Systems and methods for locally aligning map data
원문보기
IPC분류정보
국가/구분
United States(US) Patent
등록
국제특허분류(IPC7판)
G06T-007/33
G05D-001/02
G01C-021/34
G06T-007/11
G06T-007/32
G01C-021/32
출원번호
US-0427263
(2017-02-08)
등록번호
US-10217232
(2019-02-26)
발명자
/ 주소
Harada, Masahiro
출원인 / 주소
Toyota Motor Engineering & Manufacturing North America, Inc.
대리인 / 주소
Darrow, Christopher G.
인용정보
피인용 횟수 :
0인용 특허 :
8
초록▼
System, methods, and other embodiments described herein relate to locally aligning features within a map. In one embodiment, a method includes segmenting map data that forms the map using a first grid to divide the map into a first set of tiles and a second grid to divide the map into a second set o
System, methods, and other embodiments described herein relate to locally aligning features within a map. In one embodiment, a method includes segmenting map data that forms the map using a first grid to divide the map into a first set of tiles and a second grid to divide the map into a second set of tiles that are offset and overlap the first set of tiles. The method includes analyzing neighbor tiles in relation to a patch tile based, at least in part, on shared features between the patch tile and the neighbor tiles to identify external misalignments between one or more of the neighbor tiles and the patch tile. The neighbor tiles are from a different one of the first grid and the second grid. The method includes adjusting alignment of the neighbor tiles within the map according to the external misalignments for the neighbor tiles.
대표청구항▼
1. A map alignment system for locally aligning map data of a map, comprising: one or more processors; a memory communicably coupled to the one or more processors and storing:a segmenting module including instructions that when executed by the one or more processors cause the one or more processors t
1. A map alignment system for locally aligning map data of a map, comprising: one or more processors; a memory communicably coupled to the one or more processors and storing:a segmenting module including instructions that when executed by the one or more processors cause the one or more processors to segment the map data that forms the map using a first grid to divide the map into a first set of tiles and a second grid to divide the map into a second set of tiles that are offset and overlap the first set of tiles; andan alignment module including instructions that when executed by the one or more processors cause the one or more processors to analyze neighbor tiles in relation to a patch tile based, at least in part, on shared features between the patch tile and the neighbor tiles to identify external misalignments between one or more of the neighbor tiles and the patch tile,wherein the neighbor tiles are from a different one of the first grid and the second grid in relation to the patch tile and overlap with the patch tile,wherein the alignment module includes instructions to adjust alignment of the neighbor tiles within the map according to the misalignments for the one or more of the neighbor tiles; andcontrolling the vehicle to navigate according to at least a current title and overlapping tiles from the first set and the second set, wherein the current tile is associated with a current position of the vehicle and the overlapping tiles overlap at least a portion of the current tile. 2. The map alignment system of claim 1, wherein the alignment module further includes instructions to, prior to analyzing the neighbor tiles, identify internal misalignments for tiles of the first set and of the second set by individually analyzing the tiles of the first set and the second set according to tile data that separately comprises the tiles, wherein the alignment module includes instructions to individually adjust the tiles of the first set and the second set by correcting the internal misalignments to locally align separate data elements within the tiles of the first set and the second set, andwherein the alignment module includes instructions to load the neighbor tiles and the patch tile into a memory upon selecting the patch from the map. 3. The map alignment system of claim 2, wherein the alignment module includes instructions to analyze the neighbor tiles in relation to the patch tile by correlating the shared features to identify the external misalignments for one or more of the neighbor tiles, and wherein the alignment module includes instructions to identify the internal misalignments by applying a simultaneous localization and mapping (SLAM) function to the tiles of the first set and the second set. 4. The map alignment system of claim 1, wherein the segmenting module further includes instructions to segment the map data into the first grid by generating the first grid using horizontal and vertical dividing lines to define the first set of tiles using a first grid structure, and wherein the segmenting module further includes instructions to segment the map data into the second grid by generating the second grid to be offset from the first grid to partially overlap tiles of the first set using a second grid structure. 5. The map alignment system of claim 1, wherein the alignment module further includes instructions to adjust the alignment by correcting the external misalignments between the patch tile and the neighbor tiles using locally identified errors between the patch tile and the neighbor tiles to reposition the neighbor tiles, and wherein the alignment module further includes instructions to adjust the alignment by modifying coordinates of the neighbor tiles to change respective locations within the map according to coordinate systems defined by the first grid and the second grid. 6. The map alignment system of claim 1, wherein the segmenting module further includes instructions to segment the map data in response to electronically retrieving the map data from a data store, wherein the map data is comprised of topological map information, and wherein the shared features are identifying aspects of the patch tile and the neighbor tiles. 7. The map alignment system of claim 1, wherein the mapping module including instructions to load, into the memory of a vehicle, the current tile from one of the first set and the second set of tiles upon determining that the current tile is a closest one of the tiles of the first set and the second set to a current position of the vehicle, and wherein the mapping module further includes instructions to load, into the memory, the overlapping tiles from a different one of the first set and the second set in relation to the current tile. 8. A non-transitory computer-readable medium storing instructions that when executed by one or more processors cause the one or more processors to: segment map data that forms a map using a first grid to divide the map into a first set of tiles and a second grid to divide the map into a second set of tiles that are offset and overlap the first set of tiles;analyze neighbor tiles in relation to a patch tile based, at least in part, on shared features between the patch tile and the neighbor tiles to identify external misalignments between one or more of the neighbor tiles and the patch tile,wherein the neighbor tiles are from a different one of the first grid and the second grid in relation to the patch tile and overlap with the patch tile;adjust alignment of the neighbor tiles within the map according to the misalignments for the one or more of the neighbor tiles; andcontrol the vehicle to navigate according to at least a current tile and overlapping tiles from the first set and the second set, wherein the current tile is associated with a current position of the vehicle and the overlapping tiles overlap at least a position of the current tile. 9. The non-transitory computer-readable medium of claim 8, wherein the instructions to segment the map data include instructions to electronically retrieve the map data from a data store, wherein the map data is comprised of topological map information, wherein the shared features are identifying aspects of the patch tile and the neighbor tiles, and wherein the instructions to segment the map data into the first grid include instructions to generate the first grid using horizontal and vertical dividing lines to define the first set of tiles, and wherein the instructions to segment the map data include instructions to generate the second grid to be offset from the first grid to partially overlap tiles of the first set using a second grid structure. 10. The non-transitory computer-readable medium of claim 8, wherein the instructions to adjust the alignment include instructions to correct the external misalignments between the patch tile and the neighbor tiles using locally identified errors between the patch tile and the neighbor tiles to reposition the neighbor tiles, and wherein the instructions to adjust the alignment include instructions to modify coordinates of the neighbor tiles to change respective locations within the map according to coordinate systems defined by the first grid and the second grid. 11. The non-transitory computer-readable medium of claim 8, wherein the instructions further include instructions to, prior to analyzing the neighbor tiles, identify internal misalignments for tiles of the first set and of the second set by individually analyzing the tiles of the first set and the second set according to data that separately comprises the tiles, wherein the instructions further include instructions to individually adjust the tiles of the first set and the second set by correcting the internal misalignments to locally align separate data elements within the tiles of the first set and the second set. 12. The non-transitory computer-readable medium of claim 11, wherein the instructions to analyze the neighbor tiles include instructions to correlate the shared features to identify the external misalignments for the one or more of the neighbor tiles, and wherein the instructions to identify the internal misalignments include instructions to apply a simultaneous localization and mapping (SLAM) function to the tiles of the first set and the second set. 13. A method of locally aligning features within a map, comprising: segmenting map data that forms the map using a first grid to divide the map into a first set of tiles and a second grid to divide the map into a second set of tiles that are offset and overlap the first set of tiles;analyzing neighbor tiles in relation to a patch tile based, at least in part, on shared features between the patch tile and the neighbor tiles to identify external misalignments between one or more of the neighbor tiles and the patch tile,wherein the neighbor tiles are from a different one of the first grid and the second grid in relation to the patch tile and overlap with the patch tile;adjusting alignment of the neighbor tiles within the map according to the external misalignments for the neighbor tiles; andcontrolling the vehicle to navigate according to at least a current tile and overlapping tiles from the first set and the second set, wherein the current tile is associated with a current position of the vehicle and the overlapping tiles overlap at least a portion of the current tile. 14. The method of claim 13, further comprising: prior to analyzing the neighbor tiles, identifying internal misalignments for tiles of the first set and of the second set by analyzing the tiles of the first set and the second set according to tile data that separately comprises the tiles;adjusting the tiles of the first set and the second set by modifying the tiles to correct the internal misalignments and to locally align separate data elements within the tiles; andloading the neighbor tiles and the patch tile into a memory upon selecting the patch from the map. 15. The method of claim 14, wherein analyzing the neighbor tiles in relation to the patch tile includes correlating the shared features to identify the external misalignments for the one or more of the neighbor tiles, and wherein identifying the internal misalignments includes applying a simultaneous localization and mapping (SLAM) function to the tiles of the first set and the second set. 16. The method of claim 13, wherein segmenting the map data into the first grid includes generating the first grid using horizontal and vertical dividing lines to define the first set of tiles using a first grid structure, and wherein segmenting the map data into the second grid includes generating the second grid to be offset from the first grid to partially overlap tiles of the first set using a second grid structure. 17. The method of claim 13, wherein adjusting the alignment includes correcting the external misalignments between the patch tile and the neighbor tiles using locally identified errors between the patch tile and the neighbor tiles to reposition the neighbor tiles, and wherein adjusting the alignment includes modifying coordinates of the neighbor tiles to change respective locations within the map according to coordinate systems defined by the first grid and the second grid. 18. The method of claim 13, wherein segmenting the map data occurs in response to electronically retrieving the map data from a data store, wherein the map data is comprised of topological map information, and wherein the shared features are identifying aspects of the patch tile and the neighbor tiles. 19. The method of claim 13, further comprising: loading, into a memory of a vehicle, a current tile from one of the first set and the second set of tiles upon determining that the current tile is a closest one of the tiles of the first set and the second set to a current position of the vehicle;loading, into the memory, overlapping tiles from a different one of the first set and the second set in relation to the current tile, wherein the overlapping tiles ovlerap at least a portion of the current tile; andcontrolling the vehicle to navigate according to at least the current tile and the overlapping tiles.
연구과제 타임라인
LOADING...
LOADING...
LOADING...
LOADING...
LOADING...
이 특허에 인용된 특허 (8)
Padwick, Christopher G.; Wallerius, John W.; Smith, II, James T., Automated geospatial image mosaic generation with automatic source selection.
※ AI-Helper는 부적절한 답변을 할 수 있습니다.