CN112669459B

CN112669459B - Satellite image optimal mosaic line generation method based on feature library intelligent decision

Info

Publication number: CN112669459B
Application number: CN202011564393.0A
Authority: CN
Inventors: 丁一帆; 赵薇薇; 王战举; 王霜; 王艳; 周颖; 王红钢; 陈雪华; 韩冰; 杨劲林; 杨宇科; 董文军; 冯鑫
Original assignee: Beijing Aerospace Titan Technology Co ltd; Beijing Institute of Remote Sensing Information
Current assignee: Beijing Aerospace Titan Technology Co ltd; Beijing Institute of Remote Sensing Information
Priority date: 2020-12-25
Filing date: 2020-12-25
Publication date: 2023-05-05
Anticipated expiration: 2040-12-25
Also published as: CN112669459A

Abstract

The invention discloses a satellite image optimal mosaic line generation method based on intelligent decision of a feature library, which specifically comprises the steps of acquiring an effective overlapping area of a satellite image; firstly, acquiring an image effective area of a satellite image, and then acquiring an overlapping area between adjacent images; generating an initial mosaic reference line; an initial mosaic reference line is generated by adopting an automatic mosaic line extraction method based on pixel difference eclosion; an overlapping area dividing method based on a primary feature library; automatically generating a section embedded line based on a secondary feature library; the connection of the embedded lines among the sections; after generating a complete mosaic line, storing the mosaic line into a feature library, and taking the mosaic line as the mosaic line for the next mosaic in the area under the condition of small variation difference of ground features; and under the condition that the ground feature variation difference is large, taking the obtained complete embedded line as an embedded reference line, and repeating the steps. The method can better meet the production requirements of large-area DOM in the global scope with automation, rapidness and high quality and the related research requirements.

Description

Satellite image optimal mosaic line generation method based on feature library intelligent decision

Technical Field

The invention belongs to the technical field of remote sensing data processing and image mosaic, and particularly relates to a satellite image optimal mosaic line generation method based on intelligent decision of a feature library.

Background

The current large-area mosaic field of remote sensing satellite images lacks independent intellectual property and lacks rapid intelligent mosaic processing and data updating capability for multi-source satellite data, along with the continuous development of platforms with increased means, higher resolution and stronger functions, remote sensing satellites are continuously developed, and the acquired data volume is gradually increased in TB magnitude. For example, besides the traditional photogrammetry method, liDAR and InSAR technologies are new means for acquiring digital elevation models, and the development of various remote sensing data acquisition technologies accelerates the acquisition speed of high-resolution earth observation data and digital orthographic image (digital orthophoto map, DOM) data, but the current image mosaic line selection still has the problems of low efficiency, uneven quality, strong manual intervention, low automatic acquisition degree, low mosaic speed and the like, so how to quickly, reasonably and automatically acquire digital geographic information products with industrial significance, especially the automatic high-quality production of digital orthographic images in a large area range, has become an urgent problem to be solved in the fields of satellite remote sensing image update mosaic production and research.

Over the last decade researchers have developed over 30 damascene line automatic generation and optimization methods. These mosaic methods can be roughly classified into 4 types, including a method based on the difference of images in the overlapping region, a method based on auxiliary data, a method based on homonymies, a method based on morphology, and the like. In the method for generating the embedded line by image embedding, the following limitations are mainly existed at present: firstly, the automatic embedded line generation method based on the image difference of the overlapping area only considers the pixel difference, and cannot consider the geometric difference problem of the pixels with the same name of the actual ground object, so that the generated result is not true; secondly, the method based on the homonymy point does not consider image radiation information when generating the mosaic line, and the mosaic line obtained by the method easily passes through a building area, so that the obtained mosaic result is not ideal; thirdly, in the method based on auxiliary data, road vectors are easy to update and change and lack of behavior, in an actual image, the problem that a building is projected onto a road is caused due to shooting angles and the like, in addition, when point cloud data is adopted as the auxiliary data, the method relates to a Digital Surface Model (DSM) and optical image accurate matching technology, and has great technical difficulty; fourth, the geometric registration accuracy of the remote sensing image is higher by the morphology-based method, but the accuracy of the geometric registration accuracy is difficult to judge aiming at the classification of the ground features and the boundaries of the ground features, and the acquired mosaic lines still need to be continuously optimized by the conventional method based on the point cloud auxiliary data. Therefore, the existing method cannot completely avoid the characteristics of the high-altitude ground feature of the building and cannot ensure the integrity of the ground feature target element, and more parts need to be manually edited in the DOM generation process, so that the improvement of the generation quality of the mosaic line and the automation degree of image update mosaic are the problems to be solved urgently at present.

In the existing automatic mosaic line generation method, for example, the automatic mosaic line extraction method based on the image difference of the overlapping area mainly comprises the following steps: 1) Calculating pixel differences, local texture differences and projection differences; 2) Mosaic line selection is performed according to a search strategy, wherein the search strategy comprises a progressive search, a dynamic planning method, a shortest path method, an optimal spanning tree method, a dynamic contour model method and the like, and the seam line is obtained according to the search strategy. However, this type of search strategy has several major problems: 1) The line-by-line search cannot specify the starting point and the end point of the seam line; 2) The path trend of dynamic planning is not globally optimal, and the embedded line can be selected only according to the horizontal trend and the vertical trend; 3) The calculation amount of the exhaustion method of the shortest path method is extremely large; 4) The optimal spanning tree method cannot guarantee that the starting point and the ending point of the initial mosaic line can be prevented from falling on a building; 5) The dynamic profile selection policy may lose the global minimum.

In summary, although the current mosaic line generation methods are various, the search strategy speed of most mosaic methods is slow, and the large area mosaic is generated by repeated feature classification, most of the methods do not have intelligent decision effect, the situation is weak, and no research on the mosaic line generation method based on intelligent decision of a feature library is currently performed, so that the calculation amount of the mosaic line optimization strategy generated by the existing method is large and the universality is to be enhanced, and the mosaic line generation method applied to the area mosaic with other large-scale and multi-feature is worthy of further research. At present, a satellite image region mosaic method based on intelligent decision of a feature library is not applied to rapid intelligent mosaic of a large-scale multi-feature region, and particularly, the most important optimization of a mosaic line rapid generation method in the mosaic process is urgently researched. According to the method, the search strategy based on the feature library intelligent decision method is adopted to generate the optimal embedded line, the rapid automatic selection function is achieved, the generated embedded line is put in storage, the embedded efficiency for the next area is improved, and under the condition that the change difference of ground features in the area is small.

Disclosure of Invention

Aiming at the problems that the searching strategy speed of the mosaic line generating method of the existing remote sensing satellite image is slower, the large-area mosaic can be generated by repeated ground object classification for many times, the intelligent decision effect is not achieved, and the situation is weak, the invention provides a method for generating the optimal mosaic line based on the segmentation selection of a primary feature library and the mosaic line selection strategy of a secondary feature library, and aims to improve the calculation efficiency of mosaic line generation, promote the searching speed, maximally enable the mosaic line to be generated reasonably, maximally enable the mosaic line to pass through the building and the area with larger geometric difference along the road vector data in the secondary feature library. Compared with the traditional mosaic line generation method, the method is a rapid intelligent mosaic method which fully utilizes feature data of the feature classification, the calculated amount in the mosaic process is small, the mosaic quality is high, and the strategy of the feature library can be suitable for various feature type image segments, and has good universality and expansibility.

The invention discloses a satellite image optimal mosaic line generation method based on intelligent decision of a feature library, which comprises the following specific steps of,

S1, acquiring an effective overlapping area of a satellite image; firstly, acquiring an image effective area of a satellite image, and then acquiring an overlapping area between adjacent images;

the method for acquiring the image effective area of the satellite image comprises the following steps:

if the obtained satellite image is a multiband image, adding multiband pixel values of each pixel point in the satellite image, and storing the multiband pixel values, and if the added value is 0, judging that the added value is an invalid value;

if the obtained satellite image is a single-band image, sequentially circulating from the first row to the subsequent row according to the number of rows and columns of the image, judging each pixel value from left to right in each row, continuously judging the next pixel value if the value is 0, recording the corresponding row number and column number of the first non-0 pixel value, setting the point as an upper left corner C point, acquiring the upper left corner, sequentially circulating from the last row to the last row, judging whether each pixel value is 0 from right to left in each row, and setting the point as a lower right corner D point when the first non-0 pixel value is encountered; and continuously circulating each column of the image from left to right, judging each pixel value from bottom to top in each column, continuously judging the next pixel value if the value is 0, recording the corresponding row number column number when the first non-0 pixel value is judged, setting the pixel as a lower left corner B point, circularly reading the image column from right to left from the rightmost column after the lower left corner point is obtained, judging each pixel value from top to bottom in each column, and setting the pixel as an upper right corner A point when the first non-0 pixel value is encountered.

And creating a vector file according to the obtained row and column numbers of the four corner points, and completing the acquisition of the effective area of the satellite image.

And acquiring an overlapping region between adjacent images, acquiring phase-amplitude data vectors corresponding to the two adjacent images, respectively overlapping vector data of the two images, and then calculating an overlapping edge of the two images to acquire an effective overlapping region range which is ACBD.

S2, generating an initial mosaic reference line.

According to the overlapping area between the adjacent images obtained in the step S1, an initial mosaic reference line is generated by adopting an automatic mosaic line extraction method based on pixel difference eclosion, and the initial mosaic reference line is represented by adopting a coordinate sequence in the overlapping area;

the pixel difference eclosion-based mosaic line automatic extraction method is realized by adopting a Dijkstra method, wherein the Dijkstra method is used for solving the shortest path of the minimum cost (cost) between two points, the cost is replaced by a pixel path with weight, the brightness difference delta I (I, j) and the gradient of an overlapped area are used for distinguishing normalized pixel values s (I, j) as pixel difference values of the overlapped area, the gray gradient difference delta g is the absolute value of the gradient value difference, the gradient value is the change value g of the gray scale of an image in the horizontal and vertical directions, and the gradient direction is

The specific calculation process of the brightness difference Δi (I, j) is as follows:

ΔIΔI(i，j)＝|imgl(i，j)-img2(i，j)|，

wherein img1 (i, j), img2 (i, j) respectively represent brightness values of points (i, j) of the overlapping area on two adjacent images to be inlaid; gradient value magnitude g (i, j) and gradient direction of point (i, j) of overlap region

The calculation formula of (2) is as follows:

g(i，j)＝[(img(i，j)-img(i+1，j) ² +(img(i，j)-img(i，j+1)) ² )] ^1/2

img (i, j) represents the luminance value of the point (i, j) of the overlap region, and the gradation gradient difference Δg is represented as:

wherein Δg (i, j) represents the gray gradient difference value at the pixel point (i, j), g1 (i, j) represents the gradient value of the pixel point in one of the two adjacent images, g2 (i, j) represents the gradient value of the pixel point in the other of the two adjacent images,

respectively representing the gradient directions corresponding to the two pixel points. The maximum value of the normalized brightness difference or gray gradient difference represents the maximum difference value of the pixels in the overlapping region, the value is max (delta I (I, j), delta g (I, j)), the corresponding pixel points are called pixel difference points, the set of the pixel difference points is CC, the upper right corner of the boundary of the image in the overlapping region is taken as the starting point A of the overlapping region, the coordinates of the starting point A are (x 0, y 0), and the starting point A is from the starting point A to all the pixel difference points a in the overlapping region _i E CC, the shortest path in the overlap region is denoted Dist [ A, a ] _i ]And i is a subscript, and then:

Dist[A，a _i ]≥short[A，a _i ]，

wherein short [ A, a ] _i ]Representing the difference point a from the start point A of the overlapping region to all pixels _i E shortest path of CC; by continuously calculating Dist [ A, a ] _i ]Up to a _i To the point B at the left lower corner of the overlap region in step S1, at which time Dist [ A, a ] _i ]＝short[A，a _i ]And obtaining an initial mosaic reference line according to the mosaic line generation method based on the pixel feathering difference.

S3, an overlapping area dividing method based on a primary feature library comprises the following specific steps:

s31, selecting and grading a feature library;

and classifying the ground feature characteristics in the overlapping area in the acquired satellite image, and establishing a primary characteristic library and a secondary characteristic library, wherein the primary characteristic library is used for determining segmentation limit, and the secondary characteristic library is used for determining the selection of the embedded line. Selecting a primary feature library and a secondary feature library from the feature library data, wherein:

the first-level feature library comprises dense boundary vector data and sparse boundary vector data;

the secondary feature library comprises road vector data, water system vector data, abrupt bank data and the like.

Dividing the overlapping region into a dense segment, a sparse segment and a mixed segment according to dense boundary vector data and sparse boundary vector data in a primary feature library according to feature types and feature library data established by various remote sensing geographic information resources, obtaining the dense segment according to the dense boundary vector data in the overlapping region according to the step S32, obtaining the sparse segment according to the sparse boundary vector data in the overlapping region according to the step S33, wherein the regions except the dense segment and the sparse segment are collectively called the mixed segment, and a specific segmentation method is shown in the steps S32 and S33; the dense boundary vector data establishes a feature library according to data (first level and second level) of buildings, high-level roads and the like, noctilucent remote sensing data, water system vectors and the like, and the sparse boundary vector data establishes a feature library according to desert vectors, DEM data, highways, low-level road data (third level and fourth level), water system vectors and the like.

S32, dense segment segmentation, which specifically comprises:

s321, firstly judging whether dense boundary vector data exist in an overlapping area, if so, segmenting according to the step S322, and if not, directly selecting an initial mosaic reference line in the step S2 without segmenting;

s322, constructing a minimum circumscribed rectangle according to dense boundary vector data in a primary feature library, wherein the specific construction process is as follows:

the minimum circumscribed rectangle is drawn by using a minAreRect () function meter in opencv, and the minimum circumscribed rectangle comprises four endpoints and four edges: firstly, defining an circumscribed rectangle set as a circumscribed rectangle set formed by the contours of dense boundary vectors and sparse boundary vectors in an overlapping area ABCD in the step S1, and defining a minimum circumscribed rectangle set as a set formed by the minimum circumscribed rectangles of the contours of each dense boundary vector and sparse boundary vector; the outline here refers to the set of points within the overlap region ABCD obtained in step S1 corresponding to the boundary vector; drawing a center point of the minimum circumscribed rectangle to obtain four end points of the minimum circumscribed rectangle; finally, drawing four sides of the minimum circumscribed rectangle to finish drawing the minimum circumscribed rectangle;

and rotating the polygon formed by the endpoints of the original dense boundary vector data, obtaining a simple circumscribed rectangle of the polygon after rotating each degree from-90 degrees to 90 degrees, recording the area, vertex coordinates and rotation degree of each simple circumscribed rectangle, obtaining the circumscribed rectangle with the minimum area, and finally rotating the circumscribed rectangle by the rotation angles with the same magnitude in the opposite direction to the recorded rotation degree, thereby obtaining the minimum circumscribed rectangle.

S323, in the case where the minimum bounding rectangle obtained in step S322 intersects or contains the bounding rectangle of the remaining boundary vectors, the overlapping area in step S1 is segmented according to the dense boundary vector.

S324, expressing the minimum circumscribed tangent equation as follows according to the upper and lower boundaries of the minimum circumscribed rectangle of the dense boundary vector in the step S322: y=kx+b, and according to the coordinates of the four end points of the minimum circumscribed rectangle, the equation for solving the upper and lower two minimum circumscribed tangents is:

y ₁ =k ₁ x+b ₁

y ₂ =k ₂ x+b ₂ ，

wherein the slope k ₁ =k ₂ ，b ₁ And b ₂ Is a constant term of two minimum circumscribed tangent lines.

S33, sparse segment segmentation, which specifically comprises:

firstly judging whether the overlapped area has sparse boundary vector data, if so, constructing a minimum circumscribed rectangle according to the step S322, and obtaining a lower boundary intersection point S by intersecting the overlapped area boundary with the known sparse segment boundary, wherein for the sparse segment segmentation boundary y30, the expression is y ₃ =k ₃ x+b ₃ Its slope k ₃ =k ₁ =k ₂ Substituting the coordinates of the lower boundary intersection point S into the expression of y30 to calculate b ₃ Values, thereby obtaining an expression of the sparse segment segmentation boundary y 30; if the sparse boundary vector data does not exist, the initial mosaic reference line in the step S2 is directly selected without segmentation, and the rest segments except the dense segment and the sparse segment are collectively called a mixed segment.

S4, matching road vector, water system vector data and image pixel data, namely, pixel matrix of the overlapping region image in the secondary feature library, wherein the method specifically comprises the following steps:

the three-parameter method is to translate the X, Y and Z axes between two reference planes, and convert the road vector and the water system vector coordinate of the input overlapping region image by the three-parameter method according to the three parameters and the reference points of the translation to obtain the output vector geographic coordinate data, and the calculation is as follows:

wherein [ X, Y, Z] _new For the vector data obtained, [ X, Y, Z ]] _orgin Is the original vector coordinate, [ dX, dY, dZ ]]Is three parameters of translation.

And matching the road vector data and the water system vector data with the geographic entity coordinate points to the overlapping region image according to the prior feature data of the secondary feature library, and matching the same-name entities. The positions of the pixel points in the image are represented by pixel coordinates, when matching the road with the water system vector data and the image pixels in the overlapping area, the corresponding geographic coordinates are required to be converted into the pixel coordinates, and the conversion method comprises the following steps:

first, the geographic coordinates of the upper left corner of the overlapping region image [ GEO ] _x ，GEO _y ]Known, and resolution R of image _x ，R _y It is known to calculate the geographical coordinates of arbitrary pixel locations from the geographical coordinates of the upper left corner of the image and the resolution of the image. The formula for converting the geographic coordinates into pixel coordinates is:

The pixel coordinate is converted into a geographic coordinate formula:

X＝GEO _x +col*R _x

Y＝GEO _y +row*R _y ，

wherein X and Y in the two-coordinate conversion formula represent geographic coordinates, PIX _x ，PIX _y Representing pixel coordinates, GEO _x 、GEO _y Representing the geographical coordinates of the upper left corner of the image, col, row representing the number of columns and rows of image pixels, R _x And R is R _y Representing the row and column resolution of the image. And matching the road vector with the A point coordinate as a starting point to the image pixel, and preparing the road vector as data generated by a subsequent mosaic line.

S5, automatically generating a section embedded line based on a secondary feature library, wherein the method specifically comprises the following steps of:

s51, adopting an automatic mosaic line generation method based on a water system vector and a road vector feature vector for a dense segment, wherein the method comprises the following specific steps of:

when no secondary feature library data exists, the embedded line is an initial embedded reference line in the step S2;

when the secondary feature library data exist, acquiring an embedded line by adopting an automatic embedded line selecting method based on a water system vector and a road vector:

the starting point of the optimized embedded line in the dense section is A ₂ And fitting the water system or road vector coordinates of the secondary feature library by selecting N lower intersection points of the water system or road vector curve in the overlapping area and the dense section lower boundary straight line y20, namely the dense section road end point, so as to obtain the water system or road vector curve, wherein the water system vector curve is as follows:

Wherein ρ is ₀ ，ρ ₁ ，…，/>

Polynomial coefficients of the water system vector curve, N ₀ The order of the water system or the road vector curve is as follows:

wherein omega ₀ ，ω ₁ ，…，/>

Polynomial coefficients for a road vector curve;

s511, if the secondary feature library has water system vector data, integrating the water system vector curve f ₁ (x)，f ₂ (x)，f ₃ (x)，...，f _N (x) The method is combined with a straight line y20 to form an equation set, and an intersection point set is solved to obtain a water system vector endpoint set in the dense segment;

s512, if the secondary feature library has no water system vector data, the road vector curve h is obtained ₁ (x)，h ₂ (x)，...，h _N (x) The curve set and the straight line y20 are combined to form an equation set, and the end point set of the road vector in the dense section is obtained through solving.

S513, taking the intersection point of the overlapping area reference mosaic line obtained in the step S2 and the dense section upper boundary straight line y10 as a starting point, taking the continuous water system or road vector points which penetrate through the dense section upper and lower boundaries on the dense section lower boundary straight line y20 as end points, searching through water systems or roads according to constraint conditions, searching M water systems or roads, wherein N water systems or roads (N is less than or equal to M) are water systems or road vectors containing the end points until the continuous water systems or road vector lines which penetrate through the dense section upper and lower boundaries are searched at the end points on the dense section lower boundary straight line y20, selecting the searched continuous water systems or road vector lines which penetrate through the dense section upper and lower boundaries as optimal mosaic lines, if the penetrating water systems or roads are not searched all the time, cutting off the road or water system vector data, and selecting the initial mosaic reference line in the step S2 as the optimal mosaic lines.

S514, selecting an optimal embedded line under the condition of penetrating the water system or road vector characteristic data;

the process of judging the penetration of the water system or road vector characteristic data is that the discrete point coordinates of each water system or road are fitted into a water system or road vector curve equation, the starting point and the ending point coordinates of the boundary points of the water system or road are brought into y10 and y20 equations of boundary straight lines, when the end point coordinates of the water system or road vector are on the boundary straight lines, the water system or road vector is judged to penetrate through the upper and lower boundaries of dense or sparse sections, the penetration road or water system vector data is selected, if the end point coordinates of the water system or road vector are not on the boundary straight lines, the road or water system vector data is omitted, and the initial embedding line in the step S2 is selected as the embedding line of the sections.

The process for selecting the inlaid wire comprises the following steps: selecting a vector of the shortest path which penetrates through according to the sequence of the water system, the high-grade road and the low-grade road; according to the distance between the end point of the lower boundary of the dense section of the water system or the road vector line and the intersection point of the lower boundary of the dense section of the reference embedded line in the overlapping area, selecting the water system or the road vector corresponding to the end point of the shortest distance as the optimized embedded line according to the principle of the shortest distance priority, and selecting the distance between the end point of the lower boundary of the dense section of the water system or the road vector line and the intersection point of the lower boundary of the dense section of the reference embedded line

D as a constraint of closest distance between the water system or road vector and the reference line _i Representing the distance between the end point of the lower boundary of the ith water system or road vector line at the dense segment and the intersection point of the lower boundary of the reference embedded line at the dense segment, N ₁ The number of the searched through water system or road vector is set. The calculation method of the water system or the road path length comprises the following steps: equally dividing the water system or road curve penetrated in the step S51 into n equal parts according to the actual road vector length, and solving the total curve length of the equally divided curve by adopting a short straight line method, wherein the method comprises the following specific steps:

wherein L represents the total curve length, deltax _i For the arc length in the ith subinterval to correspond to the transverse axis length, deltay _i For the corresponding longitudinal axis length of the arc length in the ith subinterval, x _i Is the abscissa of any point in the ith subinterval, i is the number of curve divisions, where i=1, 2, …, n, f' ² (x _i ) Is the water system or the road curve equation f (x) at x _i The derivative square of the points, and finally, n equal divided curve lengths are accumulated to obtain a total curve length L, namely, the corresponding water system or road path length penetrating through the upper boundary and the lower boundary of the dense segment, and the curve lengths of all water systems or road vectors in the dense segment are calculated to ensure that L _min ＝mm{L ₁ ，L ₂ ，...，L _N Selecting L _min As a water system or road vector shortest path constraint. Obtaining an optimal mosaic line segment A according to the method in the step S514 ₂ A ₅ 。

S515, the shortest water system or road vector line end point coincides with the reference mosaic line lower boundary end point coordinates at the dense segment or d _min Under the condition that the distance between the shortest water system or road vector line end point and the embedded reference line end point in the step S2 is directly connected with the end point of the embedded reference line in the dense section lower boundary y20 at the distance of less than or equal to 10m, a buffer circle is established by taking the distance between two intersection points of the shortest water system or road vector line and the embedded reference line at the dense section lower boundary as the diameter of a circle, the midpoint of the two points is the center of the circle, and the line segment A is optimally embedded in the step S514 in the circle ₂ A ₅ Intersection point A with dense segment lower boundary y20 ₅ (xx 5, yy 5) as a starting point, searching for a water system or a road vector, and selecting a mosaic line A in the segment according to steps S511 to S514 ₅ A ₃ 。

S516, obtaining the dense-segment embedded line as the optimal embedded line according to the steps from S511 to S515.

S52, aiming at the sparse segment, when no secondary feature library vector exists, reserving the original initial mosaic reference line in the step S2 as a mosaic line segment of the area;

when the secondary feature library vector exists, an automatic mosaic line generating method based on road and water system vectors is adopted, and the method specifically comprises the following steps:

s521, obtaining the optimal embedded line AA with the road priority according to steps S511-S515 by taking the point A as the starting point according to the road-first and water-second through road selection mode in the sparse land section ₄ If there is no road or water system satisfying the condition in step S514, the initial mosaic reference line in step S2 is selected.

S522, the shortest road or water system vector line end point coincides with the reference embedded line at the sparse segment boundary coordinates or d _min Under the condition of less than or equal to 10m, directly connecting the shortest road or water system vector line end point in the lower boundary y30 of the sparse segment with the embedded reference line end point in the step S2; distance d between end point and reference line _min When the distance between the shortest water system or road vector line and the reference embedded line is more than 10m, the distance between two intersection points of the lower boundary of the sparse segment is taken as the diameter of a circle, the midpoint of the two intersection points is taken as the center of the circle, a buffer circle is established, and the optimal embedded line segment AA in the sparse segment is taken in the circle ₄ Intersection point A with sparse segment boundary y30 ₄ As a starting point, searching for water system or road vector, selecting the inlaid strand A according to steps S511-S514 ₄ A ₁ 。

S523, obtaining AA according to the mosaic lines in the penetrating sparse segment obtained in the steps S521 to S522 as the optimal mosaic lines ₄ 、A ₄ A ₁ As the optimal mosaic line segment.

S53, aiming at the generation of embedded lines in the mixed segment except the dense segment and the sparse segment, respectively selecting intersection points of the initial embedded reference line and the sparse segment in the step S2, and taking the initial embedded reference line or the intersection point of the boundary between the initial embedded reference line and the dense segment as a starting point and an end point of an embedded line segment, and taking the initial embedded reference line in the step S2 as the embedded line.

S6, connecting embedded wires among the sections.

The mixing section uses an initial mosaic reference line; the dense segment uses the feature vector optimal mosaic line segment A obtained in the step S51 ₂ A ₅ 、A ₅ A ₃ The method comprises the steps of carrying out a first treatment on the surface of the Sparse segment uses the optimal mosaic line AA acquired in step S52 ₄ 、A ₄ A ₁ . And finally merging the end points of the embedded lines between the sections on the sectional boundary lines of the mixed section, the sparse section and the dense section into an optimal embedded line vector which is complete in the overlapping area.

S7, after the complete mosaic line is generated, storing the mosaic line into a feature library, and taking the mosaic line as the mosaic line for the next mosaic in the area under the condition of small difference of the ground feature changes; under the condition that the ground feature variation difference is large, taking the obtained complete mosaic line as a mosaic reference line, continuing to perform the mosaic line optimization algorithm of the steps S1 to S6, and further improving the mosaic line generation efficiency in regional mosaic.

The beneficial effects of the invention are as follows:

1. compared with the traditional mosaic line generation method, the mosaic line generation method has the advantages that the mosaic line is built based on the feature library, the existing remote sensing resource data can be fully utilized, the feature classification data such as road vector data with strong behavior can be fully utilized, the feature classification data and other data in the corresponding feature library can be updated in real time, compared with other mosaic line generation methods, the mosaic line generation method reduces the calculation amount of the mosaic line generation and improves the mosaic quality, has strong behavior, fully and reasonably utilizes the integrated geographic information data resource and remote sensing data resource, and has the advantage of higher information availability.

2. Compared with the traditional method, the method for embedding the mosaic lines in the areas of the satellite images has higher accuracy and stronger universality, for example, the method can provide an intelligent strategy for the existing satellite images to be embedded, and intelligent decision selection of the mosaic lines is carried out according to the feature library data of the object areas, so that the method is beneficial to solving the problem that the mosaic lines pass through high-altitude ground features such as buildings, can ensure the quality precision of the selected mosaic lines, and has a certain practical significance for generating high-quality and automatic digital orthographic products (DOM) in the whole world, particularly in a large area range.

3. The method of the invention is aimed at the secondary region mosaic of the same region of different source data, and can directly utilize the existing mosaic line in the library to perform region mosaic or be used as the mosaic line reference line of the region mosaic. After the mosaic line search is completed, the mosaic line is put in storage, and the mosaic lines generated by the method are inlaid in the same area for different source data, so that the mosaic line generated by the method can be used as a reference mosaic line of the area next time, or under the conditions that the ground feature of the area is not greatly changed and the overlapping areas of images are generally consistent, the existing mosaic lines in the storage can be directly utilized for rapid and intelligent area mosaic.

Drawings

FIG. 1 is a flow chart of a method for selecting an optimal mosaic line based on a feature library;

FIG. 2 is a schematic diagram of an image to be mosaic and an overlapping area;

FIG. 3 is a schematic diagram of an overlapping region reference mosaic line;

FIG. 4 is a schematic view of a minimum bounding rectangle;

FIG. 5 is a segmented schematic diagram of an overlap region;

FIG. 6 is a schematic diagram of the relationship between two reference planes between three parameters;

FIG. 7 is a schematic diagram of a transformation of geographic coordinates and pixel coordinates;

FIG. 8 is a schematic diagram of a dense segment and sparse segment tessellation line generation process;

fig. 9 is a graph of optimal mosaic line merging effect.

Detailed Description

For a better understanding of the present disclosure, an embodiment is presented herein.

The embodiment discloses a satellite image optimal mosaic line generating method based on intelligent decision of a feature library, wherein a flow chart of the method is shown in figure 1, and the specific steps comprise,

s1, acquiring an effective overlapping area of a satellite image; firstly, an image effective area of a satellite image is obtained, then an overlapping area between adjacent images is obtained, and an image to be inlaid and an overlapping area schematic diagram are shown in fig. 2.

For the acquisition of the overlapping area between the adjacent images, the corresponding phase-amplitude data vectors of the two adjacent images are acquired, the respective vector data of the two images are respectively overlapped, the overlapping edges of the two images are calculated, and the effective overlapping area range is ACBD, as shown in fig. 2.

S2, generating an initial mosaic reference line.

According to the overlapping area between the adjacent images acquired in the step S1, an initial mosaic reference line is generated by adopting an automatic mosaic line extraction method based on pixel difference eclosion, and the initial mosaic reference line is represented by adopting a coordinate sequence in the overlapping area, wherein the coordinate sequence is as follows:

(x1,y1),(x2,y2),(x3,y3),...,(xi,yi)，

wherein (x 1, y 1) is the value of the first coordinate point in the coordinate sequence, the values of the subsequent coordinate points are analogized in sequence, and (xi, yi) is the value of the ith coordinate point in the coordinate sequence;

ΔI(i，j)＝|img1(i，j)-img2(i，j)|，

The calculation formula of (2) is as follows:

g(i，j)＝[(img(i，j)-img(i+1，j) ² +(img(i，j)-img(i，j+1)) ² )] ^1/2

respectively representing the gradient directions corresponding to the two pixel points. The maximum value of the normalized brightness difference or gray gradient difference represents the maximum difference value of the pixels in the overlapping region, the value is max (delta I (I, j), delta g (I, j)), the corresponding pixel points are called pixel difference points, the set of the pixel difference points is CC, the upper right corner of the boundary of the image in the overlapping region is taken as the starting point A of the overlapping region, the coordinates of the starting point A are (x 0, y 0), and the starting point A is from the starting point A to all the pixel difference points a in the overlapping region _i E CC, the shortest path in the overlap region is denoted Dist [ A, a ] _i I is a subscript, and there are:

Dist[A，a _i ]≥short[A，a _i ]，

wherein short [ A, a ] _i ]Representing the difference point a from the start point A of the overlapping region to all pixels _i E shortest path of CC; by continuously calculating Dist [ A, a ] _i ]Up to a _i To the point B at the left lower corner of the overlap region in step S1, at which time Dist [ A, a ] _i ]＝short[A，a _i ]Obtaining an initial mosaic reference line according to the mosaic line generation method based on the pixel feathering difference, and obtaining an initial mosaic reference line schematic diagram by taking the overlapping area obtained in the step S1 as input is shown in fig. 3.

s31, selecting and grading a feature library;

S32, dense segment segmentation, which specifically comprises:

the minimum bounding rectangle, as shown in fig. 4, is drawn by using a minarea rect () function meter in opencv, including four endpoints and four sides of the minimum bounding rectangle: firstly, defining an circumscribed rectangle set as a circumscribed rectangle set formed by the contours of dense boundary vectors and sparse boundary vectors in an overlapping area ABCD in the step S1, and defining a minimum circumscribed rectangle set as a set formed by the minimum circumscribed rectangles of the contours of each dense boundary vector and sparse boundary vector; the outline here refers to the set of points within the overlap region ABCD obtained in step S1 corresponding to the boundary vector; drawing a center point of the minimum circumscribed rectangle to obtain four end points of the minimum circumscribed rectangle; finally, drawing four sides of the minimum circumscribed rectangle to finish drawing the minimum circumscribed rectangle; the specific formula is as follows: the point (x 1, y 1) on the plane of the overlapping area after being rotated counterclockwise by an angle θ around the other point (x 0, y 0) is (x 2, y 2), and then the following formula is given:

x2＝(x1-x0)cosθ-(y1-y0)sinθ+x0

y2＝(x1-x0)sinθ+(yl-y0)cosθ+y0，

S323, in the case where the minimum bounding rectangle obtained in step S322 intersects or contains the bounding rectangle of the remaining boundary vectors, the overlapping area in step S1 is segmented according to the dense boundary vector. Fig. 5 is a segment schematic diagram of an overlapping region.

y ₁ ＝k ₁ x+b ₁

y ₂ ＝k ₂ x+b ₂ ，

wherein the slope k ₁ ＝k ₂ ，b ₁ And b ₂ Is a constant term of two minimum circumscribed tangent lines.

S33, sparse segment segmentation, which specifically comprises:

firstly judging whether the overlapped area has sparse boundary vector data, if so, constructing a minimum circumscribed rectangle according to the step S322, and obtaining a lower boundary intersection point S by intersecting the overlapped area boundary with the known sparse segment boundary, wherein for the sparse segment segmentation boundary y30, the expression is y ₃ ＝k ₃ x+b ₃ Its slope k ₃ ＝k ₁ ＝k ₂ Substituting the coordinates of the lower boundary intersection point S into the expression of y30 to calculate b ₃ Values, thereby obtaining an expression of the sparse segment segmentation boundary y 30; if the sparse boundary vector data does not exist, the initial mosaic reference line in the step S2 is directly selected without segmentation, and the rest segments except the dense segment and the sparse segment are collectively called a mixed segment.

the coordinates in different spatial reference coordinate systems are different on different reference planes, the reference planes form a part of the coordinate system, because the reference planes involve translation or rotation relative to the earth center during positioning, the conversion cannot be directly performed, a conversion parameter is needed, the parameters are also based on different models, the three-parameter method in the embodiment is that translation of X, Y and Z axes is performed between two reference planes, fig. 6 is a schematic diagram of the relationship between two reference planes between three parameters, and the three-parameter method is used for converting road vector and water vector coordinates of an input overlapping region image according to the three parameters and the reference points of translation to obtain output vector geographic coordinate data, which is calculated as follows:

first, the geographic coordinates of the upper left corner of the overlapping region image [ GEO ] _x ，GEO _y ]Known, and resolution R of image _x ，R _y It is known to calculate the geographical coordinates of arbitrary pixel locations from the geographical coordinates of the upper left corner of the image and the resolution of the image. If a geographic coordinate is given, the position of the geographic coordinate on the image can also be calculated, and fig. 7 is a schematic diagram of conversion of the geographic coordinate and the pixel coordinate.

The formula for converting the geographic coordinates into pixel coordinates is:

the pixel coordinate is converted into a geographic coordinate formula:

X＝GEO _x +col*R _x

Y＝GEO _y +row*R _y ，

wherein X and Y in the two-coordinate conversion formula represent geographic coordinates, PIPIX _x ，PIX _y Representing pixel coordinates, GEO _x 、GEO _y Representing the geographical coordinates of the upper left corner of the image, col, row representing the number of columns and rows of image pixels, R _x And R is R _y Representing the row and column resolution of the image. And matching the road vector with the A point coordinate as a starting point to the image pixel, and preparing the road vector as data generated by a subsequent mosaic line.

wherein ρ is ₀ ，ρ ₁ ，…，/>

wherein omega ₀ ，ω ₁ ，…，/>

Polynomial coefficients for a road vector curve;

D as a constraint of closest distance between the water system or road vector and the reference line _i Representing the distance between the end point of the lower boundary of the ith water system or road vector line at the dense segment and the intersection point of the lower boundary of the reference embedded line at the dense segment, N ₁ For searchingNumber of lines penetrating through the water system or road vector. The calculation method of the water system or the road path length comprises the following steps: equally dividing the water system or road curve penetrated in the step S51 into n equal parts according to the actual road vector length, and solving the total curve length of the equally divided curve by adopting a short straight line method, wherein the method comprises the following specific steps:

S516, according toS511 to S515 steps obtain dense section embedded lines as optimal embedded lines, A is obtained in the embodiment ₂ A ₅ 、A ₅ A ₃ Fig. 8 is a schematic diagram of a dense segment and sparse segment mosaic line generation process as an optimal mosaic line segment.

S52, for the sparse segment, when no secondary feature library vector exists, the original initial mosaic reference line in the step S2 is reserved as a mosaic line segment of the region, namely an intersection point A of an initial upper right corner A point of an overlapping region and a sparse segment boundary in the step S1 ₁ Formed curve AA ₁ ；

S523, obtaining AA according to the mosaic lines in the penetrating sparse segment obtained in the steps S521 to S522 as the optimal mosaic lines ₄ 、A ₄ A ₁ As an optimal mosaic line segment, as shown in fig. 8.

S6, connecting embedded wires among the sections.

The mixing section uses an initial mosaic reference line; the dense segment uses the feature vector optimal mosaic line segment A obtained in the step S51 ₂ A ₅ 、A ₅ A ₃ The method comprises the steps of carrying out a first treatment on the surface of the Sparse segment uses the optimal mosaic line AA acquired in step S52 ₄ 、A ₄ A ₁ . Finally, the segment-to-segment embedded lines are combined into complete optimal embedded line vectors in the overlapping area at the upper end points of the segment boundary lines of the mixed segment, the sparse segment and the dense segment, as shown in fig. 9, namely complete embedded line AA in the embodiment ₄ A ₁ A ₂ A ₅ A ₃ B。

S7, after the complete mosaic line is generated, storing the mosaic line into a feature library, and taking the mosaic line as the mosaic line for the next mosaic in the area under the condition of small difference of the ground feature changes; under the condition that the ground feature variation difference is large, the obtained complete mosaic line is used as a mosaic reference line, and the mosaic line optimization algorithm of the steps S1 to S6 in the embodiment is continued, so that the mosaic line generation efficiency in regional mosaic is improved.

The foregoing is merely exemplary of the present application and is not intended to limit the present application. Various modifications and changes may be made to the present application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc. which are within the spirit and principles of the present application are intended to be included within the scope of the claims of the present application.

Claims

1. A satellite image optimal mosaic line generation method based on feature library intelligent decision is characterized by comprising the following specific steps of,

s2, generating an initial mosaic reference line;

s31, selecting and grading a feature library;

s32, dense segment segmentation;

s33, sparse segment segmentation;

s4, matching road vector, water system vector data and image pixel data in the secondary feature library,

When the secondary feature library data exist, acquiring an embedded line by adopting an automatic embedded line selecting method based on a water system vector and a road vector;

wherein (1)>

wherein (1)>

Polynomial coefficients for a road vector curve;

s511, if the secondary feature library has water system vector data, integrating the water system vector curve f ₁ (x)，f ₂ (x)，f ₃ (x),...,f _N (x) The method is combined with a straight line y20 to form an equation set, and an intersection point set is solved to obtain a water system vector endpoint set in the dense segment;

s512, if the secondary feature library has no water system vector data, the road vector curve h is obtained ₁ (x),h ₂ (x),...,h _N (x) The curve set and the straight line y20 are combined to form an equation set, and a road vector endpoint set in the dense section is obtained through solving;

s513, taking the intersection point of the overlapping area reference mosaic line obtained in the step S2 and the dense section upper boundary straight line y10 as a starting point, taking a continuous water system or road vector point which penetrates through the dense section upper and lower boundaries on the dense section lower boundary straight line y20 as an end point, searching penetrating water systems or roads according to constraint conditions, searching M water systems or roads, wherein N water systems or roads (N is less than or equal to M) are water systems or road vectors containing the end points until the continuous water systems or road vector lines which penetrate through the dense section upper and lower boundaries are searched at the end point on the dense section lower boundary straight line y20, selecting the searched continuous water systems or road vector lines which penetrate through the dense section upper and lower boundaries as optimal mosaic lines, if the penetrating water systems or roads are not searched all the time, cutting off the road or water system vector data, and selecting the initial mosaic reference line in the step S2 as the optimal mosaic lines;

the process of judging that the water system or road vector characteristic data penetrate is that the discrete point coordinates of each water system or road are fitted into a water system or road vector curve equation, the starting point and the ending point coordinates of the boundary points of the water system or road are brought into y10 and y20 equations of boundary straight lines, when the end point coordinates of the water system or road vector are on the boundary straight lines, the water system or road vector penetrates through the upper and lower boundaries of dense or sparse sections, through-road or water system vector data are selected, if the end point coordinates of the water system or road vector are not on the boundary straight lines, the road or water system vector data are cut off, and the initial embedded line in the step S2 is selected as the embedded line of the sections;

D as a constraint of closest distance between the water system or road vector and the reference line _i Representing the distance between the end point of the lower boundary of the ith water system or road vector line at the dense segment and the intersection point of the lower boundary of the reference embedded line at the dense segment, N ₁ The number of the searched through water system or road vector; the calculation method of the water system or the road path length comprises the following steps: equally dividing the water system or road curve penetrated in the step S51 into n equal parts according to the actual road vector length, and solving the total curve length of the equally divided curve by adopting a short straight line method, wherein the method comprises the following specific steps:

wherein L represents the total curve length, deltax _i For the arc length in the ith subinterval to correspond to the transverse axis length, deltay _i For the corresponding longitudinal axis length of the arc length in the ith subinterval, x _i Is the abscissa of any point in the ith subinterval, i is the number of curve divisions, where i=1, 2, …, n, f' ² (x _i ) Is the water system or the road curve equation f (x) at x _i Derivative level of pointsAnd finally, accumulating n equal curve lengths to obtain a total curve length L, namely the corresponding water system or road path length penetrating through the upper boundary and the lower boundary of the dense section, and calculating the curve lengths of all water systems or road vectors in the dense section to ensure that L _min ＝min{L ₁ ,L ₂ ,...,L _N Selecting L _min As a water system or road vector shortest path constraint; obtaining an optimal mosaic line segment A according to the method in the step S514 ₂ A ₅ ；

S515, the shortest water system or road vector line end point coincides with the reference mosaic line lower boundary end point coordinates at the dense segment or d _min Under the condition that the distance between the shortest water system or road vector line end point and the embedded reference line end point in the step S2 is directly connected with the end point of the embedded reference line in the dense section lower boundary y20 at the distance of less than or equal to 10m, a buffer circle is established by taking the distance between two intersection points of the shortest water system or road vector line and the embedded reference line at the dense section lower boundary as the diameter of a circle, the midpoint of the two points is the center of the circle, and the line segment A is optimally embedded in the step S514 in the circle ₂ A ₅ Intersection point A with dense segment lower boundary y20 ₅ (xx 5, yy 5) as a starting point, searching a water system or a road vector, and selecting an optimal mosaic line segment A in the segment according to the steps S511 to S514 ₅ A ₃ ；

S516, obtaining the dense-segment embedded line as an optimal embedded line according to the steps S511 to S515;

s52, for the sparse segment, when no secondary feature library vector exists, reserving the original initial mosaic reference line in the step S2 as an initial mosaic line of the sparse segment;

s521, obtaining an optimal mosaic line segment AA with the priority of the road according to the through road selection mode of the road and the water system in the sparse ground object section and taking the point A as the starting point and the steps S511 to S515 ₄ If the road and the water system meeting the conditions in the step S514 are not available, selecting an initial mosaic reference line in the step S2;

s522, the shortest road or water system vector line end point coincides with the reference embedded line at the sparse segment boundary coordinates or d _min Under the condition of less than or equal to 10m, directly connecting the shortest road or water system vector line end point in the lower boundary y30 of the sparse segment with the embedded reference line end point in the step S2; distance d between end point and reference line _min When the distance between the shortest water system or road vector line and the reference embedded line is more than 10m, the distance between two intersection points of the lower boundary of the sparse segment is taken as the diameter of a circle, the midpoint of the two intersection points is taken as the center of the circle, a buffer circle is established, and the optimal embedded line segment AA in the sparse segment is taken in the circle ₄ Intersection point A with sparse segment boundary y30 ₄ As a starting point, searching water system or road vector, selecting the optimal mosaic line segment A according to the steps S511 to S514 ₄ A ₁ ；

S523, obtaining AA according to the mosaic lines in the penetrating sparse segment obtained in the steps S521 to S522 as the optimal mosaic lines ₄ 、A ₄ A ₁ As an optimal mosaic line segment;

s53, aiming at the generation of embedded lines in the mixed section except the dense section and the sparse section, respectively selecting intersection points of the initial embedded reference line and the sparse section in the step S2, and taking the initial embedded reference line or the intersection point of the boundary between the initial embedded reference line and the dense section as a starting point and an end point of an embedded line section, and taking the initial embedded reference line in the step S2 as the embedded line;

S6, connecting embedded lines among the sections;

the mixing section uses an initial mosaic reference line; the dense segment uses the optimal mosaic line segment A obtained in the step S51 ₂ A ₅ 、A ₅ A ₃ The method comprises the steps of carrying out a first treatment on the surface of the The sparse segment uses the optimal mosaic line segment AA acquired in the step S52 ₄ 、A ₄ A ₁ The method comprises the steps of carrying out a first treatment on the surface of the Finally, merging the end points of the embedded lines between the sections on the sectional boundary lines of the mixed section, the sparse section and the dense section into an integral optimal embedded line vector in the overlapping area;

s7, after the complete mosaic line is generated, storing the mosaic line into a feature library, and taking the mosaic line as the mosaic line for the next mosaic in the area under the condition of small difference of the ground feature changes; under the condition that the ground feature variation difference is large, the obtained complete mosaic line is used as a mosaic reference line, and the mosaic line optimization method of the steps S1 to S6 is continued, so that the mosaic line generation efficiency in regional mosaic is improved.

2. The method for generating a satellite image optimal mosaic line based on intelligent decision of feature library as set forth in claim 1, wherein in step S1, for obtaining an image effective area of a satellite image, the method comprises the steps of:

If the obtained satellite image is a single-band image, sequentially circulating from the first row to the subsequent row according to the number of rows and columns of the image, judging each pixel value from left to right in each row, continuously judging the next pixel value if the value is 0, recording the corresponding row number and column number of the first non-0 pixel value, setting the point as an upper left corner C point, acquiring the upper left corner, sequentially circulating from the last row to the last row, judging whether each pixel value is 0 from right to left in each row, and setting the point as a lower right corner D point when the first non-0 pixel value is encountered; continuously circulating each column of the image from left to right, judging each pixel value from bottom to top in each column, continuously judging the next pixel value if the value is 0, recording the corresponding row number column number when the first non-0 pixel value is judged, setting the pixel as a lower left corner B point, circularly reading the image column from right to left from the rightmost column after acquiring the lower left corner point, judging each pixel value from top to bottom in each column, and setting the pixel as an upper right corner A point when the first non-0 pixel value is encountered;

creating a vector file according to the obtained four corner row and column numbers, and completing the acquisition of an effective area of a satellite image;

3. A method for generating a satellite image optimal mosaic line based on intelligent decision of feature library as set forth in claim 1 or 2, wherein the steps ofThe automatic mosaic line extraction method based on pixel difference eclosion described in the step S2 is implemented by adopting a Dijkstra method, wherein the Dijkstra method is used for solving the shortest path of the minimum cost (cost) between two points, the cost is replaced by a pixel path with weight, the brightness difference delta I (I, j) of an overlapping area and the pixel value S (I, j) after the normalization of gradient distinction are used, the gray gradient difference delta g is the absolute value of the difference of the gradient value, the gradient value is the change value g of the gray scale of an image in the horizontal and vertical directions, and the gradient direction is

The specific calculation process of the brightness difference Δi (I, j) is as follows: />

ΔI(i,j)＝|img1(i,j)-img2(i,j)|，

The calculation formula of (2) is as follows:

g(i,j)＝[(img(i,j)-img(i+1,j) ² +(img(i,j)-img(i,j+1)) ² )] ^1/2

respectively representing the gradient directions corresponding to the two pixel points; the maximum value of the normalized brightness difference or gray gradient difference represents the maximum difference value of the pixels in the overlapping region, the value is max (delta I (I, j), delta g (I, j)), the corresponding pixel points are called pixel difference points, the set of the pixel difference points is CC, the upper right corner of the boundary of the image in the overlapping region is taken as the starting point A of the overlapping region, the coordinates of the starting point A are (x 0, y 0), and the starting point A is from the starting point A to all the pixel difference points a in the overlapping region _i E CC, the shortest path in the overlap region is denoted Dist [ A, a ] _i ]And i is a subscript, and then:

Dist[A,a _i ]≥short[A,a _i ]，

wherein short [ A, a ] _i ]Representing the difference point a from the start point A of the overlapping region to all pixels _i E shortest path of CC; by continuously calculating Dist [ A, a ] _i ]Up to a _i To the point B at the left lower corner of the overlap region in step S1, at which time Dist [ A, a ] _i ]＝short[A,a _i ]And obtaining an initial mosaic reference line according to the mosaic line generation method based on the pixel difference eclosion.

4. The method for generating the optimal mosaic line of the satellite image based on intelligent decision of the feature library as claimed in claim 1, wherein the step S31 is characterized by comprising the following steps:

classifying ground feature characteristics in an overlapping area in the acquired satellite image, and establishing a primary characteristic library and a secondary characteristic library, wherein the primary characteristic library is used for determining segmentation limit, and the secondary characteristic library is used for determining the selection of an embedded line; selecting a primary feature library and a secondary feature library from the feature library data, wherein:

the secondary feature library comprises road vector data, water system vector data and abrupt bank data;

dividing the overlapping region into a dense segment, a sparse segment and a mixed segment according to dense boundary vector data and sparse boundary vector data in a primary feature library according to feature types and feature library data established by various remote sensing geographic information resources, obtaining the dense segment according to the dense boundary vector data in the overlapping region according to the step S32, obtaining the sparse segment according to the sparse boundary vector data in the overlapping region according to the step S33, wherein the regions except the dense segment and the sparse segment are collectively called the mixed segment, and a specific segmentation method is shown in the steps S32 and S33; the dense boundary vector data establishes a feature library according to a building, high-grade road data, noctilucent remote sensing data, water system vectors and the like, and the sparse boundary vector data establishes a feature library according to a desert vector, DEM data, expressways, low-grade road data, water system vectors and the like.

5. The method for generating the optimal mosaic line of the satellite image based on intelligent decision of the feature library as claimed in claim 1, wherein the step S32 of dense segment segmentation specifically comprises:

drawing a small circumscribed rectangle by using a minAreRect () function meter in opencv, wherein the small circumscribed rectangle comprises four endpoints and four edges of the minimum circumscribed rectangle; firstly, defining an circumscribed rectangle set as a circumscribed rectangle set formed by the contours of dense boundary vectors and sparse boundary vectors in an overlapping area ABCD in the step S1, and defining a minimum circumscribed rectangle set as a set formed by the minimum circumscribed rectangles of the contours of each dense boundary vector and sparse boundary vector; the outline here refers to the set of points within the overlap region ABCD obtained in step S1 corresponding to the boundary vector; drawing a center point of the minimum circumscribed rectangle to obtain four end points of the minimum circumscribed rectangle; finally, drawing four sides of the minimum circumscribed rectangle to finish drawing the minimum circumscribed rectangle;

Rotating the polygon formed by the endpoints of the original dense boundary vector data, obtaining a simple circumscribed rectangle of the polygon after rotating each degree from-90 degrees to 90 degrees, recording the area, vertex coordinates and rotation degree of each simple circumscribed rectangle, obtaining the circumscribed rectangle with the smallest area, and finally rotating the circumscribed rectangle by a rotation angle with the same size in the opposite direction to the recorded rotation degree, thus obtaining the smallest circumscribed rectangle;

s323, segmenting the overlapped area in the step S1 according to the dense boundary vector under the condition that the minimum circumscribed rectangle obtained in the step S322 is intersected with or included in the circumscribed rectangle of the rest boundary vector;

y ₁ ＝k ₁ x+b ₁

y ₂ ＝k ₂ x+b ₂ ，

6. The method for generating the optimal mosaic line of the satellite image based on intelligent decision of the feature library as claimed in claim 1, wherein the step S33 of sparse segment segmentation specifically comprises:

Firstly judging whether the overlapped area has sparse boundary vector data, if so, constructing a minimum circumscribed rectangle according to the step S322, and obtaining a lower boundary intersection point S by intersecting the overlapped area boundary with the known sparse segment boundary, wherein for the sparse segment segmentation boundary y30, the expression is y ₃ ＝k ₃ x+b ₃ Its slope k ₃ ＝k ₁ ＝k ₂ Substituting the coordinates of the lower boundary intersection point S into the expression of y30 to calculate b ₃ Values, thereby obtaining an expression of the sparse segment segmentation boundary y 30; if the sparse boundary vector data does not exist, the initial in the S2 step is directly selected without segmentationThe mosaic reference lines refer to the remaining segments except for the dense segment and the sparse segment collectively as a hybrid segment.

7. The method for generating the optimal mosaic line of the satellite image based on intelligent decision of the feature library as claimed in claim 1, wherein the step S4 specifically comprises:

wherein [ X, Y, Z] _new For the vector data obtained, [ X, Y, Z ] ] _orgin Is the original vector coordinate, [ dX, dY, dZ ]]Three parameters of translation;

matching road vector data and water system vector data with geographic entity coordinate points to the image of the overlapping region according to prior feature data of the secondary feature library, and matching homonymous entities; the positions of the pixel points in the image are represented by pixel coordinates, when matching the road with the water system vector data and the image pixels in the overlapping area, the corresponding geographic coordinates are required to be converted into the pixel coordinates, and the conversion method comprises the following steps:

first, the geographic coordinates of the upper left corner of the overlapping region image [ GEO ] _x ,GEO _y ]Known, and resolution R of image _x ,R _y The geographical coordinates of any pixel position are calculated through the geographical coordinates of the upper left corner of the image and the resolution of the image; the formula for converting the geographic coordinates into pixel coordinates is:

the pixel coordinate is converted into a geographic coordinate formula:

X＝GEO _x +col*R _x

Y＝GEO _y +row*R _y ，

wherein X and Y in the two-coordinate conversion formula represent geographic coordinates, PIX _x ，PIX _y Representing pixel coordinates, GEO _x 、GEO _y Representing the geographical coordinates of the upper left corner of the image, col, row representing the number of columns and rows of image pixels, R _x And R is R _y Representing the row and column resolution of the image; and matching the road vector with the A point coordinate as a starting point to the image pixel, and preparing the road vector as data generated by a subsequent mosaic line.