CN109827585B - Method for rapidly filling concave area in grid map - Google Patents
Method for rapidly filling concave area in grid map Download PDFInfo
- Publication number
- CN109827585B CN109827585B CN201910126765.2A CN201910126765A CN109827585B CN 109827585 B CN109827585 B CN 109827585B CN 201910126765 A CN201910126765 A CN 201910126765A CN 109827585 B CN109827585 B CN 109827585B
- Authority
- CN
- China
- Prior art keywords
- grid
- block
- filling
- map
- layer
- Prior art date
- Legal status (The legal status 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 status listed.)
- Active
Links
Images
Landscapes
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention belongs to the field of robot path planning, and particularly relates to a method for quickly filling a concave area in a grid map, which comprises the following steps: acquiring an occupation grid map, selecting a gravity action direction which is usually selected to be four directions of up, down, left and right, and selecting the gravity direction to be downward in the following steps; traversing the map to find a grid unit with supporting force, namely an obstacle grid unit is arranged right below the grid unit, and the grid unit is not an obstacle; and starting to expand in the horizontal direction from the left side to the right side of the grid unit until the whole horizontal layer is filled, namely, the horizontal layer is expanded until the left side and the right side are both provided with obstacles, continuously checking whether the obstacles exist above a certain grid or the obstacles do not exist below the certain grid in the process, and returning to the previous step if the obstacles exist. The invention is independent of the convex hull problem in computer graphics, and is a method which is rapid, intuitive and convenient to apply.
Description
Technical Field
The invention belongs to the field of robot path planning, and particularly relates to a method for quickly filling a concave area in a grid map.
Background
The path planning is one of core problems of autonomous robot research, and aims to plan a safe path from a starting point to a target point through the perception of the robot on the surrounding environment or through the existing map information. The target range of the robot path planning can be divided into global path planning and local path planning. Where global path planning typically requires that the initial pose, environmental information, and targets of the robot be known. The map is used as a basic element of path planning, and the running time and the planning quality of the global planning algorithm are directly influenced. Common map representation methods in robotics are grid maps, topological maps, and semantic maps, where grid maps are easy to build, maintain, and can provide accurate metric maps. In many practical scenarios, such as SLAM, games, and mars expedition divers, grid maps are the most common and basic form of maps. The grid map, however, contains redundant information in the metric map for accurate mapping and thus cannot process data efficiently. This also results in that path planning on a grid map tends to take a lot of time.
One particular case in path planning is an invalid search for a concave region. The concave area refers to an area formed by irregular concave polygons such as rooms, office compartments or corners in a map, and if one such area exists on a path between a starting point and an ending point, a large amount of invalid search time and memory occupation of a path planning algorithm are caused, and even navigation fails. It is therefore necessary to mark this type of area in the grid map in advance by a preprocessing means.
Disclosure of Invention
The invention provides a map preprocessing method for quickly eliminating a concave obstacle area, aiming at the problems that a concave area in a grid map can cause invalid search time consumption and memory occupation of a path planning algorithm. The method is inspired by physical phenomena, is not limited by the specific shape of the barrier, and has the characteristics of intuition, rapidness and convenient implementation.
A method for rapidly filling concave areas in a grid map comprises the following steps:
(1) acquiring an occupation grid map, selecting a gravity action direction which is usually selected to be four directions of up, down, left and right, and selecting the gravity direction to be downward in the following steps;
(2) traversing the map to find a grid unit with supporting force, namely an obstacle grid unit is arranged right below the grid unit, and the grid unit is not an obstacle;
(3) starting to expand in the horizontal direction from the left side to the right side of the grid unit until the whole horizontal layer is filled, namely, expanding until the left side and the right side are both provided with obstacles, continuously checking whether the situation that an obstacle exists above a certain grid or no obstacle exists below the certain grid in the process, and if so, returning to the step (2);
(4) if the horizontal layer is the first layer of the current filling, marking the filling layer as a new filling area; otherwise, marking the grid of the layer as a filling area which is the same as the grid of the next layer, rising the horizontal plane by one layer, and returning to the step (3);
(5) establishing association for adjacent filling blocks, and using a block association table to represent the association structure;
(6) building an exclusion list L exc 。
The association is created for the adjacent filling blocks, and the block association table is used for representing the association structure, and the method comprises the following steps:
a filling block b n Is marked with a category number of T b (b n ) The following is the construction rule of the block association table:
(5.1) applying an array space as an index directory of the block association table, wherein data items in the array are head nodes of a single linked list, which is abbreviated as B (n), wherein n is an index sequence number, and the data items of the single linked list are category sequence numbers of the filling blocks;
(5.2) Each filling Block b in the Table n Occupying index unit T b (b n ) As the program runs, more blocks are added into the index table and the array space needs to be reallocated;
(5.3) if one is building block b new Is a block b adjacent to the previously established block old And constructed, then T b (b new ) Will become B (T) b (b old ) New node under the head node.
Said establishing an exclusion list L exc The method comprises the following steps:
(1) establishing a null exclusion list L exc ；
(2) Filling block b with starting point or target point start And b end T of b (b start ) And T b (b end ) Adding into L exc Querying the corresponding head node B (T) according to the block association table b (b start ) And B (T) b (b end ) And all their successor nodes are also added to L) exc ；
(3) All the back connection nodes are used as head nodes, the block association table is inquired, and the back connection nodes are added into the exclusion list L exc This process is repeated until the last added node has no successor nodes.
The invention has the beneficial effects that:
(1) the phenomenon of interaction force of the container and fluid flowing into the container is abstracted, and a method for filling a concave obstacle area in a grid map is obtained. The FC algorithm in the invention is insensitive to the barrier edge, and filling with a better result can be performed once as long as the concave barrier exists in the map.
(2) The invention is independent of the convex hull problem in computer graphics, and is a method which is rapid, intuitive and convenient to apply.
(3) If there is no starting point and no target point in a concave obstacle, the grid in the concave obstacle does not appear on the optimal path. Filling concave obstacles occupying the grid map can greatly reduce invalid search time consumption and memory occupation of the path planning algorithm.
(4) The building block association table and the exclusion list can enable repeated path planning on the same map without repeatedly using an FC algorithm.
Drawings
FIG. 1 is a block association table structure diagram;
FIG. 2 is a schematic diagram of a point on a map having a supporting force;
FIG. 3 is a schematic diagram of horizontal layer expansion;
FIG. 4 is a schematic view of the horizontal layer lacking support force, terminating expansion;
FIG. 5 is a schematic diagram of a process for completing filling of a recessed area;
Detailed Description
The invention is further described below with reference to the accompanying drawings.
The invention relates to the field of robot path planning, in particular to a method for quickly filling a concave area in a grid map.
The path planning is one of the core problems of autonomous robot research, and aims to plan a safe path from a starting point to a target point through the perception of the robot on the surrounding environment or through the existing map information. The target range of the robot path planning can be divided into global path planning and local path planning. Where global path planning typically requires that the initial pose, environmental information, and targets of the robot be known. The map is used as a basic element of path planning, and the running time and the planning quality of the global planning algorithm are directly influenced. Common map representation methods in robotics are grid maps, topological maps, and semantic maps, where grid maps are easy to build, maintain, and can provide accurate metric maps. In many practical scenarios, such as SLAM, games, and mars expedition divers, grid maps are the most common and basic form of maps. The grid map, however, contains redundant information in the metric map for accurate mapping and thus cannot process data efficiently. This also results in that path planning on a grid map tends to take a lot of time.
One particular case in path planning is an invalid search for a concave region. The concave area refers to an area formed by irregular concave polygons such as rooms, office compartments or corners in a map, and if one such area exists on a path between a starting point and an ending point, a large amount of invalid search time and memory occupation of a path planning algorithm are caused, and even navigation fails. It is therefore necessary to mark this type of area in the grid map in advance by a preprocessing means.
The invention provides a map preprocessing method for quickly eliminating a concave obstacle area, aiming at the problems that a concave area in a grid map can cause invalid search time consumption and memory occupation of a path planning algorithm. The method is inspired by physical phenomena, is not limited by the specific shape of the barrier, and has the characteristics of intuition, rapidness and convenient implementation.
In order to achieve the above object, the present invention provides the following solutions:
an FC (fill The container) algorithm that quickly fills in a concave area. The algorithm is not an optimal scheme, can not ensure complete concave polygon completion, but is insensitive to the shape of a concave obstacle, and can achieve better effect in most cases. Abstracting the phenomenon of interaction force of the container and fluid flowing into the container to obtain a step of filling a concave obstacle area in the grid map:
step 1: an occupancy grid map is acquired, a gravity action direction is selected, and the direction can be selected to be four directions, namely, an upper direction, a lower direction, a left direction and a right direction, and the following steps are based on the gravity direction being downward.
Step 2: the map is traversed to find a grid cell that possesses the supporting force, i.e. a grid cell with an obstacle directly below it, and which is not an obstacle itself.
And step 3: starting to expand the grid unit in the horizontal direction from the left side to the right side until the whole horizontal layer is filled, namely, expanding the grid unit until the left side and the right side are both provided with obstacles, continuously checking whether an obstacle exists above a certain grid or no obstacle exists below the certain grid in the process, and if so, jumping back to the step 2.
And 4, step 4: if the horizontal layer is the first layer of the current padding, marking the padding layer as a new padding area, otherwise marking the grid of the layer as the same padding area as the grid of the next layer. And (5) rising the horizontal plane by one layer, and continuing to the step 3.
There are often a number of recessed areas in a real-world map, and each recessed area is filled with one or more filling blocks. The filling blocks provide supporting force for each other during the construction of the FC algorithm, which results in that only one filling block is removed in the path planning process, which is not enough to restore the path, and therefore, an association needs to be created for adjacent filling blocks, and a block association table is used for representing the association structure. A filling block b n Is marked with a category number of T b (b n ). The following is the construction rules for the block association table:
rule 1: an array space is applied as an index directory of the block association table, and data items in the array are head nodes of a single linked list, which are abbreviated as B (n), wherein n is an index sequence number. Wherein the data items of the singly linked list are the category sequence numbers of the padding blocks.
Rule 2: each filling block b in the table n Occupying index unit T b (b n ) As the program runs, more blocks are added to the index table and array space may need to be reallocated.
Rule 3: if one is building block b new Is a block b adjacent to the previously established block old And constructed, then T b (b new ) Will become B (T) b (b old ) New node under the head node.
Performing path planning on a map with padding blocks already established requires some extra processing since a randomly set starting or target point is likely to be present in a padding block. The following steps establish an exclusion list L for the padding blocks of the blocked path exc The process of (2):
step 1: establishing a null exclusion list L exc 。
Step 2: filling block b in which the starting point or the target point is located start And b end T of b (b start ) And T b (b end ) Adding into L exc From block-to-block association table look-upCorresponding head node B (T) b (b start ) And B (T) b (b end ) And all their successor nodes are also added to L) exc 。
And step 3: all the back connection nodes are used as head nodes, the block association table is inquired, and the back connection nodes are added into the exclusion list L exc This process is repeated until the last added node has no successor nodes.
The exclusion list L is not required to be excluded when the path planning algorithm is running exc The padded blocks represented by the middle class number act as obstacles, while the other non-acquired blocks still act as obstacles.
The present invention will be described in detail below with reference to the accompanying drawings, and it should be noted that the described embodiments are only intended to illustrate the present invention and do not constitute a limitation of the present invention.
The cell relationship in the two-dimensional grid map space is set to be represented by a rectangular plane coordinate system, and the position of each grid cell is identified by an abscissa x and an ordinate y. Further, each cell has a value to represent its type of category, which is different from the conventional grid map, for example, in a binary grid map, 1 represents an obstacle and 0 represents a free area; or in a probability grid map, the number of a grid represents the occupation probability of that grid. For each cell occupying the grid map, its category is defined as follows:
(1) if a trellis is unoccupied, its category type is 0, or record T (x, y) ═ 0;
(2) if a trellis is occupied, its type is 1, or T (x, y) is 1;
(3) if a cell is in the padding area of the FC algorithm, its category is greater than or equal to 2, and the specific value is related to the processing order of the algorithm itself, and usually, cells in the same padding area have the same category.
There are often several recessed areas in a real-world map, and usually, each recessed area is filled with one or more filling blocks. As in fig. 5, the gray area is a patch. These padding blocks, or blocks for short, are at FC is created at the same stage of the algorithm run. Each block is composed of several grid cells of the same type, each block can be identified using the class number, and one padding block b n Is marked as T b (b n ). As the algorithm runs, more blocks are created and their class numbers increase progressively from 2.
To facilitate subsequent processing, the padding blocks are not completely independent of each other, and therefore, associations must be created for adjacent blocks. This does not require a quadtree structure like in a multi-resolution grid map, and an index table, called a block association table, is usually sufficient to handle this association. Each shim block in the table occupies an index cell, and as the program runs, more blocks are added to the index table. Each index unit is a head node of a single linked list, which is abbreviated as b (type), that is, a head node of a type number of the index table. If one is building block b new Is a block b adjacent to the previously established block old And constructed, then T b (b new ) Will become B (T) b (b old ) New node under the head node. This structure is similar to a hash table, except that there is no hash function mapping. The structure of the table can be seen in fig. 1.
The following process is a construction process of the block association table shown in fig. 1:
step 1: an array space is applied as an index directory of the block association table, and data items in the array are head nodes of a single linked list. Where index entry No. 0 and index entry No. 1 will not have a successor linked list structure because the trellis cells labeled type 0 and type 1 are not padding blocks.
Step 2: as the FC algorithm runs, the padding block b 0 Is created, and T b (b 0 ) When the index number is 2, an index 2 unit is added in the index directory, and a default padding block b is adopted n Index number of T b (b n ). Then filling in block b 1 Is created, and T b (b 1 ) (ii) when b is equal to 3 0 And b 1 No edge contact occurs and thus will not be the successor nodes to each other.
And step 3: filling block b 2 Is created and at b 2 During the construction of (a) and (b) 0 Contact occurs. Because in the supporting force determination process of the FC algorithm, the native obstacles, namely the grid cells with type ≧ 1, and all the grids with type ≧ 1 can provide supporting force for the current determination grid, this causes a situation where the constructed shim block comes into contact with the newly constructed shim block. At this point, a new singly linked list node is created with its data items set to T b (b 2 ) 4, and sets the node as a successor node of B (2). Filling block b 2 Is not in contact with other blocks and is therefore only present in the successor nodes of B (2).
And 4, step 4: filling block b 3 Is created and at b 3 During the construction of (a) and (b) 0 And b 1 Contact occurs. Constructing two singly linked list nodes with data items set to T b (b 3 ) They are set as successor nodes of B (2) and B (3), respectively, 5.
And 5: and continuously adding the table entries and creating the linked list nodes until the FC algorithm is completely operated.
The specific action process of the FC algorithm comprises the following steps:
step 1: and acquiring an occupation grid map M, and generating an empty block association table B. The current class number type is set to 2.
Step 2: one gravity direction is selected, and the gravity direction can be selected to be four directions, namely, up, down, left and right directions, or four diagonal directions. The following steps are all based on the gravity direction being downward.
And step 3: traversing all grid cells except the boundary grid one by one, and setting a block extension identifier exp b And juxtaposition is true. And recording the extension level by using the ordinate y of the current unit.
And 4, step 4: if the block extension identifier exp b This means that a new shim block can be constructed, or one or more horizontal layer expansion shims have been completed, and the previous round of shim block construction can continue. Each trellis (x, y) is attempted to be expanded to build a padding block. Checking whether the lower grid is an obstacle, and checkingWhether the grid cell is an obstacle. If T (x, y-1) ≠ 1 and T (x, y) ≠ 1, as in the case of fig. 2, then step 5 is entered, otherwise the padding procedure is terminated, returning to step 3.
And 5: the left end flag l and the right end flag r of the horizontal layer expansion are recorded by using the abscissa x of the current cell, and the horizontal direction expansion is attempted, and the result of one horizontal layer expansion is shown in fig. 3.
Step 6: a left-hand expansion is attempted. If the left end is not an obstacle, i.e., T (l-1, level) ═ 0, and the block extension identifier exp b If true, the horizontal shim layer is expanded one frame to the left, i.e., l is set to l-1. Then checking whether the upper and lower layers of the left end cell are obstacles, as shown in fig. 4, the lower layer obstacles or the previously constructed shim cells will provide supporting force for the current grid cell to prevent the current grid cell from being overlapped; inspection of upper level obstacles can avoid overfilling overwhelming the traversable path. If T (l, level-1) ≠ 0 or T (l, level +1) ≠ 0 is satisfied, the current left-direction extension is terminated, and the block extension identifier exp is set b Is false. This step is repeated until the extension to the left end is an obstacle or block extension identifier exp b False or until the left boundary of the map is reached.
And 7: a right-hand expansion is attempted. If the right end is not an obstacle, i.e., T (r +1, level) ═ 0, and the block extension identifier exp b If true, the horizontal shim layer is expanded to the right by one, i.e., r is set to r + 1. Then, whether the upper layer and the lower layer of the right-end grid are obstacles is checked, if T (r, level-1) ≠ 0 or T (l, level +1) ≠ 0 is satisfied, the current right-direction expansion is terminated, and a block expansion identifier exp is set b Is false. This step is repeated until the extension to the right is an obstacle or block extension identifier exp b False, or until the right boundary of the map is reached.
And 8: if the block extension identifier exp b False, and level ≠ y, then sets the current class number type to type + 1. Since meeting the above two conditions means that a new padding block has been constructed and the last padding block b n T of b (b n ) Type. If the block is not satisfiedExtended identifier exp b If false, go to step 9.
And step 9: if the block extension identifier exp is satisfied b This means that the concave obstacle is already being filled during the horizontal layer expansion in step 5 as well as in step 6. Whether the lowest layer of a padding block or other layers, all of the bins will share the same type of class number. For all i ∈ (l, r +1), type is assigned to T (i, level).
Step 10: it is checked whether this horizontal layer extension is in contact with other already built shim blocks. For all i ∈ (l, r +1), if T (i, level) ≠ 1 and
a new node is created under B (T (i, level-1)) and type is taken as its data item. Further, it is checked whether the left and right ends of the horizontal extension layer are in contact with other blocks. If B (T (l-1, level)) ≠ 1 or B (T (r +1, level)) ≠ 1, then a new node is created under B (T (l-1, level)) or B (T (r +1, level)) with type as its data item.
Step 11: at this time, the horizontal layer expansion is finished once, level is increased by 1, that is, the horizontal plane rises by one layer, the step 4 is returned, and the filling of the concave area is continued until one filling block is completed, as shown in fig. 5.
Performing path planning on a map with padding blocks already established requires additional processing because randomly set starting points or target points are likely to exist in a padding block, which is usually simple and usually does not require secondary processing of the map. The following steps establish an exclusion list L for the padding blocks of the blocked path exc The process of (2):
step 1: establishing a null exclusion list L exc 。
Step 2: filling block b with starting point or target point start And b end T of b (b start ) And T b (b end ) Adding into L exc Querying the corresponding head node B (T) according to the block-to-block association table b (b start ) And B (T) b (b end ) And all their successor nodes are also added to L) exc 。
And step 3: all the back connection nodes are used as head nodes, the block association table is inquired, and the back connection nodes are added into the exclusion list L exc This process is repeated until the last added node has no successor nodes.
This procedure does not result in overly complex recursive calls because the class number of the head node is always smaller than the class number of the successor node, because the padding block represented by the head node is established earlier. The time loss of this process is low, since the size of the block association table is typically small relative to the map. The exclusion list L is not required to be excluded when the path planning algorithm is running exc The padded blocks represented by the middle class number act as obstacles, while the other non-acquired blocks still act as obstacles.
Claims (1)
1. A method for rapidly filling a concave area in a grid map is characterized by comprising the following steps:
(1) acquiring an occupation grid map, selecting a gravity action direction, and selecting a gravity direction to be downward;
(2) traversing the map to find a grid unit with supporting force, namely an obstacle grid unit is arranged right below the grid unit, and the grid unit is not an obstacle;
(3) starting to expand in the horizontal direction from the left side to the right side of the grid unit until the whole horizontal layer is filled, namely, expanding until the left side and the right side are both provided with obstacles, continuously checking whether the obstacle exists above a certain grid or the obstacle does not exist below the certain grid in the process, and if so, returning to the step (2);
(4) if the horizontal layer is the first layer of the current filling, marking the filling layer as a new filling area; otherwise, marking the grid of the layer as a filling area which is the same as the grid of the next layer, rising the horizontal plane by one layer, and returning to the step (3);
(5) establishing association for adjacent filling blocks, and using a block association table to represent the association structure;
a filling block b n Is marked with a category number of T b (b n ) The following is the construction rule of the block association table:
(5.1) applying an array space as an index directory of the block association table, wherein data items in the array are head nodes of a single linked list, which is abbreviated as B (n), wherein n is an index sequence number, and the data items of the single linked list are category sequence numbers of the filling blocks;
(5.2) Each filling Block b in the Table n Occupying index unit T b (b n ) As the program runs, more blocks are added into the index table and the array space needs to be reallocated;
(5.3) if one is building block b new Is a block b adjacent to the previously established block old And constructed, then T b (b new ) Will become B (T) b (b old ) New node under the head node;
(6) building an exclusion list L exc ；
(6.1) establishing an empty exclusion list L exc ；
(6.2) filling block b with the starting point or the target point start And b end T of b (b start ) And T b (b end ) Adding into L exc Querying the corresponding head node B (T) according to the block association table b (b start ) And B (T) b (b end ) And all their successor nodes are also added to L) exc ；
(6.3) taking all the back connection nodes as head nodes, inquiring the block association table, and adding the back connection nodes into the exclusion list L exc This process is repeated until no subsequent nodes are added to the last node.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910126765.2A CN109827585B (en) | 2019-02-20 | 2019-02-20 | Method for rapidly filling concave area in grid map |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910126765.2A CN109827585B (en) | 2019-02-20 | 2019-02-20 | Method for rapidly filling concave area in grid map |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109827585A CN109827585A (en) | 2019-05-31 |
| CN109827585B true CN109827585B (en) | 2022-08-02 |
Family
ID=66863927
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910126765.2A Active CN109827585B (en) | 2019-02-20 | 2019-02-20 | Method for rapidly filling concave area in grid map |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109827585B (en) |
Families Citing this family (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110726416A (en) * | 2019-10-23 | 2020-01-24 | 西安工程大学 | Reinforced learning path planning method based on obstacle area expansion strategy |
| CN114035581A (en) * | 2021-11-15 | 2022-02-11 | 上海交通大学宁波人工智能研究院 | Traversal path planning method for flaw detection robot with concave obstacle map |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US5537495A (en) * | 1994-03-11 | 1996-07-16 | Hewlett-Packard Company | Pixel correction and smoothing method |
| EP1329692A1 (en) * | 2002-01-17 | 2003-07-23 | Robert Bosch Gmbh | Method for determining map data |
| CN106248097A (en) * | 2015-12-21 | 2016-12-21 | 北京智谷睿拓技术服务有限公司 | Air navigation aid and equipment |
| CN106480215A (en) * | 2016-12-05 | 2017-03-08 | 北京华生恒业科技有限公司 | A kind of method of analyzing and processing crops SSR marker collection of illustrative plates and its device |
| CN109000646A (en) * | 2018-04-25 | 2018-12-14 | 深圳臻迪信息技术有限公司 | Unmanned boat paths planning method and system based on concave polygon scanning area |
-
2019
- 2019-02-20 CN CN201910126765.2A patent/CN109827585B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US5537495A (en) * | 1994-03-11 | 1996-07-16 | Hewlett-Packard Company | Pixel correction and smoothing method |
| EP1329692A1 (en) * | 2002-01-17 | 2003-07-23 | Robert Bosch Gmbh | Method for determining map data |
| DE10201519A1 (en) * | 2002-01-17 | 2003-07-31 | Bosch Gmbh Robert | Method for determining card data |
| CN106248097A (en) * | 2015-12-21 | 2016-12-21 | 北京智谷睿拓技术服务有限公司 | Air navigation aid and equipment |
| CN106480215A (en) * | 2016-12-05 | 2017-03-08 | 北京华生恒业科技有限公司 | A kind of method of analyzing and processing crops SSR marker collection of illustrative plates and its device |
| CN109000646A (en) * | 2018-04-25 | 2018-12-14 | 深圳臻迪信息技术有限公司 | Unmanned boat paths planning method and system based on concave polygon scanning area |
Non-Patent Citations (6)
| Title |
|---|
| Extraction of topological map based on the characteristic comers from grid environment exploration;Su, Liying;《2006 IEEE CONFERENCE ON ROBOTICS, AUTOMATION AND MECHATRONICS》;20061231;1-6 * |
| Improved techniques for grid mapping with Rao-Blackwellized particle filters;Grisetti, G;《IEEE Trans. Robot》;20071231;第23卷(第2期);34-46 * |
| 基于栅格的移动机器人区域分解遍历算法;朱宝艳;《山东理工大学学报(自然科学版)》;20170504;第31卷(第04期);13-18 * |
| 结合粒子群算法和改进蚁群算法的机器人混合路径规划;宋彬;《中国优秀博硕士学位论文全文数据库(硕士)信息科技辑》;20190215(第02期);I140-176 * |
| 自主移动机器人动态环境感知与构图研究;肖瑶;《中国优秀博硕士学位论文全文数据库(硕士)信息科技辑》;20180315(第03期);I140-654 * |
| 蚁群算法在路径规划中的应用;张元良;《数字技术与应用》;20130515(第05期);159-160 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109827585A (en) | 2019-05-31 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US11922574B2 (en) | Method and device for determining plurality of layers of bounding boxes, collision detection method and device, and motion control method and device | |
| JPH0452856A (en) | Method and device for allocating processing of document | |
| CN109827585B (en) | Method for rapidly filling concave area in grid map | |
| CN107092978B (en) | Shortest path layered planning method for virtual earth | |
| CN111080786A (en) | BIM-based indoor map model construction method and device | |
| CN110763247A (en) | Robot path planning method based on combination of visual algorithm and greedy algorithm | |
| US9910878B2 (en) | Methods for processing within-distance queries | |
| CN110909961A (en) | BIM-based indoor path query method and device | |
| CN113865589A (en) | Long-distance rapid path planning method based on terrain slope | |
| KR101575639B1 (en) | Tile image update system for map service and method thereof | |
| AU2013399651B2 (en) | Global grid building in reverse faulted areas by an optimized unfaulting method | |
| CN114404984A (en) | Data processing method and device for game scene, computer equipment and medium | |
| CN115779424B (en) | Navigation grid path finding method, device, equipment and medium | |
| CN113034515A (en) | Bounding box tree-based polygon clipping method, electronic device and storage medium | |
| CN110975291B (en) | Path extraction method and system | |
| KR20210077975A (en) | Spatial indexing method and apparatus for blockchain-based geospatial data | |
| CN116090395A (en) | Data processing method, data structure generating method and query method | |
| CN114817775A (en) | Method and device for generating spatial index | |
| CN114861444A (en) | Unstructured grid earth system mode observation sparsification method based on KD tree | |
| CN110880180B (en) | Virtual cutting algorithm of deformable object based on grid optimization | |
| JP2008299618A (en) | Device, method and program for enhancing image quality | |
| KR20220099745A (en) | A spatial decomposition-based tree indexing and query processing methods and apparatus for geospatial blockchain data retrieval | |
| JP2644735B2 (en) | Drawing information management method | |
| KR20200068901A (en) | Method and apparatus for encoding map image based on condition | |
| Yin et al. | Incremental construction of generalized Voronoi diagrams on pointerless quadtrees |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |