WO2022111609A1 - 一种网格编码方法及计算机系统 - Google Patents

一种网格编码方法及计算机系统 Download PDF

Info

Publication number
WO2022111609A1
WO2022111609A1 PCT/CN2021/133339 CN2021133339W WO2022111609A1 WO 2022111609 A1 WO2022111609 A1 WO 2022111609A1 CN 2021133339 W CN2021133339 W CN 2021133339W WO 2022111609 A1 WO2022111609 A1 WO 2022111609A1
Authority
WO
WIPO (PCT)
Prior art keywords
square
target
projection
plane
grids
Prior art date
Application number
PCT/CN2021/133339
Other languages
English (en)
French (fr)
Inventor
汪超亮
Original Assignee
华为技术有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 华为技术有限公司 filed Critical 华为技术有限公司
Publication of WO2022111609A1 publication Critical patent/WO2022111609A1/zh

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T9/00—Image coding
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T17/00—Three dimensional [3D] modelling, e.g. data description of 3D objects
    • G06T17/20—Finite element generation, e.g. wire-frame surface description, tesselation

Definitions

  • the present application relates to the field of computers, and in particular, to a trellis coding method and a computer system.
  • the discrete global grid is a spherical (or ellipsoid)-based earth fitting grid that can be subdivided infinitely without changing its shape.
  • DGG has the characteristics of hierarchy and global continuity, and overcomes many constraints and uncertainties that limit the application of geographic information systems, so that spatial data of any resolution (different precision) acquired at any location on the earth can be expressed and expressed in a standardized manner. analysis, and can perform multi-scale operations with deterministic precision.
  • DGG includes GeoSOT grid.
  • GeoSOT grid belongs to quadtree grid system with equal latitude and longitude. First, the space on the earth's surface is expanded into a square plane space. For the plane space grid with the intersection as the center, GeoSOT divides the 1-level grid into four equal parts on the basis of 0-level, and the subsequent division levels follow the principle of the above quadtree and so on.
  • GeoSOT belongs to a quadtree grid system with equal latitude and longitude
  • the Euclidean distance of the grid is different in each level, and the size of the grid gradually decreases from the equator to the poles, which is a generalization of the calculation that is sensitive to the grid size. more difficult.
  • the present application provides a trellis coding method, the method comprising:
  • the target projection is the projection of the earth's surface on a plane, and the area difference between any two areas of the same area on the earth's surface corresponding to the target projection is less than a preset value; you can use the map projection method to convert the earth's surface to a plane to obtain the target projection, where the map projection method can be a global projection or a zonal projection.
  • the earth's surface plane of equal latitude and longitude is obtained by the reduction and deformation of the earth's surface by a certain proportion.
  • the reduction ratio of the area far from the equator is much smaller than that of the area close to the equator, that is to say, The area of the Earth's surface that is far from the equator and the area that is close to the equator is very different between the two areas on the Earth's surface plane at equal latitude and longitude.
  • the area difference between any two regions of the same area on the surface of the earth corresponding to the two projection regions on the target projection is smaller than the preset value, that is, any two regions on the surface of the earth are smaller than the preset value.
  • the target projection is obtained by reducing an area of the same area by a certain ratio, and the difference in the reduction ratio of any two areas of the same area on the surface of the earth is very small.
  • the present application does not limit that the areas between any two regions of the same area on the surface of the earth corresponding to the two projection regions on the target projection are completely consistent, but the difference is smaller than a preset value.
  • a square plane is determined, and the area where the target projection is located is covered by the square plane; after the target projection is obtained, a square plane can be determined according to the target projection, and the area where the target projection is located is covered by the Square flat overlay.
  • the area of the square plane should be larger than the target projection, and the sides of the square plane should be outside the area of the target projection.
  • the target projection is non-square, so the square plane includes the area that overlaps with the target projection, and also includes the area that does not overlap with the target projection.
  • the plurality of square grids are encoded to obtain a plurality of codes, each code corresponds to a square grid, and each code is used to indicate an area covered by the corresponding square grid.
  • a group of square grids can be obtained, and then the group of square grids can be coded to obtain the After encoding, the next bisection can be performed, that is, each square grid in a group of square grids obtained by the previous division can be divided into two equal parts in the horizontal and vertical directions. Divide to get a new set of square grids, which can then be encoded to get the encoding of the set of square grids, and so on.
  • each code obtained after the square grid is coded is used to indicate the area covered by the corresponding square grid, and the present application does not limit a specific coding manner.
  • grid division is performed on the square plane constructed on the target projection, so that the division size of the earth surface corresponding to the square grid obtained by the same level of division is basically the same, which is conducive to the calculation sensitive to the size of grid data.
  • the first target point and the second target point are any two points on the target projection
  • the first target point is the projection of the first position point on the surface of the earth
  • the second target point is the projection of the second position point on the surface of the earth
  • the distance between the first target point and the second target point is the first distance
  • the distance between the second position points is the second distance
  • the ratio between the first distance and the second distance is within a preset range.
  • the Euclidean distance of each point on the earth's surface can be more accurately represented by the target projection.
  • the distance between any two points on the target projection can represent the Euclidean distance between the corresponding two physical location points on the earth's surface.
  • the target projection is an approximate undistorted projection of the earth's surface.
  • the distance between each point on the target projection is obtained by reducing the earth's surface by a certain percentage, and the reduction ratio of the distance between each point on the earth's surface is basically the same. Therefore, through the target Projection can more accurately represent the Euclidean distance of each point on the earth's surface.
  • the earth surface includes an equatorial latitude and a target longitude perpendicular to the equatorial latitude
  • the target projection includes the projected equatorial latitude and the projected target longitude
  • the horizontal axis of the square plane The direction is consistent with the direction in which the projected equatorial latitude is located, and the direction of the longitudinal axis of the square plane is consistent with the direction in which the projected target meridian is located.
  • the direction of the square plane needs to be determined.
  • the so-called direction refers to the directions of the horizontal axis and the vertical axis included in the square plane.
  • the square plane can include a horizontal axis and a vertical axis, wherein the horizontal axis can be parallel to one side of the square plane, and the vertical axis is perpendicular to the horizontal axis.
  • the horizontal axis can be parallel to one side of the square plane and pass through the square plane.
  • the axis of the center, the vertical axis is the axis perpendicular to the horizontal axis and passing through the center of the square plane.
  • the determining a square plane according to the target projection includes:
  • the side length of the square plane is a preset multiple of an integer power of 2; the reason why the side length of the square plane is a preset multiple of an integer power of 2 is because the horizontal axis of the square plane needs to be performed during subsequent encoding.
  • the side length of the square plane can be Wherein, L is the length in the equatorial latitude direction in the target projection, and S is the scaling factor, that is, the above-mentioned preset multiple. According to the side length of the square plane, the square plane is determined so that the area where the target projection is located is covered by the square plane. In this embodiment, the size of the square grid obtained after division can be controlled by controlling the size of the preset multiple.
  • determining the side length of the square plane according to the target length projected by the target in the direction of the projected equatorial latitude includes:
  • the preset multiple is acquired, and the side length of the square plane is determined according to the target length projected by the target in the direction of the projected equatorial latitude and the preset multiple.
  • the central transverse axis of the square plane overlaps the projected equatorial latitude.
  • the square plane is divided into two equal parts in the horizontal axis direction and the vertical axis direction multiple times to obtain a plurality of square grids, including:
  • the square plane is divided into two equal parts in the horizontal axis direction and the vertical axis direction for many times to obtain multiple groups of square grids, each group of square grids includes a plurality of square grids, wherein the multiple times of bisection
  • the subdivision includes the Nth bisection division and the N+1th bisection division, and after the Nth bisection division is performed, M square grids are obtained, and the N+1th bisection is performed. After bisection, M*4 square grids are obtained, and the M*4 square grids are the horizontal axis direction and the vertical axis direction of each square grid in the M square grids obtained by bisection division;
  • the encoding the multiple square grids to obtain multiple codes includes: encoding the multiple square grids included in each group of square grids in the multiple groups of square grids to obtain multiple sets of codes , and each set of codes includes multiple codes.
  • the target projection is obtained by performing a global projection or a zonal projection on the surface of the earth.
  • the present application provides a trellis encoding device, the device comprising:
  • an acquisition module for acquiring a target projection, the target projection being the projection of the earth's surface on a plane;
  • a determining module configured to determine a square plane according to the target projection, and the area where the target projection is located is covered by the square plane;
  • a meshing module configured to perform multiple bisection divisions on the square plane in the horizontal and vertical directions to obtain a plurality of square meshes
  • an encoding module configured to encode the plurality of square grids to obtain a plurality of codes, each code corresponds to a square grid, and each code is used to indicate an area covered by the corresponding square grid.
  • the first target point and the second target point are any two points on the target projection
  • the first target point is the projection of the first position point on the surface of the earth
  • the second target point is the projection of the second position point on the surface of the earth
  • the distance between the first target point and the second target point is the first distance
  • the sum of the first position point and the obtained The distance between the second position points is the second distance
  • the ratio between the first distance and the second distance is within a preset range.
  • the earth surface includes an equatorial latitude and a target longitude perpendicular to the equatorial latitude
  • the target projection includes the projected equatorial latitude and the projected target longitude
  • the horizontal axis of the square plane The direction is consistent with the direction in which the projected equatorial latitude is located, and the direction of the longitudinal axis of the square plane is consistent with the direction in which the projected target meridian is located.
  • the determining module is configured to determine the side length of the square plane according to the target length projected by the target in the direction of the projected equatorial latitude, wherein the square The side length of the plane is greater than or equal to the target length, and the side length of the square plane is a preset multiple of an integer power of 2;
  • the square plane is determined so that the area where the target projection is located is covered by the square plane.
  • the determining module is configured to obtain the preset multiple, and determine the target length according to the projected target length in the direction of the projected equatorial latitude and the preset multiple. The side length of the square plane.
  • the central transverse axis of the square plane overlaps the projected equatorial latitude.
  • the meshing module is configured to perform multiple bisection divisions on the square plane in the horizontal axis direction and the vertical axis direction to obtain multiple sets of square meshes, each set of The square grid includes a plurality of square grids, wherein the multiple bisections include the Nth bisection and the N+1th bisection, and the Nth bisection is performed After the division, M square grids are obtained, and after the N+1th bisection division is performed, M*4 square grids are obtained, and the M*4 square grids are for the M Each square grid in the square grid is obtained by bisection in the horizontal and vertical directions;
  • the encoding module is configured to encode multiple square grids included in each group of square grids in the multiple groups of square grids, so as to obtain multiple sets of codes, and each group of codes includes multiple codes.
  • the target projection is obtained by performing a global projection or a zonal projection on the surface of the earth.
  • the present application provides a computer system, the computer system includes a memory and a processor, the memory is used for storing computer-readable instructions (or referred to as a computer program), and the processor is used for reading the Computer readable instructions to implement the methods provided by any of the foregoing implementations.
  • the present application provides a computer storage medium, which may be non-volatile.
  • Computer-readable instructions are stored in the computer storage medium, and when the computer-readable instructions are executed by the processor, implement the method provided by any of the foregoing implementations.
  • the present application provides a computer program product, the computer program product includes computer-readable instructions, and when the computer-readable instructions are executed by a processor, implements the method provided by any of the foregoing implementations.
  • An embodiment of the present application provides a grid coding method, the method includes: acquiring a target projection, where the target projection is a projection of the earth surface on a plane, and any two regions of the same area on the earth surface are The area difference between the two corresponding projection areas on the target projection is less than a preset value; according to the target projection, a square plane is determined, and the area where the target projection is located is covered by the square plane; Perform multiple bisections along the horizontal axis and the vertical axis to obtain a plurality of square grids; encode the plurality of square grids to obtain a plurality of codes, each code corresponding to a square grid , and each code is used to indicate the area covered by the corresponding square grid.
  • Fig. 1 is a schematic diagram of a GeoSOT grid
  • Fig. 2 is a schematic diagram of a GeoSOT grid
  • FIG. 3 is a schematic flowchart of a trellis coding method provided by an embodiment of the present application.
  • FIG. 4 is a schematic diagram of a target projection provided by an embodiment of the present application.
  • FIG. 5 is a schematic diagram of a square plane provided by an embodiment of the present application.
  • FIG. 6 is a schematic diagram of the division of a square grid provided by an embodiment of the present application.
  • FIG. 7 is a schematic diagram of the division of a square grid provided by an embodiment of the present application.
  • FIG. 8 is a schematic structural diagram of a trellis encoding apparatus provided by an embodiment of the present application.
  • FIG. 9 is a schematic structural diagram of a computer system provided in this embodiment.
  • FIG. 10 is a schematic structural diagram of an NPU according to this embodiment.
  • plural means two or more.
  • the term “and/or” or the character “/” in this application is only an association relationship to describe associated objects, indicating that there can be three relationships, for example, A and/or B, or A/B, which can indicate: There are three cases where A exists alone, A and B exist at the same time, and B exists alone.
  • Discrete Global Grid is a spherical (or ellipsoid)-based earth fitting grid that can be subdivided infinitely without changing its shape. The purpose of simulating the surface of the earth. DGG has the characteristics of hierarchy and global continuity, and overcomes many constraints and uncertainties that limit the application of geographic information systems, so that spatial data of any resolution (different precision) acquired at any location on the earth can be expressed and expressed in a standardized manner. analysis, and can perform multi-scale operations with deterministic precision.
  • the global discrete grid system can be roughly divided into four categories: the equal latitude and longitude global grid, the variable longitude and latitude global grid, the adaptive global grid and the regular polyhedron global grid system.
  • Global grid, degenerate quadtree grid is a variable latitude and longitude global grid
  • digital elevation model (DEM) data level of detail (levels of detail, LOD) model is adaptive global grid
  • triangular grid rhombic mesh
  • hexagonal mesh etc. are regular polyhedral global meshes.
  • GeoSOT grid is a quadtree grid system with equal latitude and longitude.
  • the space of 360°*180° on the earth's surface is expanded to a plane space of 512°*512°.
  • the 512°*512° plane space grid with the intersection of the prime meridian as the center GeoSOT divides the level 1 grid into four equal parts on the basis of level 0, and the size of each level 1 grid is 256°*256°, Specifically, it can be shown in Figure 1.
  • GeoSOT divides the level 2 grid into four equal parts on the basis of level 1, and the size of each level 1 grid is 128°*128°. The following division levels follow the principle of the quadtree above and so on.
  • GeoSOT divides 9-level grids with a size of 1°*1°. Grids above 9-level are GeoSOT “degree”-level grids, and the 10th to 15th-level grids are “sub-level” grids. The starting point of the patch is the 1° patch of the 9-level grid. The initial numerical space of the grid is extended from 60' to 64'.
  • the GeoSOT 10-level mesh is divided into four equal parts with a size of 64'*64'. The size of each 10-level grid is 32'*32'.
  • the 10 to 15-level grids are "divided" grids, and the division method is recursive according to the above rules, as shown in Figure 2.
  • Levels 16 to 21 are “second” level grids, and the “second” level grid division method refers to the hierarchical grid, that is, the value range of the 15-level 1'*1' patch is extended to 64"*64".
  • the grades are divided according to the quartering method.
  • the 22-32 grids are divided in strict accordance with the quadruple method at the second level, and the size of the 32nd grid is 1/2048”*1/2048”.
  • FIG. 3 is a schematic flowchart of a trellis coding method provided by an embodiment of the present application. As shown in FIG. 3 , the present application implements Examples of trellis encoding methods provided include:
  • the target projection is a projection of the earth surface on a plane, and the area difference between any two regions of the same area on the earth surface corresponding to the target projection is less than default value.
  • a map projection method can be used to convert the earth's surface to a plane to obtain a target projection, wherein the map projection method can be an overall projection or a zonal projection, that is to say, the target projection can be It is obtained by performing an integral projection or a zonal projection on the surface of the earth.
  • the GeoSOT grid is obtained by directly performing grid division on a square surface of equal latitude and longitude obtained by deforming the earth surface, and the Euclidean distance in different grids divided by latitude and longitude is different.
  • a square plane that can cover the target projection is re-determined outside the target projection.
  • the earth's surface plane with equal latitude and longitude is obtained by reducing and deforming the earth's surface by a certain proportion.
  • the reduction ratio of the area far from the equator is much smaller than that of the area close to the equator, that is to say,
  • the area of the Earth's surface that is far from the equator and the area that is close to the equator is very different between the two areas on the Earth's surface plane at equal latitude and longitude.
  • the area difference between any two regions of the same area on the surface of the earth corresponding to the two projection regions on the target projection is smaller than the preset value, that is, any two regions on the surface of the earth are smaller than the preset value.
  • the target projection is obtained by reducing an area of the same area by a certain ratio, and the difference in the reduction ratio of any two areas of the same area on the surface of the earth is very small.
  • the Euclidean distance of each point on the earth's surface can be more accurately represented by the target projection.
  • the distance between any two points on the target projection can represent the Euclidean distance between the corresponding two physical location points on the earth's surface.
  • Distance the so-called Euclidean distance, refers to the Euclidean metric, and in mathematics, the Euclidean distance or Euclidean metric is the "ordinary" (ie, straight-line) distance between two points in Euclidean space.
  • the distance between two points on the plane of equal longitude and latitude on the earth surface in Fig. 1 cannot accurately represent the Euclidean distance of each point on the earth surface.
  • the equal latitude and longitude plane of the earth's surface is obtained by reducing and deforming the earth's surface by a certain percentage.
  • the reduction ratio of the distance between two points far from the equator is much smaller than that of the one close to the equator.
  • the reduction ratio of the distance between two points such as the distance between two points close to the equator on the equal latitude and longitude plane, the distance between the corresponding actual physical location points on the earth's surface, compared to the distance between the two points far from the equator on the equal latitude and longitude plane
  • the distance between the corresponding actual physical location points on the earth surface represented by two points with the same distance is longer.
  • the target projection is an approximate undistorted projection of the earth's surface
  • the distance between each point on the target projection is obtained by reducing the earth's surface by a certain percentage
  • the reduction ratio of the distance between each point on the earth's surface It is basically the same, so the Euclidean distance of each point on the earth's surface can be more accurately represented by the target projection.
  • the first target point and the second target point are any two points on the target projection
  • the first target point is the projection of the first position point on the surface of the earth
  • the second target point is the projection of the second position point on the surface of the earth
  • the distance between the first target point and the second target point is the first distance
  • the first position point and the second position point are obtained
  • the distance between them is a second distance
  • the ratio between the first distance and the second distance is within a preset range.
  • the preset range can be selected according to the actual situation.
  • the preset range can represent the reduction ratio of the earth's surface during plane projection. Since the reduction ratio of the distance between any two points on the earth's surface is within the preset range, then The Euclidean distance of each point on the earth's surface can be more accurately represented by the target projection.
  • the present application does not limit the areas between any two regions of the same area on the surface of the earth corresponding to the two projection regions on the target projection are completely consistent, but the difference is smaller than a preset value.
  • the target projection as an example obtained by performing a universal transverse mercator (UTM) projection on the surface of the earth on a plane.
  • UDM universal transverse mercator
  • the UTM projection is a horizontal-axis equiangular secant elliptical cylinder projection.
  • the elliptical cylinder cuts the earth at two contour circles at 80° south latitude and 84° north latitude, and the two secant meridians are not deformed after projection.
  • the length ratio on the central meridian is 0.9996.
  • the projection of the central meridian of the projection is the ordinate axis, and the straight line after the projection of the equatorial latitude is the horizontal axis.
  • the globe is divided into 60 projection zones, each with a longitude difference of 6°.
  • the first belt is from 180°W to 174°W, numbered continuously eastward.
  • T tan 2 B
  • C e′ 2 cos 2 B
  • A (LL 0 )cosB
  • a is the semi-major axis of the earth ellipsoid
  • b is the semi-minor axis of the earth ellipsoid
  • e is the first eccentricity
  • e' is the second eccentricity
  • L 0 is the central longitude of the zoning.
  • FIG. 4 is a schematic diagram of a target projection provided by an embodiment of the application.
  • the target projection shown in FIG. 4 is obtained by performing UTM projection on the earth surface, wherein the central meridian of the target projection is projected as an ordinate axis, the straight line projected from the equatorial latitude is the abscissa axis.
  • the UTM projection divides the globe into 60 projection bands, each spanning 6° of longitude.
  • a square plane may be determined according to the target projection, and the area where the target projection is located is covered by the square plane.
  • the area of the square plane should be larger than the target projection, and the sides of the square plane should be outside the area of the target projection.
  • the target projection is non-square, so the square plane includes the area that overlaps with the target projection, and also includes the area that does not overlap with the target projection.
  • the so-called direction refers to the directions of the horizontal axis and the vertical axis included in the square plane.
  • the square plane can include a horizontal axis and a vertical axis, wherein the horizontal axis can be parallel to one side of the square plane, and the vertical axis is perpendicular to the horizontal axis.
  • the horizontal axis can be parallel to one side of the square plane and pass through the square plane.
  • the axis of the center, the vertical axis is the axis perpendicular to the horizontal axis and passing through the center of the square plane.
  • the earth surface includes an equatorial line of latitude and a target meridian perpendicular to the equatorial line of latitude
  • the target projection includes the projected equatorial line of latitude and the projected target longitude
  • the horizontal axis direction of the square plane is the same as the The direction of the projected equatorial latitude is the same
  • the direction of the longitudinal axis of the square plane is the same as the direction of the projected target meridian.
  • FIG. 5 is a schematic diagram of a square plane provided by an embodiment of the present application. In FIG.
  • the horizontal axis of the square plane may be the axis x(m), the vertical axis may be y(m), and the square
  • the direction of the horizontal axis of the plane is consistent with the direction of the projected equatorial latitude, and the direction of the vertical axis of the square plane is consistent with the direction of the projected target longitude.
  • the side length of the square plane can be determined. Specifically, in one implementation, the side length of the square plane can be determined according to the target length projected by the target in the direction of the projected equatorial latitude. Wherein, the side length of the square plane is greater than or equal to the target length, and the side length of the square plane is a preset multiple of an integer power of 2, and the square plane is determined according to the side length of the square plane, so that the area where the target projection is located is covered by the square plane. In one implementation, the central transverse axis of the square plane overlaps the projected equatorial latitude.
  • the side length of the square plane is a preset multiple of the integer power of 2
  • the square plane needs to be divided into two equal parts in the horizontal and vertical directions. After that, square grids of an integer power of 2 can be obtained, and it is necessary to ensure that the side length of the square plane is a preset multiple of the integer power of 2.
  • the side length of the square plane can be Wherein, L is the length in the equatorial latitude direction in the target projection, and S is the scaling factor, that is, the preset multiple in the above-mentioned embodiment.
  • the preset multiple may be obtained, and the side length of the square plane may be determined according to the target length projected by the target in the direction of the projected equatorial latitude and the preset multiple,
  • the size of the square grid obtained after division can be controlled by controlling the size of the preset multiple.
  • the straight line after the projection of the equatorial latitude in the target projection can be used as the horizontal axis, and the projected meridian perpendicular to it is used as the vertical axis to establish a plane rectangular coordinate system, and the target projection can be divided into four parts, which are the northeast projection respectively. Facets, Southeast Projected Facets, Northwest Projected Facets, and Southwest Projected Facets.
  • the target projection as an example obtained by performing UTM projection on the surface of the earth on a plane.
  • the straight line projected by the equatorial latitude as the horizontal axis that is, coincident with the horizontal axis coordinates of the UTM projection, and the coordinate origin as the prime meridian projection as the vertical axis
  • a two-dimensional plane rectangular coordinate system is established, and the earth projection plane is divided into four parts, They are the northeast projection facet, the southeast projection facet, the northwest projection facet and the southwest projection facet, as shown in Figure 5.
  • Let the scaling factor s 1, Extend the northeast projection facet, southeast projection facet, northwest projection facet and southwest projection facet to a square grid of W*W size.
  • the square plane may be divided into two equal parts in the horizontal axis direction and the vertical axis direction multiple times to obtain multiple groups of square grids, each group of square grids includes a plurality of square grids, wherein , each time the square plane is divided into two equal parts, a set of square grids can be obtained.
  • the square plane can be divided into two equal parts in the horizontal axis direction and the vertical axis direction multiple times to obtain multiple groups of square grids, each group of square grids includes a plurality of square grids, wherein the Multiple bisection divisions include the Nth bisection division and the N+1th bisection division, and after the Nth bisection division is performed, M square grids are obtained. After N+1 bisection division, M*4 square grids are obtained, and the M*4 square grids are the horizontal axis direction of each square grid in the M square grids obtained by bisection along the longitudinal axis.
  • a group of square grids can be obtained after the square plane is divided into two equal parts in the horizontal direction and the vertical axis direction.
  • Each square grid in the obtained set of square grids is divided into two equal parts in the horizontal axis direction and the vertical axis direction to obtain a new set of square grids.
  • 16 square grids as shown in FIG. 6 can be obtained, and then in the next mesh division process, the Each of the 16 square grids shown is divided into two equal parts in the direction of the horizontal axis and the vertical axis, so as to obtain 64 square grids (specifically, as shown in FIG. 7 ).
  • each code corresponds to a square grid, and each code is used to indicate an area covered by the corresponding square grid.
  • the plurality of square grids can be encoded,
  • each code corresponds to a square grid, and each code is used to indicate the area covered by the corresponding square grid.
  • multiple square grids included in each group of square grids in the multiple groups of square grids may be coded to obtain multiple sets of codes, and each set of codes includes multiple codes.
  • step 303 and step 304 is not limited in the embodiment of the present application.
  • a A group of square grids, and then the group of square grids can be coded to obtain the code of the group of square grids, and then the next bisection division can be performed, that is, the one obtained by the previous division can be obtained.
  • Each square grid in the set of square grids is divided into two equal parts in the horizontal and vertical axis directions to obtain a new set of square grids, which can then be coded to obtain the The encoding of the group square grid, and so on.
  • each code obtained after the square grid is coded is used to indicate the area covered by the corresponding square grid, and the present application does not limit a specific coding manner.
  • the target projection as an example obtained by performing UTM projection on the surface of the earth on a plane.
  • each new square lattice generated by the above division continues to be divided into two equal parts in the horizontal and vertical directions to form four square lattices of equal size, and the square lattices generated by each division are divided according to certain rules for encoding. This cycle is repeated until the coding of the square grid generated by each subdivision meets the requirements, for example, the side length of the smallest square grid is 0.5 meters.
  • a square grid is obtained by dividing a square plane, which has the characteristics of a global discrete grid such as global coverage, uniqueness, hierarchy, hierarchy membership, and coding operation, and is sampled or counted according to grid sampling or statistics.
  • the size of the data grid is square, and the size of the square can be selected according to the needs of the application, or the appropriate scaling factor s can be selected when constructing the mesh, so as to meet the application requirements of the algorithm sensitive to the grid size.
  • grid division is performed on the square plane constructed on the target projection, so that the division size of the earth surface corresponding to the square grid obtained by the same level of division is basically the same, which is beneficial to the grid data size.
  • Sensitive computations such as AI models with convolutional neural networks (CNN) as feature extraction layers.
  • CNN convolutional neural networks
  • a height dimension can also be added, a three-dimensional rectangular coordinate can be constructed, and the height dimension can be divided into equal distances, which can be extended to a 3D equidistant global discrete grid.
  • An embodiment of the present application provides a grid coding method, the method includes: acquiring a target projection, where the target projection is a projection of the earth surface on a plane, and any two regions of the same area on the earth surface are The area difference between the two corresponding projection areas on the target projection is less than a preset value; according to the target projection, a square plane is determined, and the area where the target projection is located is covered by the square plane; Perform multiple bisections along the horizontal axis and the vertical axis to obtain a plurality of square grids; encode the plurality of square grids to obtain a plurality of codes, each code corresponding to a square grid , and each code is used to indicate the area covered by the corresponding square grid.
  • FIG. 8 is a schematic structural diagram of a trellis encoding apparatus provided by an embodiment of the present application.
  • a trellis encoding apparatus 800 provided by an embodiment of the present application includes:
  • the obtaining module 801 is configured to obtain a target projection, where the target projection is a projection of the earth surface on a plane.
  • step 301 For the specific description of the obtaining module 801, reference may be made to step 301 and the description of the corresponding embodiment, which will not be repeated here.
  • a determination module 802 configured to determine a square plane according to the target projection, and the area where the target projection is located is covered by the square plane;
  • step 302 For the specific description of the determination module 802, reference may be made to step 302 and the description of the corresponding embodiment, which will not be repeated here.
  • a meshing module 803 configured to perform multiple bisection divisions on the square plane in the horizontal axis direction and the vertical axis direction to obtain a plurality of square meshes;
  • step 303 For the specific description of the meshing module 803, reference may be made to step 303 and the description of the corresponding embodiment, which will not be repeated here.
  • the encoding module 804 is configured to encode the plurality of square grids to obtain a plurality of codes, each code corresponds to a square grid, and each code is used to indicate an area covered by the corresponding square grid.
  • step 304 For the specific description of the encoding module 804, reference may be made to step 304 and the description of the corresponding embodiment, which will not be repeated here.
  • the first target point and the second target point are any two points on the target projection
  • the first target point is the projection of the first position point on the surface of the earth
  • the second target point is the projection of the second position point on the surface of the earth
  • the distance between the first target point and the second target point is the first distance
  • the sum of the first position point and the obtained The distance between the second position points is the second distance
  • the ratio between the first distance and the second distance is within a preset range.
  • the earth surface includes an equatorial latitude and a target longitude perpendicular to the equatorial latitude
  • the target projection includes the projected equatorial latitude and the projected target longitude
  • the horizontal axis of the square plane The direction is consistent with the direction in which the projected equatorial latitude is located, and the direction of the longitudinal axis of the square plane is consistent with the direction in which the projected target meridian is located.
  • the determining module is configured to determine the side length of the square plane according to the target length projected by the target in the direction of the projected equatorial latitude, wherein the square The side length of the plane is greater than or equal to the target length, and the side length of the square plane is a preset multiple of an integer power of 2;
  • the square plane is determined so that the area where the target projection is located is covered by the square plane.
  • the determining module is configured to obtain the preset multiple, and determine the target length according to the projected target length in the direction of the projected equatorial latitude and the preset multiple. The side length of the square plane.
  • the central transverse axis of the square plane overlaps the projected equatorial latitude.
  • the meshing module is configured to perform multiple bisection divisions on the square plane in the horizontal axis direction and the vertical axis direction to obtain multiple sets of square meshes, each set of The square grid includes a plurality of square grids, wherein the multiple bisections include the Nth bisection and the N+1th bisection, and the Nth bisection is performed After the division, M square grids are obtained, and after the N+1th bisection division is performed, M*4 square grids are obtained, and the M*4 square grids are for the M Each square grid in the square grid is obtained by bisection in the horizontal and vertical directions;
  • the encoding module is configured to encode multiple square grids included in each group of square grids in the multiple groups of square grids, so as to obtain multiple sets of codes, and each group of codes includes multiple codes.
  • the target projection is obtained by performing a global projection or a zonal projection on the surface of the earth.
  • An embodiment of the present application provides a grid coding device, including: an acquisition module 801, configured to acquire a target projection, where the target projection is a projection of the earth surface on a plane; a determination module 802, configured to, according to the target projection, A square plane is determined, and the area where the target projection is located is covered by the square plane; the meshing module 803 is used to perform multiple bisection divisions on the square plane in the horizontal and vertical directions to obtain obtaining a plurality of square grids; the encoding module 804 is configured to encode the plurality of square grids to obtain a plurality of codes, each code corresponds to a square grid, and each code is used to indicate a corresponding square The area covered by the grid. Grid division is performed on the square plane constructed on the target projection, so that the division size of the earth surface corresponding to the square grid obtained by the same level of division is basically the same, which is conducive to the calculation that is sensitive to the size of grid data.
  • the present application also provides a non-volatile computer-readable storage medium, the non-volatile computer-readable storage medium contains computer instructions, and when the computer instructions are executed by a computer, the trellis coding method in the above embodiments can be implemented .
  • FIG. 9 is a schematic structural diagram of a computer system provided in this embodiment.
  • the computer system can be a terminal device (or called an intelligent terminal) or a server.
  • the computer system includes a communication module 810 , a sensor 820 , a user input module 830 , an output module 840 , a processor 850 , an audio and video input module 860 , a memory 870 and a power supply 880 .
  • the computer system provided in this embodiment may further include an NPU 890 .
  • Communication module 810 may include at least one module that enables communication between the computer system and a communication system or other computer system.
  • the communication module 810 may include one or more of a wired network interface, a broadcast receiving module, a mobile communication module, a wireless Internet module, a local area communication module, and a location (or positioning) information module, among others.
  • a wired network interface may include one or more of a wireless network interface, a broadcast receiving module, a mobile communication module, a wireless Internet module, a local area communication module, and a location (or positioning) information module, among others.
  • Sensors 820 may sense the current state of the system, such as open/closed state, position, contact with the user, direction, and acceleration/deceleration, and may generate sensed signals for controlling the operation of the system.
  • the user input module 830 is used to receive input digital information, character information or contact touch operation/non-contact gesture, and receive signal input related to user settings and function control of the system.
  • User input module 830 includes a touch panel and/or other input devices.
  • the output module 840 includes a display panel for displaying information input by the user, information provided to the user, various menu interfaces of the system, and the like.
  • the display panel may be configured in the form of a liquid crystal display (liquid crystal display, LCD) or an organic light-emitting diode (organic light-emitting diode, OLED).
  • the touch panel may cover the display panel to form a touch display screen.
  • the output module 840 may further include an audio output module, an alarm, a haptic module, and the like.
  • the audio and video input module 860 is used for inputting audio signals or video signals.
  • the audio and video input module 860 may include a camera and a microphone.
  • the power supply 880 may receive external power and internal power under the control of the processor 850 and provide power required for the operation of the various components of the system.
  • the processor 850 includes one or more processors, for example, the processor 850 may include a central processing unit and a graphics processor.
  • the central processing unit has multiple cores and belongs to a multi-core processor. These multiple cores can be integrated on the same chip, or can each be independent chips.
  • the memory 870 stores computer programs including an operating system program 872, an application program 871, and the like.
  • Typical operating systems such as Microsoft's Windows, Apple's MacOS, etc. are used for desktop or notebook systems, and another example is the Android-based system developed by Google, which is used for mobile terminals.
  • the methods provided in the foregoing embodiments may be implemented in software, which may be considered as the specific implementation of the operating system program 872 .
  • the memory 870 may be one or more of the following types: flash memory, hard disk type memory, micro multimedia card type memory, card memory (eg SD or XD memory), random access memory , RAM), static random access memory (static RAM, SRAM), read only memory (read only memory, ROM), electrically erasable programmable read only memory (electrically erasable programmable read only memory, EEPROM), programmable read only memory Memory (programmable ROM, PROM), rollback protected memory block (replay protected memory block, RPMB), magnetic storage, magnetic disk or optical disk.
  • the memory 870 can also be a network storage device on the Internet, and the system can perform operations such as updating or reading on the memory 870 on the Internet.
  • the processor 850 is used to read the computer program in the memory 870, and then execute the method defined by the computer program. For example, the processor 850 reads the operating system program 872 to run the operating system and implement various functions of the operating system in the system, or read the operating system program 872. One or more application programs 871 are taken to run the application on the system.
  • the memory 870 also stores other data 873 in addition to computer programs.
  • the NPU 890 is mounted on the main processor 850 as a co-processor for executing tasks assigned to it by the main processor 850.
  • the NPU 890 can be called by one or more sub-threads of the face recognition TA to implement some complex algorithms involved in the face recognition.
  • the sub-thread of the face recognition TA runs on multiple cores of the main processor 850 , and then the main processor 850 calls the NPU 890 , and the result implemented by the NPU 890 is returned to the main processor 850 .
  • connection relationship of the above modules is only an example, and the trellis coding method provided by any embodiment of the present application can also be applied to terminal devices or servers in other connection modes, for example, all modules are connected through a bus.
  • FIG. 10 is a schematic structural diagram of an NPU 900 provided in this embodiment.
  • the NPU900 is connected to the main processor and external memory.
  • the core part of the NPU 900 is the arithmetic circuit 903, which is controlled by the controller 904 to extract the data in the memory and perform mathematical operations.
  • the arithmetic circuit 903 includes multiple process engines (PE) inside.
  • PE process engines
  • arithmetic circuit 903 is a two-dimensional systolic array.
  • the arithmetic circuit 903 may also be a one-dimensional systolic array or other electronic circuitry capable of performing mathematical operations such as multiplication and addition.
  • the arithmetic circuit 903 is a general-purpose matrix processor.
  • the operation circuit 903 fetches the data corresponding to the matrix B from the weight memory 902 and buffers it on each PE of the operation circuit 903 .
  • the operation circuit 903 takes the data of the matrix A from the input memory 901 and performs the matrix operation on the matrix B, and stores the partial result or the final result of the matrix in the accumulator 908 .
  • Unified memory 906 is used to store input data and output data.
  • the weight data is directly transferred to the weight memory 902 through the storage unit access controller 905 (eg, direct memory access controller, DMAC).
  • DMAC direct memory access controller
  • Input data is also transferred to the unified storage 906 through the storage unit access controller 905 .
  • the bus interface unit 910 (bus interface unit, BIU) is used for the interaction of the AXI (advanced extensible interface) bus with the storage unit access controller 905 and the instruction fetch memory 909 (instruction fetch buffer).
  • AXI advanced extensible interface
  • the bus interface unit 910 is used for the instruction fetch memory 909 to obtain instructions from the external memory, and is also used for the storage unit access controller 905 to obtain the original data of the input matrix A or the weight matrix B from the external memory.
  • the storage unit access controller 905 is mainly used to transfer the input data in the external memory to the unified memory 906 , the weight data to the weight memory 902 , or the input data data to the input memory 901 .
  • the vector calculation unit 907 usually includes a plurality of operation processing units, and further processes the output of the operation circuit 903, such as vector multiplication, vector addition, exponential operation, logarithmic operation, and/or size comparison, etc., if necessary.
  • vector computation unit 907 can store the processed vectors into unified memory 906 .
  • the vector calculation unit 907 may apply a nonlinear function to the output of the arithmetic circuit 903, such as a vector of accumulated values, to generate activation values.
  • the vector computation unit 907 generates normalized values, merged values, or both.
  • the processed vector can be used as an activation input to arithmetic circuit 903 .
  • the instruction fetch memory 909 connected to the controller 904 is used to store the instructions used by the controller 904 .
  • the unified memory 906, the input memory 901, the weight memory 902 and the instruction fetch memory 909 are all On-Chip memories.
  • the external memory in the figure is independent of the NPU hardware architecture.
  • the method for configuring an address translation relationship can also be applied to a non-terminal computer device, such as a cloud server.
  • the disclosed system, apparatus and method may be implemented in other manners.
  • the apparatus embodiments described above are only illustrative.
  • the division of units is only a logical function division.
  • there may be other division methods for example, multiple units or components may be combined or integrated. to another system, or some features can be ignored, or not implemented.
  • the shown or discussed mutual coupling or direct coupling or communication connection may be through some interfaces, indirect coupling or communication connection of devices or units, and may be in electrical, mechanical or other forms.
  • the units described as separate components may or may not be physically separated, and components displayed as units may or may not be physical units, that is, may be located in one place, or may be distributed to multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution in this embodiment.
  • each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically alone, or two or more units may be integrated into one unit.
  • the above-mentioned integrated units may be implemented in the form of hardware, or may be implemented in the form of software functional units.
  • the integrated unit if implemented in the form of a software functional unit and sold or used as an independent product, may be stored in a computer-readable storage medium.
  • the technical solutions of the present application can be embodied in the form of software products in essence, or the parts that contribute to the prior art, or all or part of the technical solutions, and the computer software products are stored in a storage medium , including several instructions to cause a computer device (which may be a personal computer, a server, or other network device, etc.) to execute all or part of the steps of the method described in the embodiment of FIG. 3 of the present application.
  • the aforementioned storage medium includes: U disk, removable hard disk, read-only memory (ROM), random access memory (RAM), magnetic disk or optical disk and other media that can store program codes .

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Multimedia (AREA)
  • Computer Graphics (AREA)
  • Geometry (AREA)
  • Software Systems (AREA)
  • Error Detection And Correction (AREA)
  • Image Processing (AREA)

Abstract

本申请实施例公开了一种网格编码方法，包括：获取目标投影，目标投影为地球表面在平面上的投影，且地球表面上任意两个相同面积的区域在目标投影上对应的两个投影区域之间的面积差异小于预设值；根据目标投影，确定正方形平面，目标投影所在的区域被正方形平面覆盖；对正方形平面在横轴方向和纵轴方向进行多次二等分剖分，以得到多个正方形网格；对多个正方形网格进行编码，以得到多个编码，每个编码对应于一个正方形网格，且每个编码用于指示对应的正方形网格所覆盖的区域。本申请在目标投影上构建的正方形平面上进行网格剖分，使得同一层级剖分得到的正方形网格所对应的地球表面的划分大小基本一致，有利于对网格数据大小敏感的计算。

Description

一种网格编码方法及计算机系统
本申请要求于2020年11月26日提交中国专利局、申请号为202011352980.3、发明名称为“一种网格编码方法及计算机系统”的中国专利申请的优先权，其全部内容通过引用结合在本申请中。
技术领域
本申请涉及计算机领域，尤其涉及一种网格编码方法及计算机系统。
背景技术
全球离散网格(discrete global grid,DGG)是基于球面(或椭球体)的一种可以无限细分、但又不改变其形状的地球体拟合格网，当细分到一定程度时，可以达到模拟地球表面的目的。DGG具有层次性和全球连续性等特征，克服了许多限制地理信息系统应用的约束和不确定性，使得在地球上任何位置获取的任何分辨率(不同精度)的空间数据都可以规范地表达和分析，并能用确定的精度进行多尺度操作。
DGG包括GeoSOT网格，GeoSOT网格属于等经纬度的四叉树剖分网格体系，首先将地球表面的空间扩展为正方形的平面空间，GeoSOT剖分0级网格定义为以赤道与本初子午线交点为中心的的平面空间方格，GeoSOT剖分1级网格是在0级基础上平均四等分，之后的剖分层次按照上述四叉树的原则以此类推。
然而，由于GeoSOT属于等经纬度的四叉树剖分网格体系，每一个层级中，网格欧式距离大小不同，网格的大小从赤道向两极逐步减小，对网格大小敏感的计算泛化较为困难。
发明内容
第一方面，本申请提供了一种网格编码方法，所述方法包括：
获取目标投影，所述目标投影为地球表面在平面上的投影，且所述地球表面上任意两个相同面积的区域在所述目标投影上对应的两个投影区域之间的面积差异小于预设值；可以利用地图投影方法将地球表面转换到平面上，以获取目标投影，其中，地图投影方法可以是整体投影，也可以是分带投影。等经纬度的地球表面平面是地球表面经过一定比例的缩小以及形变得到的，在地球表面的等经纬度平面上，远离赤道的区域的缩小比例远远小于靠近赤道的区域的缩小比例，也就是说，地球表面上远离赤道的区域与靠近赤道的区域在等经纬度的地球表面平面上所对应的两个区域之间的面积差异很大。而本实施例中，所述地球表面上任意两个相同面积的区域在所述目标投影上对应的两个投影区域之间的面积差异小于预设值，也就是说，地球表面上任意两个相同面积的区域经过一定比例的缩小得到所述目标投影，且地球表面上任意两个相同面积的区域经过的缩小比例差异很小。本申请并不限定所述地球表面上任意两个相同面积的区域在所述目标投影上对应的两个投影区域之间的面积是完全一致的，而是差异小于预设值。根据所述目标投影，确定正方形平面，所述目标投影所在的区域被所述正方形平面覆盖；在得到目标投影之后可以根据所述目标投影，确定正方形平面，所述目标投影所在的区域被所述正方形平面覆盖。为了能将目标 投影所在的区域完全进行覆盖，正方形平面的面积应大于目标投影，且正方形平面的边应该目标投影的区域之外。应理解，目标投影为非正方形，因此正方形平面中包括与目标投影重叠的区域，也包括与目标投影不重叠的区域。对所述正方形平面在横轴方向和纵轴方向进行多次二等分剖分，以得到多个正方形网格；每次对正方形平面在横轴方向和纵轴方向进行二等分剖分之后可以得到一组正方形网格，在下一次进行二等分剖分时，可以对上一次进行剖分得到的一组正方形网格中的每个正方形网格进行在横轴方向和纵轴方向的二等分剖分，得到新的一组正方形网格。对所述多个正方形网格进行编码，以得到多个编码，每个编码对应于一个正方形网格，且每个编码用于指示对应的正方形网格所覆盖的区域。在对所述正方形平面在横轴方向和纵轴方向进行一次二等分剖分之后，可以得到一组正方形网格，之后可以对该组正方形网格进行编码，以得到该组正方形网格的编码，之后可以进行下一次的二等分剖分，也就是可以对上一次进行剖分得到的一组正方形网格中的每个正方形网格进行在横轴方向和纵轴方向的二等分剖分，得到新的一组正方形网格，之后可以对该组正方形网格进行编码，以得到该组正方形网格的编码，以此类推。
应理解，对正方形网格进行编码后得到的每个编码用于指示对应的正方形网格所覆盖的区域，本申请并不限定具体的编码方式。
本实施例在目标投影上构建的正方形平面上进行网格剖分，使得同一层级剖分得到的正方形网格所对应的地球表面的划分大小基本一致，有利于对网格数据大小敏感的计算。
在一种可能的实现中，第一目标点和第二目标点为所述目标投影上任意的两个点，所述第一目标点为所述地球表面上的第一位置点的投影，所述第二目标点为所述地球表面上的第二位置点的投影，所述第一目标点和所述第二目标点之间的距离为第一距离，所述第一位置点和得所述第二位置点之间的距离为第二距离，所述第一距离和所述第二距离之间的比值在预设范围内。通过目标投影可以较为准确的表征地球表面各个点的欧氏距离，具体的，目标投影上的任一两个点之间的距离可以表征相应的地球表面上两个物理位置点之间的欧氏距离，所谓欧氏距离，是指欧几里得度量，在数学中，欧氏距离或欧几里得度量是欧几里得空间中两点间的“普通”(即直线)距离。目标投影是地球表面的近似不失真投影，目标投影上各个点之间的距离是地球表面经过一定比例的缩小得到的，且地球表面上各个点之间的距离的缩小比例基本一致，因此通过目标投影可以较为准确的表征地球表面各个点的欧氏距离。
在一种可能的实现中，所述地球表面包括赤道纬线以及与所述赤道纬线垂直的目标经线，所述目标投影包括投影后的赤道纬线以及投影后的目标经线，所述正方形平面的横轴方向与所述投影后的赤道纬线所在的方向一致，所述正方形平面的纵轴方向与所述投影后的目标经线所在的方向一致。首先，需要确定正方形平面的方向，所谓方向是指正方形平面包括的横轴和纵轴的方向。正方形平面可以包括横轴和纵轴，其中，横轴可以是与正方形平面的一个边平行，纵轴与横轴垂直，更细节的，横轴可以是与正方形平面的一个边平行且通过正方形平面中心的轴，纵轴是与横轴垂直且通过正方形平面中心的轴。
在一种可能的实现中，所述根据所述目标投影，确定正方形平面，包括：
根据所述目标投影在所述投影后的赤道纬线所在的方向上的目标长度，确定所述正方形平面的边长，其中，所述正方形平面的边长大于或等于所述目标长度，且所述正方形平面的边长为2的整数次幂的预设倍数；之所以正方形平面的边长为2的整数次幂的预设倍数，是因为在进行后续的编码时，需要对正方形平面进行横轴方向和纵轴方向的二等分剖分，为了使剖分后可以得到2的整数次幂个正方形网格，需要保证正方形平面的边长为2的整数次幂的预设倍数，示例性的，正方形平面的边长可以为
Figure PCTCN2021133339-appb-000001
其中，L为目标投影中赤道纬线方向的长度，S为缩放系数，也就是上述的预设倍数。根据所述正方形平面的边长，确定正方形平面，以便所述目标投影所在的区域被所述正方形平面覆盖。本实施例中可以通过控制预设倍数的大小，进而控制剖分后得到的正方形网格的大小。
在一种可能的实现中，所述根据所述目标投影在所述投影后的赤道纬线所在的方向上的目标长度，确定所述正方形平面的边长，包括：
获取所述预设倍数，根据所述目标投影在所述投影后的赤道纬线所在的方向上的目标长度以及所述预设倍数，确定所述正方形平面的边长。在一种可能的实现中，所述正方形平面的中心横轴与所述投影后的赤道纬线重叠。
在一种可能的实现中，所述对所述正方形平面在横轴方向和纵轴方向进行多次二等分剖分，以得到多个正方形网格，包括：
对所述正方形平面在横轴方向和纵轴方向进行多次二等分剖分，以得到多组正方形网格，每组正方形网格包括多个正方形网格，其中，所述多次二等分剖分包括第N次二等分剖分以及第N+1次二等分剖分，且进行第N次二等分剖分后，得到M个正方形网格，在进行第N+1次二等分剖分后，得到M*4个正方形网格，且所述M*4个正方形网格为对所述M个正方形网格中的每个正方形网格在横轴方向和纵轴方向进行二等分剖分得到的；
所述对所述多个正方形网格进行编码，以得到多个编码，包括：对所述多组正方形网格中每组正方形网格包括的多个正方形网格进行编码，以得到多组编码，每组编码包括多个编码。
在一种可能的实现中，所述目标投影为对所述地球表面进行整体投影或分带投影得到的。
第二方面，本申请提供了一种网格编码装置，所述装置包括：
获取模块，用于获取目标投影，所述目标投影为地球表面在平面上的投影；
确定模块，用于根据所述目标投影，确定正方形平面，所述目标投影所在的区域被所述正方形平面覆盖；
网格剖分模块，用于对所述正方形平面在横轴方向和纵轴方向进行多次二等分剖分，以得到多个正方形网格；
编码模块，用于对所述多个正方形网格进行编码，以得到多个编码，每个编码对应于一个正方形网格，且每个编码用于指示对应的正方形网格所覆盖的区域。
在一种可能的实现中，第一目标点和第二目标点为所述目标投影上任意的两个点，所述第一目标点为所述地球表面上的第一位置点的投影，所述第二目标点为所述地球表面上的第二位置点的投影，所述第一目标点和所述第二目标点之间的距离为第一距离，所述第一位置点和得所述第二位置点之间的距离为第二距离，所述第一距离和所述第二距离之间的比值在预设范围内。
在一种可能的实现中，所述地球表面包括赤道纬线以及与所述赤道纬线垂直的目标经线，所述目标投影包括投影后的赤道纬线以及投影后的目标经线，所述正方形平面的横轴方向与所述投影后的赤道纬线所在的方向一致，所述正方形平面的纵轴方向与所述投影后的目标经线所在的方向一致。
在一种可能的实现中，所述确定模块，用于根据所述目标投影在所述投影后的赤道纬线所在的方向上的目标长度，确定所述正方形平面的边长，其中，所述正方形平面的边长大于或等于所述目标长度，且所述正方形平面的边长为2的整数次幂的预设倍数；
根据所述正方形平面的边长，确定正方形平面，以便所述目标投影所在的区域被所述正方形平面覆盖。
在一种可能的实现中，所述确定模块，用于获取所述预设倍数，根据所述目标投影在所述投影后的赤道纬线所在的方向上的目标长度以及所述预设倍数，确定所述正方形平面的边长。
在一种可能的实现中，所述正方形平面的中心横轴与所述投影后的赤道纬线重叠。
在一种可能的实现中，所述网格剖分模块，用于对所述正方形平面在横轴方向和纵轴方向进行多次二等分剖分，以得到多组正方形网格，每组正方形网格包括多个正方形网格，其中，所述多次二等分剖分包括第N次二等分剖分以及第N+1次二等分剖分，且进行第N次二等分剖分后，得到M个正方形网格，在进行第N+1次二等分剖分后，得到M*4个正方形网格，且所述M*4个正方形网格为对所述M个正方形网格中的每个正方形网格在横轴方向和纵轴方向进行二等分剖分得到的；
所述编码模块，用于对所述多组正方形网格中每组正方形网格包括的多个正方形网格进行编码，以得到多组编码，每组编码包括多个编码。
在一种可能的实现中，所述目标投影为对所述地球表面进行整体投影或分带投影得到的。
第三方面，本申请提供一种计算机系统，所述计算机系统包括存储器和处理器，所述存储器用于存储计算机可读指令(或者称之为计算机程序)，所述处理器用于读取所述计算机 可读指令以实现前述任意实现方式提供的方法。
第四方面，本申请提供一种计算机存储介质，该计算机存储介质可以是非易失性的。该计算机存储介质中存储有计算机可读指令，当该计算机可读指令被处理器执行时实现前述任意实现方式提供的方法。
第五方面，本申请提供一种计算机程序产品，该计算机程序产品中包含计算机可读指令，当该计算机可读指令被处理器执行时实现前述任意实现方式提供的方法。
本申请实施例提供了一种网格编码方法，所述方法包括：获取目标投影，所述目标投影为地球表面在平面上的投影，且所述地球表面上任意两个相同面积的区域在所述目标投影上对应的两个投影区域之间的面积差异小于预设值；根据所述目标投影，确定正方形平面，所述目标投影所在的区域被所述正方形平面覆盖；对所述正方形平面在横轴方向和纵轴方向进行多次二等分剖分，以得到多个正方形网格；对所述多个正方形网格进行编码，以得到多个编码，每个编码对应于一个正方形网格，且每个编码用于指示对应的正方形网格所覆盖的区域。通过上述方式，在目标投影上构建的正方形平面上进行网格剖分，使得同一层级剖分得到的正方形网格所对应的地球表面的划分大小基本一致，有利于对网格数据大小敏感的计算。
附图说明
图1为一种GeoSOT网格的示意；
图2为一种GeoSOT网格的示意；
图3为本申请实施例提供的一种网格编码方法的流程示意；
图4为本申请实施例提供的一种目标投影的示意；
图5为本申请实施例提供的一种正方形平面的示意；
图6为本申请实施例提供的一种正方形网格的划分示意；
图7为本申请实施例提供的一种正方形网格的划分示意；
图8为本申请实施例提供的一种网格编码装置的结构示意；
图9为本实施例提供的一种计算机系统的结构示意图；
图10为本实施例提供的一种NPU的结构示意图。
具体实施方式
下面结合附图，对本申请的实施例进行描述，显然，所描述的实施例仅仅是本申请一部分的实施例，而不是全部的实施例。本领域普通技术人员可知，随着技术的发展和新场景的出现，本申请实施例提供的技术方案对于类似的技术问题，同样适用。
本申请的说明书和权利要求书及上述附图中的术语“第一”、“第二”等是用于区别类似的对象，而不必用于描述特定的顺序或先后次序。应该理解这样使用的数据在适当情况下 可以互换，以便这里描述的实施例能够以除了在这里图示或描述的内容以外的顺序实施。此外，术语“包括”和“具有”以及他们的任何变形，意图在于覆盖不排他的包含，例如，包含了一系列步骤或模块的过程、方法、系统、产品或设备不必限于清楚地列出的那些步骤或模块，而是可包括没有清楚地列出的或对于这些过程、方法、产品或设备固有的其它步骤或模块。在本申请中出现的对步骤进行的命名或者编号，并不意味着必须按照命名或者编号所指示的时间/逻辑先后顺序执行方法流程中的步骤，已经命名或者编号的流程步骤可以根据要实现的技术目的变更执行次序，只要能达到相同或者相类似的技术效果即可。
另外，在本发明的描述中，除非另有说明，“多个”的含义是两个或两个以上。本申请中的术语“和/或”或字符“/”，仅仅是一种描述关联对象的关联关系，表示可以存在三种关系，例如，A和/或B，或A/B，可以表示：单独存在A，同时存在A和B，单独存在B这三种情况。
全球离散网格(Discrete Global Grid,DGG)是基于球面(或椭球体)的一种可以无限细分、但又不改变其形状的地球体拟合格网，当细分到一定程度时，可以达到模拟地球表面的目的。DGG具有层次性和全球连续性等特征，克服了许多限制地理信息系统应用的约束和不确定性，使得在地球上任何位置获取的任何分辨率(不同精度)的空间数据都可以规范地表达和分析，并能用确定的精度进行多尺度操作。
按照构建网格的方法的不同，全球离散网格系统可大致划分为：等经纬度全球网格、变经纬度全球网格、自适应全球网格和正多面体全球网格系统四类，如GeoSOT是等经纬度全球网格，退化四叉树网格是变经纬度全球网格，数字高程模型(digital elevation model，DEM)数据的细节层次(levels of detail，LOD)模型是自适应全球网格，三角型网格、菱形网格、六边形网格等是正多面体全球网格。
GeoSOT网格属于等经纬度的四叉树剖分网格体系，首先将地球表面360°*180°的空间扩展为512°*512°的平面空间，GeoSOT剖分0级网格定义为以赤道与本初子午线交点为中心的512°*512°的平面空间方格，GeoSOT剖分1级网格是在0级基础上平均四等分，每个1级网格大小为256°*256°，具体可以如图1所示。GeoSOT剖分2级网格是在1级基础上平均四等分，每个1级网格大小为128°*128°，以下剖分层次按照上述四叉树的原则以此类推。GeoSOT剖分9级网格大小为1°*1°，9级以上网格为GeoSOT的“度”级网格，第10～15级网格为“分”级网格，“分”级面片起始点为9级网格的1°面片，网格的起始数值空间大小由60’外延到64’，GeoSOT 10级剖分网格以64’*64’大小平均分为四份，每个10级网格大小为32’*32’。10～15级网格为“分”级网格，剖分方式按照上述规则递归，如图2所示。16～21级为“秒”级网格，“秒”级网格剖分方式参照分级网格，即15级1’*1’面片的值域范围外延为64”*64”,上述各级面片按均按照四分法进行剖分。秒级一下22～32级网格严格按照四份法进行剖分，第32级网格大小为1/2048”*1/2048”。
然而在GeoSOT的每一个层级中，网格的欧式距离大小不同，网格的大小从赤道向两极逐步减小，对网格大小敏感的计算泛化较为困难，如在靠近赤道区域训练的人工智能(artificial intelligence，AI)模型泛化应用到冰岛地区，或训练收敛困难或模型存在精度损失。
为了解决上述问题，本申请实施例提供了一种网格编码方法，参照图3，图3为本申请实施例提供的一种网格编码方法的流程示意，如图3所示，本申请实施例提供的网格编码方法包括：
301、获取目标投影，所述目标投影为地球表面在平面上的投影，且所述地球表面上任意两个相同面积的区域在所述目标投影上对应的两个投影区域之间的面积差异小于预设值。
本申请实施例中，可以利用地图投影方法将地球表面转换到平面上，以获取目标投影，其中，地图投影方法可以是整体投影，也可以是分带投影，也就是说所述目标投影可以为对所述地球表面进行整体投影或分带投影得到的。
在现有的实现中，GeoSOT网格是在通过对地球表面进行形变之后得到的等经纬度正方形表面上直接进行网格划分得到的，经纬度划分的不同的网格中的欧式距离大小不同。本申请实施例中，不是对目标投影进行形变得到正方形平面，而是在目标投影之外重新确定一个可以覆盖该目标投影的正方形平面。
等经纬度的地球表面平面是地球表面经过一定比例的缩小以及形变得到的，在地球表面的等经纬度平面上，远离赤道的区域的缩小比例远远小于靠近赤道的区域的缩小比例，也就是说，地球表面上远离赤道的区域与靠近赤道的区域在等经纬度的地球表面平面上所对应的两个区域之间的面积差异很大。而本实施例中，所述地球表面上任意两个相同面积的区域在所述目标投影上对应的两个投影区域之间的面积差异小于预设值，也就是说，地球表面上任意两个相同面积的区域经过一定比例的缩小得到所述目标投影，且地球表面上任意两个相同面积的区域经过的缩小比例差异很小。
通过目标投影可以较为准确的表征地球表面各个点的欧氏距离，具体的，目标投影上的任一两个点之间的距离可以表征相应的地球表面上两个物理位置点之间的欧氏距离，所谓欧氏距离，是指欧几里得度量，在数学中，欧氏距离或欧几里得度量是欧几里得空间中两点间的“普通”(即直线)距离。
图1中地球表面的等经纬度平面上的两个点之间的距离不能准确的表征地球表面各个点的欧氏距离。具体的，地球表面的等经纬度平面是地球表面经过一定比例的缩小以及形变得到的，在地球表面的等经纬度平面上，远离赤道的两个点之间的距离的缩小比例远远小于靠近赤道的两个点之间的距离的缩小比例，例如等经纬度平面上靠近赤道的两个点之间的距离表征的对应的地球表面实际物理位置点之间的距离，相比等经纬度平面上远离赤道的具有相同距离的两个点表征的对应的地球表面实际物理位置点之间的距离，是更长的。
而本实施例中，目标投影是地球表面的近似不失真投影，目标投影上各个点之间的距离是地球表面经过一定比例的缩小得到的，且地球表面上各个点之间的距离的缩小比例基本一致，因此通过目标投影可以较为准确的表征地球表面各个点的欧氏距离。具体的，第一目标点和第二目标点为所述目标投影上任意的两个点，所述第一目标点为所述地球表面上的第一位置点的投影，所述第二目标点为所述地球表面上的第二位置点的投影，所述第一目标点和所述第二目标点之间的距离为第一距离，所述第一位置点和得所述第二位置点之间的距离为第二距离，所述第一距离和所述第二距离之间的比值在预设范围内。预设范 围可以根据实际情况选择，预设范围可以表征进行平面投影时地球表面的缩小比例，由于地球表面上任意两个点之间的距离进行投影时的缩小比例都在预设范围内，则通过目标投影可以较为准确的表征地球表面各个点的欧氏距离。
应理解，本申请并不限定所述地球表面上任意两个相同面积的区域在所述目标投影上对应的两个投影区域之间的面积是完全一致的，而是差异小于预设值。
应理解，本实施例中并不限定在进行地球表面的投影时各个点之间距离的缩小比例都是完全一致的，而是在预设范围内。
具体的，以目标投影为对地球表面在平面上进行通用横墨卡托(universal transverse mercator，UTM)投影得到的为例。
本实施例中，UTM投影是一种横轴等角割椭圆柱投影，椭圆柱割地球于南纬80°、北纬84°的两条等高圈，投影后两条相割的经线上没有变形，而中央经线上长度比为0.9996。该投影的中央经线投影为纵坐标轴，赤道纬线投影后的直线为横轴。对于UTM投影的6°分带是将全球划分为60个投影带，每带经差为6°。从西经180°至西经174°之间为第1带，连续向东编号。
UTM平面坐标(x,y)和大地坐标(L,B)之间的关系可以通过如下公式表示：
Figure PCTCN2021133339-appb-000002
Figure PCTCN2021133339-appb-000003
其中：T＝tan 2B；C＝e′ 2cos 2B；A＝(L-L 0)cosB；
Figure PCTCN2021133339-appb-000004
Figure PCTCN2021133339-appb-000005
其中：a为地球椭球体长半轴，b为地球椭球体短半轴，e为第一偏心率，e′为第二偏心率，L 0为分带的中央经度。
如图4所示，图4为本申请实施例提供的一种目标投影的示意，图4中示出的目标投影为对地球表面进行UTM投影得到的，其中目标投影的中央经线投影为纵坐标轴，赤道纬线投影后的直线为横坐标轴。UTM投影将全球划分为60个投影带，每带的经度跨度为6°。
302、根据所述目标投影，确定正方形平面，所述目标投影所在的区域被所述正方形平面覆盖。
本申请实施例中，在得到目标投影之后可以根据所述目标投影，确定正方形平面，所述目标投影所在的区域被所述正方形平面覆盖。为了能将目标投影所在的区域完全进行覆盖，正方形平面的面积应大于目标投影，且正方形平面的边应该目标投影的区域之外。
应理解，目标投影为非正方形，因此正方形平面中包括与目标投影重叠的区域，也包括与目标投影不重叠的区域。
接下来描述如何确定可以覆盖目标投影的正方形平面。
首先，需要确定正方形平面的方向，所谓方向是指正方形平面包括的横轴和纵轴的方向。正方形平面可以包括横轴和纵轴，其中，横轴可以是与正方形平面的一个边平行，纵轴与横轴垂直，更细节的，横轴可以是与正方形平面的一个边平行且通过正方形平面中心的轴，纵轴是与横轴垂直且通过正方形平面中心的轴。
本申请实施例中，所述地球表面包括赤道纬线以及与所述赤道纬线垂直的目标经线，所述目标投影包括投影后的赤道纬线以及投影后的目标经线，所述正方形平面的横轴方向与所述投影后的赤道纬线所在的方向一致，所述正方形平面的纵轴方向与所述投影后的目标经线所在的方向一致。如图5所示，图5为本申请实施例提供的一种正方形平面的示意，图5中正方形平面的横轴可以为轴x(m)，纵轴为y(m)，且所述正方形平面的横轴方向与所述投影后的赤道纬线所在的方向一致，所述正方形平面的纵轴方向与所述投影后的目标经线所在的方向一致。
接下来可以确定正方形平面的边长，具体的，在一种实现中，可以根据所述目标投影在所述投影后的赤道纬线所在的方向上的目标长度，确定所述正方形平面的边长，其中，所述正方形平面的边长大于或等于所述目标长度，且所述正方形平面的边长为2的整数次幂的预设倍数，并根据所述正方形平面的边长，确定正方形平面，以便所述目标投影所在的区域被所述正方形平面覆盖。在一种实现中，所述正方形平面的中心横轴与所述投影后的赤道纬线重叠。
之所以正方形平面的边长为2的整数次幂的预设倍数，是因为在进行后续的编码时，需要对正方形平面进行横轴方向和纵轴方向的二等分剖分，为了使剖分后可以得到2的整数次幂个正方形网格，需要保证正方形平面的边长为2的整数次幂的预设倍数，示例性的，正方形平面的边长可以为
Figure PCTCN2021133339-appb-000006
其中，L为目标投影中赤道纬线方向的长度，S为缩放系数，也就是上述实施例中的预设倍数。
本申请实施例中，可以获取所述预设倍数，根据所述目标投影在所述投影后的赤道纬线所在的方向上的目标长度以及所述预设倍数，确定所述正方形平面的边长，本实施例中可以通过控制预设倍数的大小，进而控制剖分后得到的正方形网格的大小。
本申请实施例中，可以以目标投影中赤道纬线投影后的直线为横轴，与之垂直的投影经线为纵轴，建立平面直角坐标系，将目标投影划分为四个部分，分别为东北投影分面、东南投影分面、西北投影分面和西南投影分面。设东北投影分面与西北投影分面在横轴方 向的长度取较大者，记为L米，将东北投影分面、东南投影分面、西北投影分面和西南投影分面均扩展到W*W大小的正方形平面，其中，
Figure PCTCN2021133339-appb-000007
其中s为缩放系数，以此得到2W*2W的正方形平面。
示例性的，以目标投影为对地球表面在平面上进行UTM投影得到的为例。以赤道纬线投影后的直线为横轴，即与UTM投影的横轴坐标重合，坐标原点为本初子午线投影为纵轴，建立二维平面直角坐标系，将地球投影平面划分为四个部分，分别为东北投影分面、东南投影分面、西北投影分面和西南投影分面，如图5所示。令缩放系数s＝1，
Figure PCTCN2021133339-appb-000008
将东北投影分面、东南投影分面、西北投影分面和西南投影分面均扩展到W*W大小的正方形格子。
303、对所述正方形平面在横轴方向和纵轴方向进行多次二等分剖分，以得到多个正方形网格。
本申请实施例中，可以对所述正方形平面在横轴方向和纵轴方向进行多次二等分剖分，以得到多组正方形网格，每组正方形网格包括多个正方形网格，其中，每次对所述正方形平面进行二等分剖分后，可以得到一组正方形网格。具体的，可以对所述正方形平面在横轴方向和纵轴方向进行多次二等分剖分，以得到多组正方形网格，每组正方形网格包括多个正方形网格，其中，所述多次二等分剖分包括第N次二等分剖分以及第N+1次二等分剖分，且进行第N次二等分剖分后，得到M个正方形网格，在进行第N+1次二等分剖分后，得到M*4个正方形网格，且所述M*4个正方形网格为对所述M个正方形网格中的每个正方形网格在横轴方向和纵轴方向进行二等分剖分得到的。
本申请实施例中，每次对正方形平面在横轴方向和纵轴方向进行二等分剖分之后可以得到一组正方形网格，在下一次进行二等分剖分时，可以对上一次进行剖分得到的一组正方形网格中的每个正方形网格进行在横轴方向和纵轴方向的二等分剖分，得到新的一组正方形网格。
具体的，可以参照图6，在进过两次二等分剖分之后，可以得到如图6所示的16个正方形网格，之后在下一次的网格剖分过程中，可以对图6所示的16个正方形网格中的每个正方形网格进行在横轴和纵轴方向的二等分剖分，以得到64个正方形网格(具体可以如图7所示)。
304、对所述多个正方形网格进行编码，以得到多个编码，每个编码对应于一个正方形网格，且每个编码用于指示对应的正方形网格所覆盖的区域。
本申请实施例中，在对所述正方形平面在横轴方向和纵轴方向进行多次二等分剖分，以得到多个正方形网格之后，可以对所述多个正方形网格进行编码，以得到多个编码，每个编码对应于一个正方形网格，且每个编码用于指示对应的正方形网格所覆盖的区域。
本申请实施例中，可以对所述多组正方形网格中每组正方形网格包括的多个正方形网格进行编码，以得到多组编码，每组编码包括多个编码。
应理解，本申请实施例中并不限定步骤303和步骤304之间的时序，具体的，在对所 述正方形平面在横轴方向和纵轴方向进行一次二等分剖分之后，可以得到一组正方形网格，之后可以对该组正方形网格进行编码，以得到该组正方形网格的编码，之后可以进行下一次的二等分剖分，也就是可以对上一次进行剖分得到的一组正方形网格中的每个正方形网格进行在横轴方向和纵轴方向的二等分剖分，得到新的一组正方形网格，之后可以对该组正方形网格进行编码，以得到该组正方形网格的编码，以此类推。
应理解，对正方形网格进行编码后得到的每个编码用于指示对应的正方形网格所覆盖的区域，本申请并不限定具体的编码方式。
示例性的，以目标投影为对地球表面在平面上进行UTM投影得到的为例。对扩展后的2W*2W大小的平面，在横轴和纵轴方向上分别进行二等分剖分，形成四个大小相等的正方形格子，并对每个剖分生成的正方形格子按照一定规则进行编码。之后对上述剖分生成的每个新正方形格子继续在横轴和纵轴方向上分别进行二等分剖分，形成四个大小相等的正方形格子，并对每个剖分生成的正方形格子按照一定规则进行编码。如此循环，直至每个剖分生成正方形格子的编码满足要求为止，如最小的正方形网格的边长大小为0.5米。
本申请实施例中，对正方形平面进行剖分得到的是正方形网格，具备全球覆盖、唯一性、层次性、层次隶属性和编码运算等全球离散网格的特征，且按照网格采样或统计的数据网格大小为正方形，且正方形的大小可以根据应用需要选择合适的层级，或者在构建剖分网格时选择合适的缩放系数s，从而满足对网格大小敏感的算法的应用需求。
且本实施例中，是在目标投影上构建的正方形平面上进行网格剖分，使得同一层级剖分得到的正方形网格所对应的地球表面的划分大小基本一致，有利于对网格数据大小敏感的计算，比如以卷积神经网络(convolutional neural networks，CNN)为特征提取层的AI模型。同时对于与等距全球离散网格局部区域对齐的局部坐标系下栅格数据的转换代价低。
应理解，在上述平面直角坐标基础上，还可以增加高度维，构建三维直角坐标，对高度维进行等距划分，可扩展到3D等距全球离散网格。
本申请实施例提供了一种网格编码方法，所述方法包括：获取目标投影，所述目标投影为地球表面在平面上的投影，且所述地球表面上任意两个相同面积的区域在所述目标投影上对应的两个投影区域之间的面积差异小于预设值；根据所述目标投影，确定正方形平面，所述目标投影所在的区域被所述正方形平面覆盖；对所述正方形平面在横轴方向和纵轴方向进行多次二等分剖分，以得到多个正方形网格；对所述多个正方形网格进行编码，以得到多个编码，每个编码对应于一个正方形网格，且每个编码用于指示对应的正方形网格所覆盖的区域。通过上述方式，在目标投影上构建的正方形平面上进行网格剖分，使得同一层级剖分得到的正方形网格所对应的地球表面的划分大小基本一致，有利于对网格数据大小敏感的计算。
参照图8，图8为本申请实施例提供的一种网格编码装置的结构示意，如图8所示，本申请实施例提供的网格编码装置800包括：
获取模块801，用于获取目标投影，所述目标投影为地球表面在平面上的投影。
获取模块801的具体描述可以参照步骤301以及对应的实施例的描述，这里不再赘述。
确定模块802，用于根据所述目标投影，确定正方形平面，所述目标投影所在的区域被所述正方形平面覆盖；
确定模块802的具体描述可以参照步骤302以及对应的实施例的描述，这里不再赘述。
网格剖分模块803，用于对所述正方形平面在横轴方向和纵轴方向进行多次二等分剖分，以得到多个正方形网格；
网格剖分模块803的具体描述可以参照步骤303以及对应的实施例的描述，这里不再赘述。
编码模块804，用于对所述多个正方形网格进行编码，以得到多个编码，每个编码对应于一个正方形网格，且每个编码用于指示对应的正方形网格所覆盖的区域。
编码模块804的具体描述可以参照步骤304以及对应的实施例的描述，这里不再赘述。
在一种可能的实现中，第一目标点和第二目标点为所述目标投影上任意的两个点，所述第一目标点为所述地球表面上的第一位置点的投影，所述第二目标点为所述地球表面上的第二位置点的投影，所述第一目标点和所述第二目标点之间的距离为第一距离，所述第一位置点和得所述第二位置点之间的距离为第二距离，所述第一距离和所述第二距离之间的比值在预设范围内。
在一种可能的实现中，所述地球表面包括赤道纬线以及与所述赤道纬线垂直的目标经线，所述目标投影包括投影后的赤道纬线以及投影后的目标经线，所述正方形平面的横轴方向与所述投影后的赤道纬线所在的方向一致，所述正方形平面的纵轴方向与所述投影后的目标经线所在的方向一致。
在一种可能的实现中，所述确定模块，用于根据所述目标投影在所述投影后的赤道纬线所在的方向上的目标长度，确定所述正方形平面的边长，其中，所述正方形平面的边长大于或等于所述目标长度，且所述正方形平面的边长为2的整数次幂的预设倍数；
根据所述正方形平面的边长，确定正方形平面，以便所述目标投影所在的区域被所述正方形平面覆盖。
在一种可能的实现中，所述确定模块，用于获取所述预设倍数，根据所述目标投影在所述投影后的赤道纬线所在的方向上的目标长度以及所述预设倍数，确定所述正方形平面的边长。
在一种可能的实现中，所述正方形平面的中心横轴与所述投影后的赤道纬线重叠。
在一种可能的实现中，所述网格剖分模块，用于对所述正方形平面在横轴方向和纵轴方向进行多次二等分剖分，以得到多组正方形网格，每组正方形网格包括多个正方形网格，其中，所述多次二等分剖分包括第N次二等分剖分以及第N+1次二等分剖分，且进行第N次二等分剖分后，得到M个正方形网格，在进行第N+1次二等分剖分后，得到M*4个正方形网格，且所述M*4个正方形网格为对所述M个正方形网格中的每个正方形网格在横轴方向和纵轴方向进行二等分剖分得到的；
所述编码模块，用于对所述多组正方形网格中每组正方形网格包括的多个正方形网格进行编码，以得到多组编码，每组编码包括多个编码。
在一种可能的实现中，所述目标投影为对所述地球表面进行整体投影或分带投影得到 的。
本申请实施例提供了一种网格编码装置，包括：获取模块801，用于获取目标投影，所述目标投影为地球表面在平面上的投影；确定模块802，用于根据所述目标投影，确定正方形平面，所述目标投影所在的区域被所述正方形平面覆盖；网格剖分模块803，用于对所述正方形平面在横轴方向和纵轴方向进行多次二等分剖分，以得到多个正方形网格；编码模块804，用于对所述多个正方形网格进行编码，以得到多个编码，每个编码对应于一个正方形网格，且每个编码用于指示对应的正方形网格所覆盖的区域。在目标投影上构建的正方形平面上进行网格剖分，使得同一层级剖分得到的正方形网格所对应的地球表面的划分大小基本一致，有利于对网格数据大小敏感的计算。
本申请还提供了一种非易失性计算机可读存储介质，所述非易失性计算机可读存储介质包含计算机指令，当计算机指令被计算机执行时可以实现上述实施例中的网格编码方法。
请参考图9，为本实施例提供的一种计算机系统的结构示意图。该计算机系统可以为终端设备(或者称之为智能终端)或者服务器。如图所示，该计算机系统包括通信模块810、传感器820、用户输入模块830、输出模块840、处理器850、音视频输入模块860、存储器870以及电源880。进一步的，本实施例提供的计算机系统还可以包括NPU890。
通信模块810可以包括至少一个能使该计算机系统与通信系统或其他计算机系统之间进行通信的模块。例如，通信模块810可以包括有线网络接口，广播接收模块、移动通信模块、无线因特网模块、局域通信模块和位置(或定位)信息模块等其中的一个或多个。这多种模块均在现有技术中有多种实现，本申请不一一描述。
传感器820可以感测系统的当前状态，诸如打开/闭合状态、位置、与用户是否有接触、方向、和加速/减速，并且传感器820可以生成用于控制系统的操作的感测信号。
用户输入模块830，用于接收输入的数字信息、字符信息或接触式触摸操作/非接触式手势，以及接收与系统的用户设置以及功能控制有关的信号输入等。用户输入模块830包括触控面板和/或其他输入设备。
输出模块840包括显示面板，用于显示由用户输入的信息、提供给用户的信息或系统的各种菜单界面等。可选的，可以采用液晶显示器(liquid crystal display，LCD)或有机发光二极管(organic light-emitting diode,OLED)等形式来配置显示面板。在其他一些实施例中，触控面板可覆盖显示面板上，形成触摸显示屏。另外，输出模块840还可以包括音频输出模块、告警器以及触觉模块等。
音视频输入模块860，用于输入音频信号或视频信号。音视频输入模块860可以包括摄像头和麦克风。
电源880可以在处理器850的控制下接收外部电力和内部电力，并且提供系统的各个组件的操作所需的电力。
处理器850包括一个或多个处理器，例如，处理器850可以包括一个中央处理器和一个图形处理器。中央处理器在本申请中具有多个核，属于多核处理器。这多个核可以集成在同一块芯片上，也可以各自为独立的芯片。
存储器870存储计算机程序，该计算机程序包括操作系统程序872和应用程序871等。典型的操作系统如微软公司的Windows，苹果公司的MacOS等用于台式机或笔记本的系统，又如谷歌公司开发的基于的安卓系统等用于移动终端的系统。前述实施例提供的方法可以通过软件的方式实现，可以认为是操作系统程序872的具体实现。存储器870可以是以下类型中的一种或多种：闪速(flash)存储器、硬盘类型存储器、微型多媒体卡型存储器、卡式存储器(例如SD或XD存储器)、随机存取存储器(random access memory,RAM)、静态随机存取存储器(static RAM,SRAM)、只读存储器(read only memory,ROM)、电可擦除可编程只读存储器(electrically erasable programmable readonly memory，EEPROM)、可编程只读存储器(programmable ROM,PROM)、回滚保护存储块(replay protected memory block,RPMB)、磁存储器、磁盘或光盘。在其他一些实施例中，存储器870也可以是因特网上的网络存储设备，系统可以对在因特网上的存储器870执行更新或读取等操作。
处理器850用于读取存储器870中的计算机程序，然后执行计算机程序定义的方法，例如处理器850读取操作系统程序872从而在该系统运行操作系统以及实现操作系统的各种功能，或读取一种或多种应用程序871，从而在该系统上运行应用。
存储器870还存储有除计算机程序之外的其他数据873。
NPU 890作为协处理器挂载到主处理器850上，用于执行主处理器850给它分配的任务。在本实施例中，NPU890可以被人脸识别TA的一个或多个子线程调用从而实现人脸识别中涉及的部分复杂算法。具体的，人脸识别TA的子线程在主处理器850的多个核上运行，然后主处理器850调用NPU890，NPU890实现的结果再返回给主处理器850。
以上各个模块的连接关系仅为一种示例，本申请任意实施例提供的网格编码方法也可以应用在其它连接方式的终端设备或者服务器中，例如所有模块通过总线连接。
图10是本实施例提供的一种NPU900的结构示意图。NPU900与主处理器和外部存储器相连。NPU900的核心部分为运算电路903，通过控制器904控制运算电路903提取存储器中的数据并进行数学运算。
在一些实现中，运算电路903内部包括多个处理引擎(process engine,PE)。在一些实现中，运算电路903是二维脉动阵列。运算电路903还可以是一维脉动阵列或者能够执行例如乘法和加法这样的数学运算的其它电子线路。在另一些实现中，运算电路903是通用的矩阵处理器。
举例来说，假设有输入矩阵A，权重矩阵B，输出矩阵C。运算电路903从权重存储器902中取矩阵B相应的数据，并缓存在运算电路903的每一个PE上。运算电路903从输入存储器901中取矩阵A数据与矩阵B进行矩阵运算，得到的矩阵的部分结果或最终结果，保存在累加器(accumulator)908中。
统一存储器906用于存放输入数据以及输出数据。权重数据直接通过存储单元访问控制器905(例如direct memory access controller，DMAC)被搬运到权重存储器902中。
输入数据也通过存储单元访问控制器905被搬运到统一存储器906中。
总线接口单元910(bus interface unit，BIU)用于AXI(advanced extensible interface)总线与 存储单元访问控制器905和取指存储器909(instruction fetch buffer)的交互。
总线接口单元910用于取指存储器909从外部存储器获取指令，还用于存储单元访问控制器905从外部存储器获取输入矩阵A或者权重矩阵B的原数据。
存储单元访问控制器905主要用于将外部存储器中的输入数据搬运到统一存储器906或将权重数据搬运到权重存储器902中或将输入数据数据搬运到输入存储器901中。
向量计算单元907通常包括多个运算处理单元，在需要的情况下，对运算电路903的输出做进一步处理，如向量乘、向量加、指数运算、对数运算、和/或大小比较等等。
在一些实现中，向量计算单元907能将经处理的向量存储到统一存储器906中。例如，向量计算单元907可以将非线性函数应用到运算电路903的输出，例如累加值的向量，用以生成激活值。在一些实现中，向量计算单元907生成归一化的值、合并值，或二者均有。在一些实现中，经处理的向量能够用作运算电路903的激活输入。
与控制器904连接的取指存储器909用于存储控制器904使用的指令。
统一存储器906，输入存储器901，权重存储器902以及取指存储器909均为On-Chip存储器。图中的外部存储器与该NPU硬件架构独立。
需要说明的是，本实施例提供的地址转换关系的配置方法也可以应用于非终端的计算机设备，例如云端服务器。
所属领域的技术人员可以清楚地了解到，为描述的方便和简洁，上述描述的系统，装置和单元的具体工作过程，可以参考前述方法实施例中的对应过程，在此不再赘述。
在本申请所提供的几个实施例中，应该理解到，所揭露的系统，装置和方法，可以通过其它的方式实现。例如，以上所描述的装置实施例仅仅是示意性的，例如，单元的划分，仅仅为一种逻辑功能划分，实际实现时可以有另外的划分方式，例如多个单元或组件可以结合或者可以集成到另一个系统，或一些特征可以忽略，或不执行。另一点，所显示或讨论的相互之间的耦合或直接耦合或通信连接可以是通过一些接口，装置或单元的间接耦合或通信连接，可以是电性，机械或其它的形式。
所述作为分离部件说明的单元可以是或者也可以不是物理上分开的，作为单元显示的部件可以是或者也可以不是物理单元，即可以位于一个地方，或者也可以分布到多个网络单元上。可以根据实际的需要选择其中的部分或者全部单元来实现本实施例方案的目的。
另外，在本申请各个实施例中的各功能单元可以集成在一个处理单元中，也可以是各个单元单独物理存在，也可以两个或两个以上单元集成在一个单元中。上述集成的单元既可以采用硬件的形式实现，也可以采用软件功能单元的形式实现。
所述集成的单元如果以软件功能单元的形式实现并作为独立的产品销售或使用时，可以存储在一个计算机可读取存储介质中。基于这样的理解，本申请的技术方案本质上或者说对现有技术做出贡献的部分或者该技术方案的全部或部分可以以软件产品的形式体现出来，该计算机软件产品存储在一个存储介质中，包括若干指令用以使得一台计算机设备(可以是个人计算机，服务器，或者其他网络设备等)执行本申请图3实施例所述方法的全部或部分步骤。而前述的存储介质包括：U盘、移动硬盘、只读存储器(read-only memory，ROM)、随机存取存储器(random access memory，RAM)、磁碟或者光盘等各种可以存储程序代码的 介质。
以上所述，以上实施例仅用以说明本申请的技术方案，而非对其限制；尽管参照前述实施例对本申请进行了详细的说明，本领域的普通技术人员应当理解：其依然可以对前述各实施例所记载的技术方案进行修改，或者对其中部分技术特征进行等同替换；而这些修改或者替换，并不使相应技术方案的本质脱离本申请各实施例技术方案的范围。

Claims (19)

  1. 一种网格编码方法，其特征在于，所述方法包括：
    获取目标投影，所述目标投影为地球表面在平面上的投影，且所述地球表面上任意两个相同面积的区域在所述目标投影上对应的两个投影区域之间的面积差异小于预设值；
    根据所述目标投影，确定正方形平面，所述目标投影所在的区域被所述正方形平面覆盖；
    对所述正方形平面在横轴方向和纵轴方向进行多次二等分剖分，以得到多个正方形网格；
    对所述多个正方形网格进行编码，以得到多个编码，每个编码对应于一个正方形网格，且每个编码用于指示对应的正方形网格所覆盖的区域。
  2. 根据权利要求1所述的方法，其特征在于，第一目标点和第二目标点为所述目标投影上任意的两个点，所述第一目标点为所述地球表面上的第一位置点的投影，所述第二目标点为所述地球表面上的第二位置点的投影，所述第一目标点和所述第二目标点之间的距离为第一距离，所述第一位置点和得所述第二位置点之间的距离为第二距离，所述第一距离和所述第二距离之间的比值在预设范围内。
  3. 根据权利要求1或2所述的方法，其特征在于，所述地球表面包括赤道纬线以及与所述赤道纬线垂直的目标经线，所述目标投影包括投影后的赤道纬线以及投影后的目标经线，所述正方形平面的横轴方向与所述投影后的赤道纬线所在的方向一致，所述正方形平面的纵轴方向与所述投影后的目标经线所在的方向一致。
  4. 根据权利要求3所述的方法，其特征在于，所述根据所述目标投影，确定正方形平面，包括：
    根据所述目标投影在所述投影后的赤道纬线所在的方向上的目标长度，确定所述正方形平面的边长，其中，所述正方形平面的边长大于或等于所述目标长度，且所述正方形平面的边长为2的整数次幂的预设倍数；
    根据所述正方形平面的边长，确定正方形平面，以便所述目标投影所在的区域被所述正方形平面覆盖。
  5. 根据权利要求4所述的方法，其特征在于，所述根据所述目标投影在所述投影后的赤道纬线所在的方向上的目标长度，确定所述正方形平面的边长，包括：
    获取所述预设倍数，根据所述目标投影在所述投影后的赤道纬线所在的方向上的目标长度以及所述预设倍数，确定所述正方形平面的边长。
  6. 根据权利要求3至5任一所述的方法，其特征在于，所述正方形平面的中心横轴与所述投影后的赤道纬线重叠。
  7. 根据权利要求1至6任一所述的方法，其特征在于，所述对所述正方形平面在横轴方向和纵轴方向进行多次二等分剖分，以得到多个正方形网格，包括：
    对所述正方形平面在横轴方向和纵轴方向进行多次二等分剖分，以得到多组正方形网格，每组正方形网格包括多个正方形网格，其中，所述多次二等分剖分包括第N次二等分剖分以及第N+1次二等分剖分，且进行第N次二等分剖分后，得到M个正方形网格，在进行第N+1次二等分剖分后，得到M*4个正方形网格，且所述M*4个正方形网格为对所述M个正方形网格中的每个正方形网格在横轴方向和纵轴方向进行二等分剖分得到的；
    所述对所述多个正方形网格进行编码，以得到多个编码，包括：
    对所述多组正方形网格中每组正方形网格包括的多个正方形网格进行编码，以得到多组编码，每组编码包括多个编码。
  8. 根据权利要求1至7任一所述的方法，其特征在于，所述目标投影为对所述地球表面进行整体投影或分带投影得到的。
  9. 一种网格编码装置，其特征在于，所述装置包括：
    获取模块，用于获取目标投影，所述目标投影为地球表面在平面上的投影；
    确定模块，用于根据所述目标投影，确定正方形平面，所述目标投影所在的区域被所述正方形平面覆盖；
    网格剖分模块，用于对所述正方形平面在横轴方向和纵轴方向进行多次二等分剖分，以得到多个正方形网格；
    编码模块，用于对所述多个正方形网格进行编码，以得到多个编码，每个编码对应于一个正方形网格，且每个编码用于指示对应的正方形网格所覆盖的区域。
  10. 根据权利要求9所述的装置，其特征在于，第一目标点和第二目标点为所述目标投影上任意的两个点，所述第一目标点为所述地球表面上的第一位置点的投影，所述第二目标点为所述地球表面上的第二位置点的投影，所述第一目标点和所述第二目标点之间的距离为第一距离，所述第一位置点和得所述第二位置点之间的距离为第二距离，所述第一距离和所述第二距离之间的比值在预设范围内。
  11. 根据权利要求9或10所述的装置，其特征在于，所述地球表面包括赤道纬线以及与所述赤道纬线垂直的目标经线，所述目标投影包括投影后的赤道纬线以及投影后的目标经线，所述正方形平面的横轴方向与所述投影后的赤道纬线所在的方向一致，所述正方形平面的纵轴方向与所述投影后的目标经线所在的方向一致。
  12. 根据权利要求11所述的装置，其特征在于，所述确定模块，用于根据所述目标投影在所述投影后的赤道纬线所在的方向上的目标长度，确定所述正方形平面的边长，其中， 所述正方形平面的边长大于或等于所述目标长度，且所述正方形平面的边长为2的整数次幂的预设倍数；
    根据所述正方形平面的边长，确定正方形平面，以便所述目标投影所在的区域被所述正方形平面覆盖。
  13. 根据权利要求12所述的装置，其特征在于，所述确定模块，用于获取所述预设倍数，根据所述目标投影在所述投影后的赤道纬线所在的方向上的目标长度以及所述预设倍数，确定所述正方形平面的边长。
  14. 根据权利要求12至13任一所述的装置，其特征在于，所述正方形平面的中心横轴与所述投影后的赤道纬线重叠。
  15. 根据权利要求9至14任一所述的装置，其特征在于，所述网格剖分模块，用于对所述正方形平面在横轴方向和纵轴方向进行多次二等分剖分，以得到多组正方形网格，每组正方形网格包括多个正方形网格，其中，所述多次二等分剖分包括第N次二等分剖分以及第N+1次二等分剖分，且进行第N次二等分剖分后，得到M个正方形网格，在进行第N+1次二等分剖分后，得到M*4个正方形网格，且所述M*4个正方形网格为对所述M个正方形网格中的每个正方形网格在横轴方向和纵轴方向进行二等分剖分得到的；
    所述编码模块，用于对所述多组正方形网格中每组正方形网格包括的多个正方形网格进行编码，以得到多组编码，每组编码包括多个编码。
  16. 根据权利要求9至15任一所述的装置，其特征在于，所述目标投影为对所述地球表面进行整体投影或分带投影得到的。
  17. 一种计算机系统，其特征在于，包括存储器和处理器，其中，
    所述存储器用于存储计算机可读指令；所述处理器用于读取所述计算机可读指令并实现如权利要求1-8任意一项所述的方法。
  18. 一种计算机存储介质，其特征在于，存储有计算机可读指令，且所述计算机可读指令在被处理器执行时实现如权利要求1-8任意一项所述的方法。
  19. 一种计算机程序产品，其特征在于，包括代码，当所述代码被执行时，用于实现如权利要求1至8任一所述的方法。
PCT/CN2021/133339 2020-11-26 2021-11-26 一种网格编码方法及计算机系统 WO2022111609A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202011352980.3 2020-11-26
CN202011352980.3A CN114549671A (zh) 2020-11-26 2020-11-26 一种网格编码方法及计算机系统

Publications (1)

Publication Number Publication Date
WO2022111609A1 true WO2022111609A1 (zh) 2022-06-02

Family

ID=81668181

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2021/133339 WO2022111609A1 (zh) 2020-11-26 2021-11-26 一种网格编码方法及计算机系统

Country Status (2)

Country Link
CN (1) CN114549671A (zh)
WO (1) WO2022111609A1 (zh)

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115601466A (zh) * 2022-10-20 2023-01-13 钰深（北京）科技有限公司（Cn） 服装裁片数字化二维剖分方法、电子设备以及存储介质
CN116740303A (zh) * 2023-06-07 2023-09-12 中国人民解放军91977部队 水下目标的离散网格生成方法
CN118467666A (zh) * 2024-07-15 2024-08-09 国家基础地理信息中心 面状矢量数据网格化方法、装置及设备

Families Citing this family (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115801024B (zh) * 2022-11-24 2023-09-12 西南林业大学 局部等距性优化球面格网的编码方法、系统、装置及介质

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20070251161A1 (en) * 2004-12-21 2007-11-01 Florian Tuczek Double-curved shell
CN101606177A (zh) * 2007-01-04 2009-12-16 鸣川肇 信息处理方法
US20120110865A1 (en) * 2010-11-08 2012-05-10 Vaught Gregory V Chart specific navigation plotter and method for inexpensive production thereof
CN106780317A (zh) * 2017-01-25 2017-05-31 北京数码视讯科技股份有限公司 全景投影方法及装置

Family Cites Families (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2000181345A (ja) * 1998-12-17 2000-06-30 Toru Nishioka 世界メッシュコ―ドの作成方法、および世界メッシュコ―ド
GB2563944B (en) * 2017-06-30 2021-11-03 Canon Kk 360-Degree video encoding with block-based extension of the boundary of projected parts
CN107665242A (zh) * 2017-09-08 2018-02-06 南京王师大数据有限公司 一种区域空间多尺度网格编码方法及装置

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20070251161A1 (en) * 2004-12-21 2007-11-01 Florian Tuczek Double-curved shell
CN101606177A (zh) * 2007-01-04 2009-12-16 鸣川肇 信息处理方法
US20120110865A1 (en) * 2010-11-08 2012-05-10 Vaught Gregory V Chart specific navigation plotter and method for inexpensive production thereof
CN106780317A (zh) * 2017-01-25 2017-05-31 北京数码视讯科技股份有限公司 全景投影方法及装置

Cited By (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115601466A (zh) * 2022-10-20 2023-01-13 钰深（北京）科技有限公司（Cn） 服装裁片数字化二维剖分方法、电子设备以及存储介质
CN115601466B (zh) * 2022-10-20 2023-06-16 钰深(北京)科技有限公司 服装裁片数字化二维剖分方法、电子设备以及存储介质
CN116740303A (zh) * 2023-06-07 2023-09-12 中国人民解放军91977部队 水下目标的离散网格生成方法
CN116740303B (zh) * 2023-06-07 2023-11-07 中国人民解放军91977部队 水下目标的离散网格生成方法
CN118467666A (zh) * 2024-07-15 2024-08-09 国家基础地理信息中心 面状矢量数据网格化方法、装置及设备

Also Published As

Publication number Publication date
CN114549671A (zh) 2022-05-27

Similar Documents

Publication Publication Date Title
WO2022111609A1 (zh) 一种网格编码方法及计算机系统
JP7560021B2 (ja) 深層学習システム
US20210295169A1 (en) Deep learning based identification of difficult to test nodes
CN110392902B (zh) 使用稀疏体积数据的操作
TWI712004B (zh) 擴增實境設備的座標系校準方法及裝置
CN112990010B (zh) 点云数据处理方法、装置、计算机设备和存储介质
US10192353B1 (en) Multiresolution surface representation and compression
US8954295B1 (en) Determining an outer shell of a composite three-dimensional model
CN113936091A (zh) 用于构建射线跟踪加速结构的方法和系统
CN111414915B (zh) 一种文字识别方法以及相关设备
WO2021027692A1 (zh) 视觉特征库的构建方法、视觉定位方法、装置和存储介质
CN111127590A (zh) 一种二阶贝塞尔曲线绘制方法及装置
CN114187589A (zh) 一种目标检测方法、装置、设备和存储介质
CN110009625B (zh) 基于深度学习的图像处理系统、方法、终端、及介质
CN115984440B (zh) 对象渲染方法、装置、计算机设备和存储介质
US20230418705A1 (en) Hardware-efficient pam-3 encoder and decoder
CN113240787A (zh) 阴影渲染方法、装置及电子设备
US20230351696A1 (en) Data processing method and apparatus, device, computer-readable storage medium, and computer program product
US20230229916A1 (en) Scalable tensor network contraction using reinforcement learning
US10861174B2 (en) Selective 3D registration
Zhang et al. Pose detection of aerial image object based on constrained neural network
CN112541535B (zh) 基于互补的多分支深度学习进行三维点云分类方法
CN114140512A (zh) 一种图像处理方法以及相关设备
CN118189934B (zh) 地图更新方法、装置、计算机设备和存储介质
US12132590B2 (en) Hardware-efficient PAM-3 encoder and decoder

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 21897109

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 21897109

Country of ref document: EP

Kind code of ref document: A1