CN112762940A - Robot walking route planning method, program and storage medium - Google Patents
Robot walking route planning method, program and storage medium Download PDFInfo
- Publication number
- CN112762940A CN112762940A CN202011619748.1A CN202011619748A CN112762940A CN 112762940 A CN112762940 A CN 112762940A CN 202011619748 A CN202011619748 A CN 202011619748A CN 112762940 A CN112762940 A CN 112762940A
- Authority
- CN
- China
- Prior art keywords
- image
- boundary
- robot
- entropy
- polygon
- 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.)
- Pending
Links
Images
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01C—MEASURING DISTANCES, LEVELS OR BEARINGS; SURVEYING; NAVIGATION; GYROSCOPIC INSTRUMENTS; PHOTOGRAMMETRY OR VIDEOGRAMMETRY
- G01C21/00—Navigation; Navigational instruments not provided for in groups G01C1/00 - G01C19/00
- G01C21/20—Instruments for performing navigational calculations
Abstract
The invention discloses a robot walking route planning method, a program and a storage medium, which are characterized in that firstly, information of an RGB image T of a scene is acquired, and then the image T is converted into a gray image A; after a gray image A with the size of M multiplied by N is input and obtained, binaryzation is carried out on the image A by adopting an improved algorithm of Renyi entropy to obtain an optimal threshold value, so that the entropy value H of the image A is enabled to be equal to the maximum valueAObtaining a binarization image B at the maximum; carrying out boundary search on the binary image B to obtain an original boundary, and carrying out polygonization processing on the obtained original boundary to obtain an expanded region polygon boundary of an obstacle in a scene; identifying the vertexes of each outer convex angle of the polygon boundary of the expansion area one by one, and forming a safety fitting curve by using the vertexes; and planning the walking route of the robot by taking the safety fitting curve as a reference boundary. Therefore, the invention has the advantages of high walking efficiency, small lateral impact, accurate boundary identification and good tolerance performance.
Description
Technical Field
The present invention relates to the field of intelligent robot path planning strategies, and in particular, to a method, program and storage medium for robot path planning.
Background
The two-dimensional plane multi-robot path planning technology has achieved great results and is widely applied. In the research of the path planning technology of the existing multi-mobile robot system, a two-dimensional plane path planning solution is basically proposed aiming at the possible conflict of the mobile robot driving on the ground. The prior art, such as the chinese patent publication No. CN102929279B, discloses a robot path planning method, which is technically characterized by comprising a main thread, a planning thread, a collision avoidance thread and an execution thread which are executed in parallel in a scheduling server, wherein the main thread initializes a three-dimensional mesh path diagram and updates in real time, the planning thread divides tasks, task paths, marks task paths and sends the task paths into an execution queue, the collision avoidance thread divides a collision zone and executes a command escaping task before a command, and the execution thread sends an execution command to a mobile robot or an elevator robot. The invention realizes the automatic switching of the mobile robot and the elevator robot and the requirement of randomly accessing goods across floors by a method of jointly completing the task walking route through the relay operation of different kinds of robots, and has the characteristics of high efficiency, reasonability, instantaneity and the like.
Disclosure of Invention
The invention aims to provide a robot walking route planning method, a program and a storage medium with high walking efficiency, small lateral impact, accurate boundary identification and good tolerance performance aiming at the prior art.
The technical scheme adopted by the invention for solving the technical problems is as follows: the robot walking route planning method comprises the steps of converting an image T into a gray image A by acquiring information of an RGB image T of a scene; after a gray image A with the size of M multiplied by N is input and obtained, binaryzation is carried out on the image A by adopting an improved algorithm of Renyi entropy to obtain an optimal threshold value, so that the entropy value H of the image A is enabled to be equal to the maximum valueAObtaining a binarization image B at the maximum; carrying out boundary search on the binary image B to obtain an original boundary, and carrying out polygonization processing on the obtained original boundary to obtain an expanded region polygon boundary of an obstacle in a scene; identifying the vertexes of each outer convex angle of the polygon boundary of the expansion area one by one, and forming a safety fitting curve by using the vertexes; and planning the walking route of the robot by taking the safety fitting curve as a reference boundary. By adopting a new binarization algorithm, the outer contour boundary of the barrier is simplified and then optimized, so that a smoother path is divided and a practicable space is provided. Through new walking path logic, the transverse impact during walking is reduced, the transported object is prevented from sideslipping, and the optimized outline boundary is smooth and larger than the outline of the obstacle, so that the situation that the walking is clamped between the obstacles can be avoided.
In order to optimize the technical scheme, the adopted measures further comprise: the algorithm for obtaining the optimal threshold value by the Renyi entropy improvement algorithm is as follows:
wherein L is the number of gray levels, (t)*(α)，s*(α)) is an optimal threshold, (t, s) is a two-dimensional vector;
the Renyi entropy for the background is the entropy of,
renyi entropy as the obstacle object; epsilon is a correction coefficient used for shielding shadow influence. The problem of identification comfort of human senses to images after binarization can be solved by adopting Renyi entropy under a common condition, but the binarized images generated by adopting a general method cannot be used for robot stroke planning and need to be further denoised. The introduction of the correction factor epsilon into the whole calculation introduces a correction factor.
wherein p (i, j) is the symbiotic probability. The value of epsilon is
Where R, G, B represent the RGB values of the red, green, blue color at that point of the image T. Reducing the outline boundary of the image by the information in the color image to make the image in the later positionThe number of sharp angles generated by the regular polygon bounding algorithm is reduced, and the method can reflect the area of the robot more truly.
The polygonization processing is as follows: enclosing the original boundary of the binary image B by adopting a scanning line algorithm to obtain a polygon boundary; and expanding the radius distance of the mobile robot outside the polygon boundary to obtain the polygon boundary of the expanded area. After the polygon boundary is expanded, the robot can be simplified into one point, and the calculation force for comparing the radius distance can be saved.
The algorithm for planning the walking route of the robot is specifically as follows: and (4) taking a tangent point from the O point and the safety fitting curve, extending the radius of the robot at the tangent point, taking the tangent point from the safety fitting curve, and circulating until no barrier exists between the robot and the target area. The path planning method has small horizontal impact on the path, can keep the stability of the object on the robot conveying platform at a higher speed and avoid deviation.
Alpha is 0.81 to 0.83. Through the research on the complex environment, under the planning strategy with the extended robot radius, the alpha value cannot be used in the general value range of the general graphic processing.
The invention also provides a computer program for implementing the method.
The present invention also provides a storage medium storing the computer program.
Because the invention adopts the method of converting the image T into the gray image A; after a gray image A with the size of M multiplied by N is input and obtained, binaryzation is carried out on the image A by adopting an improved algorithm of Renyi entropy to obtain an optimal threshold value, so that the entropy value H of the image A is enabled to be equal to the maximum valueAObtaining a binarization image B at the maximum; carrying out boundary search on the binary image B to obtain an original boundary, and carrying out polygonization processing on the obtained original boundary to obtain an expanded region polygon boundary of an obstacle in a scene; identifying the vertexes of each outer convex angle of the polygon boundary of the expansion area one by one, and forming a safety fitting curve by using the vertexes; the technical scheme of planning the walking route of the robot by taking the safety fitting curve as the reference boundary, and the like, so the invention has the advantages of high walking efficiency, small lateral impact, accurate boundary identification and good tolerance performance。
Drawings
FIG. 1 is a schematic diagram of a principle of construction of a fitting curve according to an embodiment of the present invention;
FIG. 2 is a schematic diagram of an example of image binarization and fitting curve construction according to an embodiment of the invention;
FIG. 3 is a diagram illustrating an example of image path comparison according to an embodiment of the present invention.
Detailed Description
The present invention will be described in further detail with reference to the following examples.
The reference numbers illustrate: the mobile robot comprises a mobile robot O, a polygon boundary 1, an extended area polygon boundary 2, a fitting curve 3, a safety fitting curve 4, a person 5 and a target area 6.
Example 1:
referring to fig. 1-3, a robot walking route planning method includes acquiring information of an RGB image T of a scene, and then converting the image T into a gray image a; after a gray image A with the size of M multiplied by N is input and obtained, binaryzation is carried out on the image A by adopting an improved algorithm of Renyi entropy to obtain an optimal threshold value, so that the entropy value H of the image A is enabled to be equal to the maximum valueAObtaining a binarization image B at the maximum; carrying out boundary search on the binary image B to obtain an original boundary, and carrying out polygonization processing on the obtained original boundary to obtain an expanded region polygon boundary 2 of an obstacle in a scene; identifying the vertexes of each external salient angle of the extended area polygon boundary 2 one by one, and forming a safety fitting curve 4 by using the vertexes; and planning the walking route of the robot O by taking the safety fitting curve 4 as a reference boundary. By adopting a new binarization algorithm, the outer contour boundary of the barrier is simplified and then optimized, so that a smoother path is divided and a practicable space is provided. Through new walking path logic, the transverse impact during walking is reduced, the transported object is prevented from sideslipping, and the optimized outline boundary is smooth and larger than the outline of the obstacle, so that the situation that the walking is clamped between the obstacles can be avoided.
The algorithm for obtaining the optimal threshold value by the Renyi entropy improvement algorithm is as follows:
wherein L is the number of gray levels, (t)*(α)，s*(α)) is an optimal threshold, (t, s) is a two-dimensional vector;
the Renyi entropy for the background is the entropy of,
renyi entropy as the obstacle object; epsilon is a correction coefficient used for shielding shadow influence. In general, the problem of identification comfort of human senses to images after binarization can be solved by adopting Renyi entropy, but the binarized images generated by adopting a general method cannot be used for robot O-stroke planning and need to be further denoised. The introduction of the correction factor epsilon into the whole calculation introduces a correction factor.
wherein p (i, j) is the symbiotic probability. The value of epsilon is
Wherein, R, G, B represent RGB values of red, green, blue at the point of the image T. The outer coating outline boundary of the graph is reduced through the information in the color image, so that the number of acute angles generated by a post-processing polygon bounding algorithm is reduced, and the region which can pass through a robot can be reflected more truly. The polygonization processing is as follows: enclosing the original boundary of the binary image B by adopting a scanning line algorithm to obtain a polygon boundary 1; expanding the polygon boundary 1 to the radius distance of the mobile robot OThe extended region polygon limit 2 is obtained. After the polygon boundary 1 is expanded, the robot O can be simplified into a point, and the calculation force for comparing the radius distance can be saved. The algorithm for planning the walking route of the robot O specifically comprises the following steps: and (3) taking a tangent point of the point O and the safety fitting curve 4, extending the radius of the robot O at the tangent point, taking the tangent point with the safety fitting curve 4, and circulating until no barrier exists between the robot O and the target area 6. The path planning method has small horizontal impact on the path, can keep the stability of the object on the robot conveying platform at a higher speed and avoid deviation. Alpha is 0.81 to 0.83. Through the research on the complex environment, under the planning strategy with the extended robot radius, the alpha value cannot be used in the general value range of the general graphic processing.
Example 2
An RGB image T of a scene is acquired, and the image T is converted into a gray image A.
After a gray image A with the size of M multiplied by N is input and obtained, the image A is binarized by adopting an improved algorithm of Renyi entropy, and an optimal threshold value is obtained by adopting the following algorithm, so that the entropy value H of the image A is enabled to be equal to the threshold value HAMaximum:
l is the number of gray levels, (t)*(α)，s*(α)) is an optimum threshold value, and (t, s) is a two-dimensional vector.
The Renyi entropy for the background is the entropy of,
is the Renyi entropy of the obstacle. In the formula 1, the gray value is i and the 3 × 3 neighborhood mean j of the pixel point is used to represent the image, the two-dimensional vector (t, s) is used as the segmentation threshold, the related content of the above definition is a common convention in the prior art, and the obtaining method is not further described. Epsilon is a correction coefficient, and epsilon should be adjusted according to the conditions of illumination light and color so as to shield shadow influence.
The value of epsilon is
Where R, G, B represent the RGB values of the red, green, blue color at that point of the image T. And (3) performing illumination description on the color by using epsilon so as to adjust the shadow-to-edge detection error of the gray image A. The shadow is adjusted, so that the problem that invalid vertexes influencing the path planning efficiency are formed when the vertexes of each outer convex angle of the polygon boundary 2 of the extension area are identified one by one in the later stage can be avoided.
The gray level of the gray image A with the size of M multiplied by N is L, the gray value of the pixel point (M, N) is f (M, N), and f (M, N) is more than or equal to 0 and less than or equal to L-1.
The implementation pseudo-code for entropy threshold partitioning is as follows:
inputting a grayscale image
Calculating the gray level-neighborhood mean probability p (i, j) of the pixel point to find out the gray level Hist corresponding to the maximum probability of f (m, n)max；
for loop (loop variable a from HistmaxTo 254)
for loop (loop variable c from a +1 to 255)
If p (i, j) ≠ 0, the Renyi entropy of expression (2-11) is calculated
End for cycle (cycle variable c)
End for cycle (cycle variable a)
Finding out the threshold value pair (t) corresponding to the maximum Renyi entropy by the formula 1*(α)，s*(α))
The optimal segmentation threshold is t*(α), dividing the grayscale image
End up
The value of the Renyi entropy parameter alpha is 0.81 to 0.83, alpha is a real constant, alpha is more than 0, and alpha is not equal to 1.
Formula 4 is the definition of Renyi entropy, wherein b is the logarithm with base 2, and p (i, j) is the symbiotic probability; renyi entropy is commonly used for calculating a binary threshold, but a problem of meaningless expansion of a boundary can be caused by a mixed point of an image obtained by directly applying the Renyi entropy in path planning. The value of the entropy parameter alpha is generally 0.8, but dynamic or static images T are not beneficial to binarization due to light, ISO and exposure time because people 5 carrying conveying lines and moving or standing are moving objects.
The binarized image B obtained by dividing the grayscale image a by the optimal division threshold value further needs to be subjected to the obstacle describing process.
Carrying out boundary search on the binary image B to obtain an original boundary, and carrying out polygonization processing on the obtained original boundary, namely surrounding the original boundary of the binary image B by adopting a scanning line algorithm to obtain a polygonal boundary 1; and (4) expanding the radius distance of the mobile robot O outside the polygon boundary 1 to obtain an expanded region polygon boundary 2.
An obstacle M ═ M in the workspace considered as an extended region polygon boundary 2 containing a finite number of irregular shapes1，M2，...，Mn}。
Let O be a mobile robot, such as a food delivery robot or a robot that transports dishes to be washed. Each obstacle is seen as an open-set convex polygon consisting of a limited number of vertices, only the vertices of the convex polygon and the sides that make up the convex polygon, regardless of the points inside the convex polygons. Considering the mobile robot O as a circle, i.e. representing the mobile robot by a circle, the assumed benefit is to allow the sides of the circular robot to be tangent to the sides 1 of the obstacle. Preferably, the path planning may be performed by enlarging the radius of the mobile robot O to form a new traveling boundary 2 and then processing the robot O into points.
The vertices of each of the outer lobes of the extended region polygon bounding 2 are identified one by one.
Each obstacle MnAnd the extended region polygon boundary 2 is defined as O 'from the vertex of the outer lobe closest to the mobile robot O'nN is 1, 2. From O'nThe vertices of the outer lobes of the extended region polygon boundary 2 of the obstacle are connected one by one for the initial fitted points by a fitted curve 3 and by O'nFor the endpoint, a region of a closed fitted curve 3 is formed to represent each obstacle. When the extended region polygon boundary 2 and the fitting curve 3 intersect outside the vertexes of the outer convex angles, the vertexes of the two outer convex angles closest to the intersection position are taken as end points, and the fitting point between the vertexes of the two outer convex angles is deleted to form a safety fitting curve 4. The starting point of the fitted curve 3 for each identified obstacle, which assumes the vertex of the closest two outer lobes, is not normally identified in the plan for generating a tangent. Under the condition of small probability, when the initial point is identified to need to find a tangent, the initial point is replaced by a perpendicular line of an angle symmetrical bisector.
The path search algorithm adopts an improved ant colony algorithm. Different from the prior art, the logic adopted by the improved algorithm is that a tangent point is taken by the point O and the safety fitting curve 4, the radius of the robot O is extended at the tangent point, then the tangent point is taken by the point O and the safety fitting curve 4, and the operation is circulated until no barrier exists between the robot O and the target area 6. If the bending angle of the last segment is larger, the distance can be preset for prejudgment, the walking speed of the robot O is predecelerated, and the overall planning efficiency is further maintained in a mode of sacrificing the walking efficiency of the last segment. Compared with the prior ant algorithm combined with the technical scheme of quadtree decomposition, the invention has shorter path, and has smaller lateral deviation relative to the technical scheme of path planning by the polygon limit 2 vertex connection of the extended area.
The fitting curve 3 is fitted with a spline function. May be a uniform B-spline curve, a quasi-uniform B-spline curve, a segmented Bezier curve, a non-uniform B-spline. Through comparison, when the robot O walks fast to carry the tableware, the spline curves are tested in a single-factor mode, the influence on the lateral offset sliding of the tableware on the carrying table is not large, and the curves of corresponding types can be selected according to calculation force.
The route P2 obtained by performing route planning by using the conventional IACO algorithm has a more zigzag route and a larger horizontal impact, so that the speed of the robot O on most of the routes cannot be further increased. Forming path P3 in a manner that estimates horizontal impacts based on robot speed and path shape wastes time by forming a spiral-like redundant path in the last segment. By adopting the walking path of the technical scheme of the invention, the shorter path P3 can be calculated and screened out from the feasible path, and the walking path has better performance in path, horizontal impact and walking time.
Example 3
A computer program for implementing the method of embodiments 1 and 2.
Example 4
A storage medium storing the computer program of embodiment 3 described above.
While the invention has been described in connection with a preferred embodiment, it is not intended to limit the invention, and it will be understood by those skilled in the art that various changes may be made and equivalents may be substituted for elements thereof without departing from the spirit and scope of the invention.
Claims (10)
1. The robot walking route planning method comprises the steps of obtaining information of an RGB image T of a scene, and is characterized in that: converting the image T into a gray image A; after a gray image A with the size of M multiplied by N is input and obtained, binaryzation is carried out on the image A by adopting an improved algorithm of Renyi entropy to obtain an optimal threshold value, so that the entropy value H of the image A is enabled to be equal to the maximum valueAObtaining a binarization image B at the maximum; carrying out boundary search on the binary image B to obtain an original boundary, and carrying out polygonization processing on the obtained original boundary to obtain an expanded region polygon boundary (2) of an obstacle in a scene; identifying extended area polygon boundaries one by oneThe vertex of each external lobe of the limit (2) is used for forming a safety fitting curve (4); and planning the walking route of the robot (O) by taking the safety fitting curve (4) as a reference boundary.
2.… …, according to claim 1, wherein: the algorithm for obtaining the optimal threshold value by the Renyi entropy improvement algorithm is as follows:
3. wherein L is the number of gray levels, (t)*(α)，s*(α)) is an optimal threshold, (t, s) is a two-dimensional vector;
the Renyi entropy for the background is the entropy of,
renyi entropy as the obstacle object; epsilon is a correction coefficient used for shielding shadow influence.
4.… …, according to claim 2, wherein: said
The calculation method is as follows:
wherein p (i, j) is the symbiotic probability.
5.… …, according to claim 2, wherein: the value of epsilon is
Where R, G, B represent the RGB values of the red, green, blue color at that point of the image T.
6.… …, according to claim 1, wherein: the polygonization treatment comprises the following steps: enclosing the original boundary of the binary image B by adopting a scanning line algorithm to obtain a polygon boundary (1); and (3) expanding the polygon limit (1) by the radius distance of the mobile robot (O) to obtain the expanded region polygon limit (2).
7.… …, according to claim 1, wherein: the algorithm for planning the walking route of the robot (O) is specifically as follows: and (3) taking a tangent point of the O point and the safety fitting curve (4), extending the radius of the robot (O) at the tangent point, taking the tangent point with the safety fitting curve (4), and circulating until no barrier exists between the robot (O) and the target area (6).
8.… …, according to claim 1, wherein: alpha is 0.81 to 0.83.
9. A computer program for implementing the method of claim 1.
10. A storage medium storing a computer program according to claim 8.
Priority Applications (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011619748.1A CN112762940A (en) | 2020-12-30 | 2020-12-30 | Robot walking route planning method, program and storage medium |
| CN202111547317.3A CN113932818B (en) | 2020-12-30 | 2021-12-17 | Robot walking route planning method, program and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011619748.1A CN112762940A (en) | 2020-12-30 | 2020-12-30 | Robot walking route planning method, program and storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN112762940A true CN112762940A (en) | 2021-05-07 |
Family
ID=75698182
Family Applications (2)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202011619748.1A Pending CN112762940A (en) | 2020-12-30 | 2020-12-30 | Robot walking route planning method, program and storage medium |
| CN202111547317.3A Active CN113932818B (en) | 2020-12-30 | 2021-12-17 | Robot walking route planning method, program and storage medium |
Family Applications After (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111547317.3A Active CN113932818B (en) | 2020-12-30 | 2021-12-17 | Robot walking route planning method, program and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (2) | CN112762940A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115326099A (en) * | 2022-10-11 | 2022-11-11 | 禾多科技（北京）有限公司 | Local path planning method and device, electronic equipment and computer readable medium |
Family Cites Families (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP6303332B2 (en) * | 2013-08-28 | 2018-04-04 | 富士通株式会社 | Image processing apparatus, image processing method, and image processing program |
| CN103810716B (en) * | 2014-03-13 | 2016-07-06 | 北京工商大学 | Move and the image partition method of Renyi entropy based on gray scale |
| CN106546233A (en) * | 2016-10-31 | 2017-03-29 | 西北工业大学 | A kind of monocular visual positioning method towards cooperative target |
| CN107807644A (en) * | 2017-10-30 | 2018-03-16 | 洛阳中科龙网创新科技有限公司 | A kind of farm machinery consumption minimization trajectory path planning method |
| CN110865642A (en) * | 2019-11-06 | 2020-03-06 | 天津大学 | Path planning method based on mobile robot |
| CN112346463B (en) * | 2020-11-27 | 2022-09-23 | 西北工业大学 | Unmanned vehicle path planning method based on speed sampling |
-
2020
- 2020-12-30 CN CN202011619748.1A patent/CN112762940A/en active Pending
-
2021
- 2021-12-17 CN CN202111547317.3A patent/CN113932818B/en active Active
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115326099A (en) * | 2022-10-11 | 2022-11-11 | 禾多科技（北京）有限公司 | Local path planning method and device, electronic equipment and computer readable medium |
| CN115326099B (en) * | 2022-10-11 | 2023-03-07 | 禾多科技（北京）有限公司 | Local path planning method and device, electronic equipment and computer readable medium |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113932818B (en) | 2022-03-29 |
| CN113932818A (en) | 2022-01-14 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Park et al. | Pix2pose: Pixel-wise coordinate regression of objects for 6d pose estimation | |
| JP6573734B2 (en) | Robust merging of 3D textured meshes | |
| US11763459B2 (en) | Method and computing system for object identification | |
| KR100643305B1 (en) | Method and apparatus for processing line pattern using convolution kernel | |
| CN113932818B (en) | Robot walking route planning method, program and storage medium | |
| CN110648359B (en) | Fruit target positioning and identifying method and system | |
| CN114972377A (en) | 3D point cloud segmentation method and device based on moving least square method and hyper-voxels | |
| CN113269801A (en) | Method and computing system for processing candidate edges | |
| CN115330969A (en) | Local static environment vectorization description method for ground unmanned vehicle | |
| CN115311172A (en) | Map area segmentation method and related device | |
| US8717356B2 (en) | Display processing method and apparatus | |
| KR20220094089A (en) | Method, system, and non-transitory computer-readable recording medium for generating a map for a robot | |
| US20210200232A1 (en) | Method of generating scan path of autonomous mobile robot and computing device | |
| CN111915724A (en) | Point cloud model slice shape calculation method | |
| CN105701807B (en) | A kind of image partition method based on temporal voting strategy | |
| CN113343987B (en) | Text detection processing method and device, electronic equipment and storage medium | |
| CN115272465A (en) | Object positioning method, device, autonomous mobile device and storage medium | |
| WO2021220345A1 (en) | Elevator 3d data processing device | |
| CN112734783B (en) | Method and computing system for processing candidate edges | |
| CN114092498A (en) | Method for segmenting an object in an image | |
| Trisiripisal | Image approximation using triangulation | |
| Hiransakolwong et al. | FASU: A full automatic segmenting system for ultrasound images | |
| CN117611839A (en) | Method for calculating convex hull based on sub-quadrilateral structure in image processing | |
| CN114653629A (en) | Sorting method based on visual identification, intelligent sorting system and readable storage medium | |
| TW202304209A (en) | Content patch coding method and content patch decoding method |
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 | ||
| WD01 | Invention patent application deemed withdrawn after publication | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20210507 |