CN110175523A - A kind of self-movement robot animal identification and hide method and its storage medium - Google Patents
A kind of self-movement robot animal identification and hide method and its storage medium Download PDFInfo
- Publication number
- CN110175523A CN110175523A CN201910342589.6A CN201910342589A CN110175523A CN 110175523 A CN110175523 A CN 110175523A CN 201910342589 A CN201910342589 A CN 201910342589A CN 110175523 A CN110175523 A CN 110175523A
- Authority
- CN
- China
- Prior art keywords
- frame
- animal
- self
- movement robot
- movement
- 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.)
- Granted
Links
Classifications
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05D—SYSTEMS FOR CONTROLLING OR REGULATING NON-ELECTRIC VARIABLES
- G05D1/00—Control of position, course or altitude of land, water, air, or space vehicles, e.g. automatic pilot
- G05D1/02—Control of position or course in two dimensions
- G05D1/021—Control of position or course in two dimensions specially adapted to land vehicles
- G05D1/0231—Control of position or course in two dimensions specially adapted to land vehicles using optical position detecting means
- G05D1/0246—Control of position or course in two dimensions specially adapted to land vehicles using optical position detecting means using a video camera in combination with image processing means
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
Abstract
A kind of self-movement robot animal identification and hide method and its storage medium, the method be the environmental information acquired around self-movement robot, it obtains RGB and schemes and depth map, the identification using CNN to animal;The pixel of animal is removed, realizes vision inertia odometer;Calculate the transition matrix between self-movement robot b1 frame and b2 frameThe pixel of animal part is extracted, the transition matrix between b1 frame and b2 frame is calculatedThe depth map of animal is changed into point cloud chart, using ICP to b1Frame, b2Point cloud chart between frame is matched;Animal is in b1Transition matrix under frame coordinate systemDriving self-movement robot movement makes post exercise coordinate system and b1The transition matrix of frame referential isSelf-movement robot is set to keep the position orientation relation kept constant with animal.The present invention had both improved the ability of getting rid of poverty of self-movement robot, also improved self-movement robot practicability, intelligence and environmental interaction.
Description
Technical field
The present invention relates to self-movement robot fields, particularly, be related to a kind of self-movement robot to the identification of animal, estimate
It counts the movement of animal and hides the method and storage medium of animal, to improve self-movement robot practicability, intelligence and ring
Border interactivity.
Background technique
Self-movement robot works in environment indoors, and pet is the most common animal in indoor environment.From moving machine
Device people is not only influenced by animal during the motion, can also impact to the environment of animal, such as self-movement robot
It is chased in moving process by animal, not only results in the damage of self-movement robot, it is also possible to animal sheet can be hurt
Body.Indoor self-movement robots most at present do not have identification and hide the function of animal, therefore are easy to produce above ask
Topic, causes self-movement robot practicability, intelligence and environmental interaction to have certain defect.
Therefore, how to identify animal, when the pose of animal and self-movement robot is less than preset level, hide animal simultaneously
The technical issues of becoming prior art urgent need to resolve with the position orientation relation that animal keeps constant.
Summary of the invention
It, should it is an object of the invention to propose a kind of self-movement robot animal identification and hide method and its storage medium
Method enables to self-movement robot to move, the position orientation relation kept constant with animal.Both it is practical self-movement robot had been improved
Property, intelligence and environmental interaction, also enhance self-movement robot and get rid of poverty ability.
To achieve this purpose, the present invention adopts the following technical scheme:
A kind of self-movement robot animal identification and hide method, which comprises the steps of:
Animal identification step S110: RGB figure and the depth map being obtained from front of moveable robot movement pass through convolutional Neural
Network (CNN) schemes the RGB to carry out animal identification, when recognizing animal, judges animal distance from shifting by depth image
The pose of mobile robot carries out the following steps of this method when the pose is less than preset level;
Transition matrixWithCalculate step S120:
The conversion square between self-movement robot b1 frame and b2 frame is calculated using RGB figure, depth map without animal
Battle array
Extract b1、b2The corresponding depth pixel data of animal rgb pixel in frame calculate animal between b1 frame and b2 frame
Transition matrix
Transition matrix of the two frame animal point clouds under b1 frame referentialCalculate step S130: by the depth of two frame animals
Figure is converted to point cloud chart, to being transformed into b1Two frame animal point clouds in frame coordinate system are iterated, and calculate two frame animal point clouds
Transition matrix under b1 frame referential
Actuation step S140: driving self-movement robot movement makes post exercise coordinate system and b1The conversion square of frame referential
Battle array beThe position orientation relation for keeping constant self-movement robot and animal.
Optionally, described that the RGB is schemed to carry out by convolutional neural networks (CNN) in animal identification step S110
Animal identification specifically: the convolutional neural networks utilize convolutional layer, pond layer, full articulamentum, generate classifier and are predicted
Identification;Output matrix is obtained by being multiplied with convolution kernel in the convolutional layer, feature is extracted from image;The pond layer reduces
Feature vector dimension reduces over-fitting, reduces noise transmitting;The full articulamentum the tensor of pond layer be cut into
Amount, is multiplied by weight, is used for ReLU activation primitive, with gradient descent method Optimal Parameters, generates classifier;Eventually by described
Classifier carries out Forecasting recognition.
Optionally, it after identifying animal, also using the RGB figure and depth map obtained in advance, respectively obtains without dynamic
The RGB of object schemes and depth map, and RGB figure and depth map containing only animal, for estimating animal movement initial value.
Optionally, wherein transition matrixCalculating specifically: using IUM obtain self-movement robot angular speed and plus
IMU data pre-integration between b1 frame and b2 frame is obtained IMU between b1 frame and b2 frame and measures residual error, according to re-projection by speed
The residual error of error calculation image detects latest frame b using slip window sampling2Frame previous frame b therewith1Whether frame has stable spy
Sign, if there is stable feature then by latest frame, is added in sliding window, using used based on sliding window close coupling vision
Property odometer (vision VIO) calculates the transition matrix between b1 frame and b2 frameAnd/or
Wherein transition matrixCalculating specifically: extract b1、b2The corresponding depth pixel number of animal rgb pixel in frame
According to calculating conversion of the animal between b1 frame and b2 frame by the RGB of animal figure and depth map using direct linear transformation (DLT)
Matrix
Optionally, transition matrix of the two frame animal point clouds under b1 frame referentialCalculate step S130 specifically:
By reference frame b1, present frame b2The depth map of animal switchs to point cloud chart in frame, passes through two field pictures b1Frame and b2Frame it
Between transition matrixBy present frame b2The point cloud data of animal switchs to reference frame b in frame1The point cloud of animal in frame coordinate system,
Using ICP (Iterative Closest Point iteration closest approach) algorithm by being transformed into b1Two frames in frame coordinate system
Animal point cloud iteration, willAs the initial value of above-mentioned ICP iteration, which can make two frame animal point cloud fast convergences,
Calculate transition matrix of the two frame animal point clouds under b1 frame referential
Optionally, the self-movement robot has depth camera and IMU, and the depth camera is for acquiring from moving machine
Environmental information around device people, obtains RGB figure and depth map, and the IMU is used to obtain the angular speed of self-movement robot and adds
Speed.
Optionally, the self-movement robot circular flow step S110-S140, the self-movement robot circular flow
Step S110-S140 obtains next frame animal RGB figure, depth map, identifies animal, calculatesCalculate two frame animals
Transition matrix of the point cloud under b1 frame referentialMake post exercise coordinate system and b1The transition matrix of frame referential isDriving self-movement robot movement makes self-movement robot keep the position orientation relation kept constant with animal.
The invention also discloses a kind of storage mediums, for storing computer executable instructions, it is characterised in that:
The computer executable instructions executed when being executed by processor above-mentioned self-movement robot animal identification with
Hide method.
The present invention further discloses a kind of self-movement robots, with above-mentioned storage medium, it is characterised in that: institute
Storage medium is stated to execute above-mentioned self-movement robot animal identification and hide method.
The present invention further discloses a kind of self-movement robots, it is characterised in that: the self-movement robot has deep
Camera and IMU are spent, and is able to carry out above-mentioned self-movement robot animal identification and hides method.
To sum up, self-movement robot can identify animal, estimation animal movement in the present invention, hide animal and protect with animal
Hold constant position orientation relation.Both the ability of getting rid of poverty for having improved self-movement robot, also improves self-movement robot practicability, intelligence
It can property and environmental interaction.Overwhelming majority self-movement robot does not have this kind of function at present, this function can make from moving machine
The interactivity that device people and animal keep friends with.
Detailed description of the invention
Fig. 1 is the self-movement robot animal identification of specific embodiment according to the present invention and the flow chart for hiding method;
Fig. 2 is the iteration initial value of specific embodiment according to the present inventionThe step of;
The step of Fig. 3 is the calculating animal movement of specific embodiment according to the present invention and drives self-movement robot.
Specific embodiment
The present invention is described in further detail with reference to the accompanying drawings and examples.It is understood that this place is retouched
The specific embodiment stated is used only for explaining the present invention rather than limiting the invention.It also should be noted that in order to just
Only the parts related to the present invention are shown in description, attached drawing rather than entire infrastructure.
The invention reside in so that self-movement robot has depth camera and IMU (Inertial Measurement Unit), wherein depth phase
Machine is used to acquire environmental information around self-movement robot, obtains RGB figure and depth map, animal and estimates animal for identification
Movement.When the pose of animal and self-movement robot is less than preset level, self-movement robot will be moved, and be protected with animal
Constant position orientation relation is held, makes animal will not be further towards self-movement robot.
Specifically, realizing identification to animal using convolutional neural networks, the pixel of animal is removed, using IMU and
Vision inertia odometer is realized in camera fusion;Calculate the transition matrix between self-movement robot b1 frame and b2 frameIt extracts dynamic
The pixel of object part calculates transition matrix of the animal between b1 frame and b2 frameTwo frame animal point clouds are calculated to join in b1 frame
Examine the transition matrix under beingDriving self-movement robot movement makes post exercise coordinate system and b1The conversion square of frame referential
Battle array beSelf-movement robot is set to keep the position orientation relation kept constant with animal.
Specifically, showing self-movement robot animal identification referring to Fig. 1 and hiding the flow chart of method, including such as
Lower step:
Animal identification step S110: RGB figure and the depth map being obtained from front of moveable robot movement pass through convolutional Neural
Network (CNN) schemes the RGB to carry out animal identification, when recognizing animal, judges animal distance from shifting by depth image
The pose of mobile robot carries out the following steps of this method when the pose is less than preset level.
In an alternative embodiment, described that the RGB is schemed by convolutional neural networks (CNN) to carry out animal identification
Specifically: the convolutional neural networks utilize convolutional layer, pond layer, full articulamentum, generate classifier and carry out Forecasting recognition;It is described
Output matrix is obtained by being multiplied with convolution kernel in convolutional layer, feature is extracted from image;The pond layer reduces feature vector
Dimension reduces over-fitting, reduces noise transmitting;The tensor of pond layer is cut into vector by the full articulamentum, is multiplied by power
Weight, is used for ReLU activation primitive, with gradient descent method Optimal Parameters, generates classifier;Eventually by the classifier into
Row Forecasting recognition.
Further, it after identifying animal, also using the RGB figure and depth map obtained in advance, respectively obtains and is free of
The RGB of animal schemes and depth map, and RGB figure and depth map containing only animal, with the calculating for subsequent conversion matrix.
In the present invention, the self-movement robot has depth camera and IMU, and the depth camera is for acquiring from shifting
Environmental information around mobile robot, obtains RGB figure and depth map, and the IMU is used to obtain the angular speed of self-movement robot
And acceleration.
Transition matrixWithCalculate step S120:
The step including the use of RGB figure, depth map without animal be calculated self-movement robot b1 frame and b2 frame it
Between transition matrix
Extract b1、b2The corresponding depth pixel data of animal rgb pixel in frame calculate animal between b1 frame and b2 frame
Transition matrix
Wherein transition matrixCalculating specifically: use IUM obtain self-movement robot angular speed and acceleration, will
IMU data pre-integration between b1 frame and b2 frame obtains IMU between b1 frame and b2 frame and measures residual error, according to re-projection error meter
The residual error of nomogram picture detects latest frame b using slip window sampling2Frame previous frame b therewith1Whether frame has stable feature, if
There are stable features then by latest frame, is added in sliding window, using based on sliding window close coupling vision inertia mileage
Meter (vision VIO) calculates the transition matrix between b1 frame and b2 frame
Wherein transition matrixCalculating specifically: extract b1、b2The corresponding depth pixel number of animal rgb pixel in frame
According to calculating conversion of the animal between b1 frame and b2 frame by the RGB of animal figure and depth map using direct linear transformation (DLT)
Matrix
In this step, the calculating of two transition matrixes is to lay base in the calculating of next step iteration initial value
Plinth.
In the present invention, RGB figure and depth map, therefore, present frame b are shot simultaneously using depth camera2With reference frame b1More
More is at the time of indicating shooting image.
Referring to fig. 2, estimation transition matrix is shownWithAnd initial valueRequired corresponding steps.
The format of two transition matrixes is
In formula: R is spin matrix, and t is translation vector.
Transition matrix of the two frame animal point clouds under b1 frame referentialCalculate step S130: by the depth of two frame animals
Figure is converted to point cloud chart, to being transformed into b1Two frame animal point clouds in frame coordinate system are iterated, and calculate two frame animal point clouds
Transition matrix under b1 frame referential
Specifically: by reference frame b1, present frame b2The depth map of animal switchs to point cloud chart in frame, passes through two field pictures b1Frame
With b2Transition matrix between frameBy present frame b2The point cloud data of animal switchs to reference frame b in frame1Animal in frame coordinate system
Point cloud, using ICP (Iterative Closest Point iteration closest approach) algorithm by being transformed into b1In frame coordinate system
Two frame animal point cloud iteration, willAs the initial value of above-mentioned ICP iteration, i.e. two corresponding dot products of matrix, the value
It can make two frame animal point cloud fast convergences, calculate transition matrix of the two frame animal point clouds under b1 frame referential
Actuation step S140: driving self-movement robot movement makes post exercise coordinate system and b1The conversion square of frame referential
Battle array beSelf-movement robot is set to keep the position orientation relation kept constant with animal.
Referring to Fig. 3, shows the calculating animal movement of specific embodiment according to the present invention and drive self-movement robot institute
The corresponding steps needed.
Therefore, by step S110-S140, the position orientation relation for enabling to self-movement robot and animal to keep constant.
Enhance the practicability intelligence and environmental interaction to environment of self-movement robot.
Further, the self-movement robot circular flow step S110-S140 obtains next frame animal RGB and schemes, is deep
Degree figure, identifies animal, calculatesCalculate transition matrix of the two frame animal point clouds under b1 frame referentialDriving
Self-movement robot movement makes post exercise coordinate system and b1The transition matrix of frame referential isMake self-movement robot
Keep the position orientation relation kept constant with animal.
Drive self-movement robot movement.
The present invention further discloses a kind of storage mediums, and for storing computer executable instructions, the computer can
It executes instruction and executes above-mentioned self-movement robot animal identification when being executed by processor and hide method.
The invention also discloses a kind of self-movement robots can execute above-mentioned shifting certainly with above-mentioned storage medium
Mobile robot animal identification and hide method.
Alternatively, a kind of self-movement robot, has depth camera and IMU, it is dynamic to be able to carry out above-mentioned self-movement robot
Object identifies and hides method.
To sum up, self-movement robot can identify animal, estimation animal movement in the present invention, hide animal and protect with animal
Hold constant position orientation relation.Both the ability of getting rid of poverty for having improved self-movement robot, also improves self-movement robot practicability, intelligence
It can property and environmental interaction.Overwhelming majority self-movement robot does not have this kind of function at present, this function can make from moving machine
The interactivity that device people and animal keep friends with.
Obviously, it will be understood by those skilled in the art that above-mentioned each unit of the invention or each step can be with general
Computing device realizes that they can concentrate on single computing device, and optionally, they can be executable with computer installation
Program code realize, be performed by computing device so as to be stored in storage device, or by they point
It is not fabricated to each integrated circuit modules, or makes multiple modules or steps in them to single integrated circuit module
It realizes.In this way, the present invention is not limited to the combinations of any specific hardware and software.
The above content is a further detailed description of the present invention in conjunction with specific preferred embodiments, and it cannot be said that
A specific embodiment of the invention is only limitted to this, for those of ordinary skill in the art to which the present invention belongs, is not taking off
Under the premise of from present inventive concept, several simple deduction or replace can also be made, all shall be regarded as belonging to the present invention by institute
Claims of submission determine protection scope.
Claims (10)
1. a kind of self-movement robot animal identification and hiding method, which comprises the steps of:
Animal identification step S110: RGB figure and the depth map being obtained from front of moveable robot movement pass through convolutional neural networks
(CNN) scheme progress animal identification to the RGB judges animal distance from moving machine when recognizing animal by depth image
The pose of device people carries out the following steps of this method when the pose is less than preset level;
Transition matrixWithCalculate step S120:
The transition matrix between self-movement robot b1 frame and b2 frame is calculated using RGB figure, depth map without animal
Extract b1、b2The corresponding depth pixel data of animal rgb pixel in frame calculate conversion of the animal between b1 frame and b2 frame
Matrix
Transition matrix of the two frame animal point clouds under b1 frame referentialIt calculates step S130: the depth map of two frame animals is turned
It is changed to point cloud chart, to being transformed into b1Two frame animal point clouds in frame coordinate system are iterated, and calculate two frame animal point clouds in b1
Transition matrix under frame referential
Actuation step S140: driving self-movement robot movement makes post exercise coordinate system and b1The transition matrix of frame referential isThe position orientation relation for keeping constant self-movement robot and animal.
2. self-movement robot animal identification according to claim 1 and hiding method, it is characterised in that:
It is described specific to RGB figure progress animal identification by convolutional neural networks (CNN) in animal identification step S110
Are as follows: the convolutional neural networks utilize convolutional layer, pond layer, full articulamentum, generate classifier and carry out Forecasting recognition;The convolution
By being multiplied to obtain output matrix with convolution kernel in layer, feature is extracted from image;The pond layer reduces feature vector dimension,
Reduce over-fitting, reduces noise transmitting;The tensor of pond layer is cut into vector by the full articulamentum, is multiplied by weight, right
It uses ReLU activation primitive, with gradient descent method Optimal Parameters, generates classifier;It is predicted eventually by the classifier
Identification.
3. self-movement robot animal identification according to claim 2 and hiding method, it is characterised in that:
After identifying animal, also using obtain in advance RGB figure and depth map, respectively obtain without animal RGB figure and
Depth map, and RGB figure and depth map containing only animal, for estimating animal movement initial value.
4. self-movement robot animal identification according to claim 1 and hiding method, it is characterised in that:
Wherein transition matrixCalculating specifically: using IUM obtain self-movement robot angular speed and acceleration, by b1 frame
IMU data pre-integration between b2 frame obtains IMU between b1 frame and b2 frame and measures residual error, calculated and schemed according to re-projection error
The residual error of picture detects latest frame b using slip window sampling2Frame previous frame b therewith1Whether frame has stable feature, if there is
Stable feature is then added to latest frame in sliding window, using based on sliding window close coupling vision inertia odometer (depending on
Feel VIO) calculate transition matrix between b1 frame and b2 frameAnd/or
Wherein transition matrixCalculating specifically: extract b1、b2The corresponding depth pixel data of animal rgb pixel in frame use
Direct linear transformation (DLT) calculates transition matrix of the animal between b1 frame and b2 frame by the RGB figure and depth map of animal
5. self-movement robot animal identification according to claim 1 and hiding method, it is characterised in that:
Transition matrix of the two frame animal point clouds under b1 frame referentialCalculate step S130 specifically:
By reference frame b1, present frame b2The depth map of animal switchs to point cloud chart in frame, passes through two field pictures b1Frame and b2Between frame
Transition matrixBy present frame b2The point cloud data of animal switchs to reference frame b in frame1The point cloud of animal in frame coordinate system uses
ICP (Iterative Closest Point iteration closest approach) algorithm is by being transformed into b1Two frame animals in frame coordinate system
Point cloud iteration, willAs the initial value of above-mentioned ICP iteration, which can make two frame animal point cloud fast convergences, calculate
Transition matrix of the two frame animal point clouds under b1 frame referential out
6. self-movement robot animal identification according to claim 1 and hiding method, it is characterised in that:
The self-movement robot has depth camera and IMU, and the depth camera is for acquiring around self-movement robot
Environmental information, obtains RGB figure and depth map, and the IMU is used to obtain the angular speed and acceleration of self-movement robot.
7. self-movement robot animal identification according to claim 1 and hiding method, it is characterised in that:
The self-movement robot circular flow step S110-S140, the self-movement robot circular flow step S110-
S140 obtains next frame animal RGB figure, depth map, identifies animal, calculatesTwo frame animal point clouds are calculated in b1
Transition matrix under frame referentialMake post exercise coordinate system and b1The transition matrix of frame referential isDriving
Self-movement robot movement makes self-movement robot keep the position orientation relation kept constant with animal.
8. a kind of storage medium, for storing computer executable instructions, it is characterised in that:
Computer executable instructions perform claim when being executed by processor requires described in any one of 1-7 from mobile
Robot animal identification and hide method.
9. a kind of self-movement robot, with storage medium according to any one of claims 8, it is characterised in that:
The storage medium perform claim requires self-movement robot animal identification and the side of hiding described in any one of 1-7
Method.
10. a kind of self-movement robot, it is characterised in that:
The self-movement robot has depth camera and IMU, and described in any one of being able to carry out claim 1-7 from
Mobile robot animal identification and hide method.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910342589.6A CN110175523B (en) | 2019-04-26 | 2019-04-26 | Self-moving robot animal identification and avoidance method and storage medium thereof |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910342589.6A CN110175523B (en) | 2019-04-26 | 2019-04-26 | Self-moving robot animal identification and avoidance method and storage medium thereof |
Publications (2)
Publication Number | Publication Date |
---|---|
CN110175523A true CN110175523A (en) | 2019-08-27 |
CN110175523B CN110175523B (en) | 2021-05-14 |
Family
ID=67690149
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201910342589.6A Active CN110175523B (en) | 2019-04-26 | 2019-04-26 | Self-moving robot animal identification and avoidance method and storage medium thereof |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN110175523B (en) |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN112884838A (en) * | 2021-03-16 | 2021-06-01 | 重庆大学 | Robot autonomous positioning method |
CN113470591A (en) * | 2020-03-31 | 2021-10-01 | 京东方科技集团股份有限公司 | Monitor toning method and device, electronic equipment and storage medium |
Citations (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN105137973A (en) * | 2015-08-21 | 2015-12-09 | 华南理工大学 | Method for robot to intelligently avoid human under man-machine cooperation scene |
EP3007025A1 (en) * | 2014-10-10 | 2016-04-13 | LG Electronics Inc. | Robot cleaner and method for controlling the same |
CN106096559A (en) * | 2016-06-16 | 2016-11-09 | 深圳零度智能机器人科技有限公司 | Obstacle detection method and system and moving object |
CN107995962A (en) * | 2017-11-02 | 2018-05-04 | 深圳市道通智能航空技术有限公司 | A kind of barrier-avoiding method, device, loose impediment and computer-readable recording medium |
US20180278820A1 (en) * | 2014-07-01 | 2018-09-27 | Brain Corporation | Optical detection apparatus and methods |
CN108805906A (en) * | 2018-05-25 | 2018-11-13 | 哈尔滨工业大学 | A kind of moving obstacle detection and localization method based on depth map |
CN108958263A (en) * | 2018-08-03 | 2018-12-07 | 江苏木盟智能科技有限公司 | A kind of Obstacle Avoidance and robot |
CN109461185A (en) * | 2018-09-10 | 2019-03-12 | 西北工业大学 | A kind of robot target automatic obstacle avoidance method suitable for complex scene |
-
2019
- 2019-04-26 CN CN201910342589.6A patent/CN110175523B/en active Active
Patent Citations (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20180278820A1 (en) * | 2014-07-01 | 2018-09-27 | Brain Corporation | Optical detection apparatus and methods |
EP3007025A1 (en) * | 2014-10-10 | 2016-04-13 | LG Electronics Inc. | Robot cleaner and method for controlling the same |
CN105137973A (en) * | 2015-08-21 | 2015-12-09 | 华南理工大学 | Method for robot to intelligently avoid human under man-machine cooperation scene |
CN106096559A (en) * | 2016-06-16 | 2016-11-09 | 深圳零度智能机器人科技有限公司 | Obstacle detection method and system and moving object |
CN107995962A (en) * | 2017-11-02 | 2018-05-04 | 深圳市道通智能航空技术有限公司 | A kind of barrier-avoiding method, device, loose impediment and computer-readable recording medium |
CN108805906A (en) * | 2018-05-25 | 2018-11-13 | 哈尔滨工业大学 | A kind of moving obstacle detection and localization method based on depth map |
CN108958263A (en) * | 2018-08-03 | 2018-12-07 | 江苏木盟智能科技有限公司 | A kind of Obstacle Avoidance and robot |
CN109461185A (en) * | 2018-09-10 | 2019-03-12 | 西北工业大学 | A kind of robot target automatic obstacle avoidance method suitable for complex scene |
Non-Patent Citations (4)
Title |
---|
MIKAEL SVENSTRUP ET AL.: "Pose Estimation and Adaptive Robot Behaviour for Human-Robot Interaction", 《2009 IEEE INTERNATIONAL CONFERENCE ON ROBOTICS AND AUTOMATION》 * |
吴国盛 等: "一种基于极坐标系下的机器人动态避碰算法", 《2006中国控制与决策学术年会论文集》 * |
吴康 等: "一种基于模糊识别的移动机器人避障算法", 《东南大学学报(自然科学版)》 * |
张毅 等: "基于深度图像的移动机器人动态避障算 法", 《控制工程》 * |
Cited By (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN113470591A (en) * | 2020-03-31 | 2021-10-01 | 京东方科技集团股份有限公司 | Monitor toning method and device, electronic equipment and storage medium |
CN113470591B (en) * | 2020-03-31 | 2023-11-14 | 京东方科技集团股份有限公司 | Monitor color matching method and device, electronic equipment and storage medium |
CN112884838A (en) * | 2021-03-16 | 2021-06-01 | 重庆大学 | Robot autonomous positioning method |
CN112884838B (en) * | 2021-03-16 | 2022-11-15 | 重庆大学 | Robot autonomous positioning method |
Also Published As
Publication number | Publication date |
---|---|
CN110175523B (en) | 2021-05-14 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN109544636B (en) | Rapid monocular vision odometer navigation positioning method integrating feature point method and direct method | |
EP3698323B1 (en) | Depth from motion for augmented reality for handheld user devices | |
US10769411B2 (en) | Pose estimation and model retrieval for objects in images | |
CN112767373B (en) | Robot indoor complex scene obstacle avoidance method based on monocular camera | |
Biswas et al. | Gesture recognition using microsoft kinect® | |
US11064178B2 (en) | Deep virtual stereo odometry | |
CN111462135A (en) | Semantic mapping method based on visual S L AM and two-dimensional semantic segmentation | |
WO2019031083A1 (en) | Method and system for detecting action | |
US20190301871A1 (en) | Direct Sparse Visual-Inertial Odometry Using Dynamic Marginalization | |
CN104794737B (en) | A kind of depth information Auxiliary Particle Filter tracking | |
WO2015126443A1 (en) | Moving object localization in 3d using a single camera | |
WO2021098765A1 (en) | Key frame selection method and apparatus based on motion state | |
JP2022546643A (en) | Image processing system and method for landmark position estimation with uncertainty | |
CN105930790A (en) | Human body behavior recognition method based on kernel sparse coding | |
CN103914855A (en) | Moving object positioning method and system | |
CN110175523A (en) | A kind of self-movement robot animal identification and hide method and its storage medium | |
CN114004883B (en) | Visual perception method and device for curling ball, computer equipment and storage medium | |
CN111798373A (en) | Rapid unmanned aerial vehicle image stitching method based on local plane hypothesis and six-degree-of-freedom pose optimization | |
CN112581540A (en) | Camera calibration method based on human body posture estimation in large scene | |
CN112488067B (en) | Face pose estimation method and device, electronic equipment and storage medium | |
CN109725699A (en) | Recognition methods, device and the equipment of identification code | |
CN110520813A (en) | It is mobile throughout the multiple agent confrontation type of label formation using the transformation of RADON cumulative distribution and canonical correlation analysis prediction | |
US11238604B1 (en) | Densifying sparse depth maps | |
WO2022228391A1 (en) | Terminal device positioning method and related device therefor | |
CN110377033B (en) | RGBD information-based small football robot identification and tracking grabbing 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 | ||
GR01 | Patent grant | ||
GR01 | Patent grant |