The third segmented region is composed of , , and limited; The fourth segmented region is composed of , , and limited.

[0063] It should be noted that the segmented regions do not overlap in the Cartesian coordinate system, and the four segmented regions within the same small rectangle collectively cover that small rectangle.

[0064] S143: Match the boundary of the segmented region with the image boundary range of the image data to be mosaicked to generate the source result.

[0065] In one specific embodiment, the region boundary of each segmented region is read, and the image boundary range of the image data to be mosaicked is read. The region boundary is matched with the image boundary range to determine the image data to be mosaicked corresponding to each segmented region, and the source result is generated.

[0066] Specifically, the logic for matching the boundary of the segmented region with the image boundaries of the image data to be mosaicked is as follows: S143.1: Read the region boundary of the segmented region and the image boundary range of the image data to be mosaicked, and perform intersection filtering on the region boundary and the image boundary range to obtain candidate image data to be mosaicked.

[0067] In a specific embodiment, let the first... The boundaries of the segmented regions include: ; in, For the first The minimum horizontal value of each segmented region For the first The horizontal maximum value of each segmented region For the first The vertical minimum value of each segmented region For the first The vertical maximum value of each segmented region This is the number of the segmented region.

[0068] The first The boundary of each segmented region is compared with the four boundaries of the image data to be mosaicked. The image data to be mosaicked that have intersection ranges are retained to obtain candidate image data to be mosaicked.

[0069] S143.2: Calculate the overlap area between the region boundary and the image boundaries of the candidate image data to be mosaicked, and divide the overlap area by the area of ​​the segmented region to obtain the region coverage value.

[0070] In one specific embodiment, the calculation of the first... The segmented region and the first The overlap area between the four boundaries of the candidate image data to be mosaicked is expressed as: ; in, For the first The segmented region and the first The overlap area value between candidate image data to be mosaicked. Number the candidate image data to be mosaicked.

[0071] No. The area value of each segmented region is represented as: ; in, For the first The area value of each segmented region.

[0072] The area coverage value is expressed as: ; in, For the first The segmented region is relative to the first The region coverage values ​​of the candidate image data to be mosaicked are within the range of [value missing]. .

[0073] S143.3: Filter candidate image data that meet the condition of complete regional coverage, and write the candidate image data and the segmented region to generate the source result.

[0074] In one specific embodiment, the condition for complete regional coverage is that the regional coverage value is greater than or equal to a preset coverage threshold. The preset coverage threshold is determined based on the accuracy of the image boundary data, and its value range is [insert range here]. ; When the When the region coverage value corresponding to the segmented region meets the condition of complete region coverage, the corresponding candidate image data to be mosaicked is compared with the first segmented region. Each segmented region is written to generate the source result. The source result includes the segmented regions and the corresponding image data to be mosaicked.

[0075] Specifically, the steps for generating a non-overlapping mosaic image are as follows: S144: Based on the source results, read the region boundaries of the segmented area and the corresponding image data to be mosaicked to obtain the single-source cropping range.

[0076] In a specific embodiment, the image data to be mosaicked corresponding to each segmented region is read from the source results, and the region boundary of the segmented region is read, and the region boundary is used as the single-source cropping range of the corresponding image data to be mosaicked.

[0077] S145: According to the single-source cropping range, crop the image content from the corresponding image data to be mosaicked to obtain the single-source image data.

[0078] In a specific embodiment, the single-source cropping range is converted into an image row and column index range based on the single-source cropping range, the image boundary range of the corresponding image data to be mosaicked, and the image content is cropped according to the image row and column index range to obtain single-source image data.

[0079] Let the single-source cutting range include: ; in, This represents the minimum horizontal value of the single-source cutting range. This represents the maximum horizontal value of the single-source cutting range. This represents the minimum vertical value of the single-source cutting range. This represents the maximum vertical value of the single-source cutting range.

[0080] When the top-left corner of the image data to be mosaicked is used as the origin of the image point coordinates, and the column direction increases from left to right while the row direction increases from top to bottom, the range of image row and column indices is represented as follows: ; ; ; ; in, To cut the minimum value of the column index, To cut the maximum value of the column index, The minimum value of the cropped row index. The maximum value of the cropped row index. This refers to the minimum horizontal value of the image data to be mosaicked. This corresponds to the maximum vertical value of the image data to be mosaicked. To correspond to the image resolution values ​​of the image data to be mosaicked, This indicates the floor function. This indicates the rounding up operation.

[0081] Based on the image row and column index range, image content is cropped from the corresponding image data to be mosaicked to obtain single-source image data. Single-source image data originates from only one corresponding image data set in the source results.

[0082] S146: Based on the row and column positions of small rectangles in the row and column segmentation data, single-source image data is stitched together to generate a non-overlapping mosaic image.

[0083] In one specific embodiment, according to the row and column positions of the small rectangles in the row and column segmentation data, the single-source image data within the same row position are stitched together in the column direction to obtain row-stitched image data. Then, according to the order of the row segmentation values, the image data of each row is stitched together to generate a non-overlapping mosaic image.

[0084] It should be noted that the row and column positions of the small rectangle have been explained in step S132. The boundary of the single-source image data comes from the segmented region, which is obtained by drawing perpendicular lines from the quarter points to the four sides of the small rectangle. The adjacent segmented regions do not overlap within the small rectangles, and the source results ensure that each segmented region corresponds to a single image data to be mosaicked. Therefore, each region in the stitched non-overlapping mosaic image corresponds to a single image source.

[0085] The above embodiments are only used to illustrate the technical methods of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical methods of the present invention without departing from the spirit and scope of the technical methods of the present invention.

Claims

1. An image mosaicking method based on latitude and longitude grid segmentation constraints, characterized in that, The method includes: Collect mosaic range lines, standard map sheet data, and standard map sheet data including map sheet row and column boundaries and inner map frame inflection points. Generate the minimum rectangular range of the mosaic area through the mosaic range lines, and filter the image data to be mosaicked from the standard map sheet data. Arrange the map sheet row and column boundaries according to the row boundary and column boundary respectively, calculate the median value of adjacent boundaries, and write the four extreme values ​​of the smallest rectangular range of the mosaic region into the sorting endpoints of the median values ​​of adjacent boundaries to obtain the row and column segmentation data. Based on the row and column segmentation data, an intermediate dividing line is generated, and the intersection of the intermediate dividing lines forms a small rectangle. The inflection points of the inner outline within the small rectangle are converted into image point coordinates to generate quarter points. The small rectangle is divided by perpendicular lines drawn from the quarter points to the four sides. The divided regions are then matched with the source data of the image to be mosaicked to generate source results. Single-source image data is extracted from the image data to be mosaicked according to the source results, and the single-source image data is connected to generate a non-overlapping mosaic image.

2. The image mosaicking method based on latitude and longitude grid segmentation constraints according to claim 1, characterized in that, The standard sectional image data includes the image's boundaries and resolution values.

3. The image mosaicking method based on latitude and longitude grid segmentation constraints according to claim 2, characterized in that, The steps for filtering the image data to be mosaicked are as follows: Extract the range boundary based on the mosaic range line, and determine the four extreme values ​​of the range surface according to the range boundary to generate the minimum rectangular range surface of the mosaic region; Extract the extreme values ​​of the four boundaries of the range surface from the minimum rectangular range surface of the mosaic region, and extract the four boundaries of the image from the standard sectional image data; The overlapping range of the image is obtained by calculating the extreme values ​​of the four boundaries of the range surface and the four boundaries of the image. Standard frame image data with non-empty overlapping areas are selected to obtain the image data to be mosaicked.

4. The image mosaicking method based on latitude and longitude grid segmentation constraints according to claim 1, characterized in that, The steps to obtain row-column split data are as follows: The map sheet is divided into row boundaries and column boundaries, and the row boundaries and column boundaries are arranged according to their values ​​to generate boundary arrangement data. Find the median value of adjacent boundaries within the boundary arrangement data to obtain the median value of adjacent boundaries; Write the four extreme values ​​of the smallest rectangular range of the mosaic region into the sorting endpoints of the intermediate values ​​of the adjacent boundaries to obtain the row and column segmentation data.

5. The image mosaicking method based on latitude and longitude grid segmentation constraints according to claim 4, characterized in that, The steps to obtain the intermediate value of adjacent boundaries are as follows: Read the adjacent row boundaries within the boundary arrangement data, calculate the mean of the row direction values ​​of the adjacent row boundaries, and obtain the row direction median value; Read the adjacent column boundaries within the boundary arrangement data, calculate the mean of the column values ​​of the adjacent column boundaries, and obtain the column median value; Arrange the row and column median values ​​according to the order of the data arranged by the boundaries to obtain the median values ​​of adjacent boundaries.

6. The image mosaicking method based on latitude and longitude grid segmentation constraints according to claim 2, characterized in that, The steps to generate quartiles are as follows: By splitting the data by rows and columns, a row-to-middle dividing line and a column-to-middle dividing line are generated; The intersection of the row-to-middle dividing line and the column-to-middle dividing line creates a small rectangle; Filter the inflection points of the inner outline that fall within the small rectangle to obtain the inflection points of the inner outline of the small rectangle; The origin and row / column directions of the image coordinates are determined based on the four boundaries of the image data to be mosaicked. The inflection points of the inner outline of the small rectangle are converted into image point coordinates according to the image resolution value, and the quarter points are generated.

7. The image mosaicking method based on latitude and longitude grid segmentation constraints according to claim 6, characterized in that, The steps to generate the source results are as follows: Read the quarter point and the four boundaries of the small rectangle, and generate perpendicular lines from the quarter point to the four boundaries respectively to obtain the four perpendicular lines; The small rectangle is divided by four perpendicular lines to obtain the segmented region; The boundaries of the segmented regions are matched with the image boundaries of the image data to be mosaicked to generate the source result.

8. The image mosaicking method based on latitude and longitude grid segmentation constraints according to claim 7, characterized in that, The logic for matching the boundary of the segmented region with the image boundary range of the image data to be mosaicked is as follows: Read the region boundaries of the segmented region and the image boundaries of the image data to be mosaicked, and perform intersection filtering on the region boundaries and image boundaries to obtain candidate image data to be mosaicked; Calculate the overlap area between the region boundary and the image boundaries of the candidate image data to be mosaicked, and divide the overlap area by the area of ​​the segmented region to obtain the region coverage value; Candidate image data that meet the condition of complete regional coverage are selected for mosaicking. The candidate image data is then written to correspond with the segmented regions to generate the source result.

9. The image mosaicking method based on latitude and longitude grid segmentation constraints according to claim 8, characterized in that, The steps to generate a non-overlapping mosaic image are as follows: Based on the source results, read the region boundaries of the segmented region and the corresponding image data to be mosaicked to obtain the single-source cropping range; According to the single-source cropping range, the image content is cropped from the corresponding image data to be mosaicked to obtain the single-source image data; Single-source image data is stitched together based on the row and column positions of small rectangles in the row and column segmentation data to generate non-overlapping mosaic images.

Images