WO2019109228A1 - 基于视觉重定位的扫地机断电续扫方法、装置及扫地机 - Google Patents
基于视觉重定位的扫地机断电续扫方法、装置及扫地机 Download PDFInfo
- Publication number
- WO2019109228A1 WO2019109228A1 PCT/CN2017/114505 CN2017114505W WO2019109228A1 WO 2019109228 A1 WO2019109228 A1 WO 2019109228A1 CN 2017114505 W CN2017114505 W CN 2017114505W WO 2019109228 A1 WO2019109228 A1 WO 2019109228A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- scene image
- power
- sweeper
- scene
- sweeping
- Prior art date
Links
- 241001417527 Pempheridae Species 0.000 title claims abstract description 157
- 238000010408 sweeping Methods 0.000 title claims abstract description 123
- 238000000034 method Methods 0.000 title claims abstract description 70
- 230000000007 visual effect Effects 0.000 title claims abstract description 36
- 238000004140 cleaning Methods 0.000 claims description 103
- 238000010586 diagram Methods 0.000 description 7
- 239000000284 extract Substances 0.000 description 3
- 230000009286 beneficial effect Effects 0.000 description 2
- 230000007812 deficiency Effects 0.000 description 1
- 230000009466 transformation Effects 0.000 description 1
- 238000000844 transformation Methods 0.000 description 1
Classifications
-
- A—HUMAN NECESSITIES
- A47—FURNITURE; DOMESTIC ARTICLES OR APPLIANCES; COFFEE MILLS; SPICE MILLS; SUCTION CLEANERS IN GENERAL
- A47L—DOMESTIC WASHING OR CLEANING; SUCTION CLEANERS IN GENERAL
- A47L11/00—Machines for cleaning floors, carpets, furniture, walls, or wall coverings
- A47L11/24—Floor-sweeping machines, motor-driven
-
- A—HUMAN NECESSITIES
- A47—FURNITURE; DOMESTIC ARTICLES OR APPLIANCES; COFFEE MILLS; SPICE MILLS; SUCTION CLEANERS IN GENERAL
- A47L—DOMESTIC WASHING OR CLEANING; SUCTION CLEANERS IN GENERAL
- A47L11/00—Machines for cleaning floors, carpets, furniture, walls, or wall coverings
- A47L11/40—Parts or details of machines not provided for in groups A47L11/02 - A47L11/38, or not restricted to one of these groups, e.g. handles, arrangements of switches, skirts, buffers, levers
Definitions
- the invention relates to the technical field of a sweeping machine, in particular to a method, a device and a sweeping machine for sweeping and power saving of a sweeping machine based on visual repositioning.
- Sweeping machine is a sweeping robot. It is an indispensable cleaning helper in modern homes.
- the sweeper is a step-by-step replacement for manual cleaning due to its ease of use and economical durability.
- the sweeping robot When the sweeping robot is working, it senses the environment and its own state through sensors, thereby achieving autonomous movement in an environment with obstacles.
- the sweeping robot system When the sweeping robot system is turned off or powered off, when the position and posture of the sweeping robot change, the sweeping robot cannot be positioned to locate its position and posture after restarting the robot. It is necessary to manually move the sweeping robot to the initial position to restart the system. After the initial position and the self-position are obtained, the cleaning is performed again, so that the previously cleaned area is cleaned again, which reduces the cleaning efficiency.
- the technical problem to be solved by the present invention is to provide a method, a device and a sweeping machine for sweeping and power-off of a sweeping machine based on visual repositioning for the above-mentioned deficiencies in the prior art.
- the technical means adopted by the present invention to solve the technical problem is to provide a method for power-off continuous cleaning of a sweeper based on visual relocation, comprising:
- the current scene image is collected, and the scene image with the highest matching degree with the current scene image in the scene map is selected, and the relative position between the sweeper and the corresponding position of the scene image with the highest matching degree is calculated. ;
- saving the scene image collected during the sweeping process of the Sweeper and the scene image at the power-off position to the scene map include:
- a scene image acquired during the cleaning process and a feature point of the scene image at the power-off position and a landmark corresponding to the feature point are saved.
- saving the scene image collected during the sweeping process of the sweeper and the scene image at the power-off position to the scene map include:
- the sweeper posture of the scene image acquired during the cleaning process and the scene image at the power-off position is calculated and saved.
- the starting the cleaning according to the preset policy includes:
- Determining whether the corresponding position of the scene image with the highest matching degree is the power-off position, and if not, driving the sweeper to the power-off position according to the posture of the sweeper of the scene image with the highest matching degree Perform a continuous sweep.
- saving the scene image collected during the sweeping process of the Sweeper and the scene image at the power-off position to the scene map include:
- a scene image in which the number of feature points exceeds a preset value range is saved to the scene map.
- the selecting a scene image that has the highest degree of matching with the current scene image in the scene map, and calculating a relative position between the sweeper and the corresponding position of the scene image with the highest matching degree includes:
- the calculating the movement route according to the relative position comprises:
- the present invention also provides a sweeping machine power-off continuous sweeping device based on visual repositioning, comprising:
- the power-off saving unit is configured to save the scene image collected during the sweeping process of the sweeper and the scene image at the power-off position to the scene map when the sweeper is powered off;
- a power-on positioning unit configured to collect a current scene image after the power of the sweeping machine, select a scene image with the highest matching degree with the current scene image in the scene map, and calculate the highest matching degree between the sweeping machine and the current scene The relative position between the corresponding positions of the scene image;
- a positioning continuous scanning unit configured to calculate a moving route according to the relative position, and drive the sweeping machine to move to a corresponding position of the scene image with the highest matching degree according to the moving route and start cleaning according to a preset strategy.
- the power-off saving unit includes:
- the feature point processing module is configured to save a scene image acquired during the cleaning process and a feature point of the scene image at the power-off position and a landmark corresponding to the feature point.
- the power-off saving unit includes:
- the attitude processing module is configured to calculate and save a scene image acquired during the cleaning process and a sweeper posture of the scene image at the power-off position.
- the positioning continuous scanning unit includes:
- a scan control module configured to determine whether a corresponding position of the scene image with the highest matching degree is the power-off position, and if not, driving the sweeper according to the posture of the sweeper of the scene image with the highest matching degree Move to the power-off position for continuous sweeping.
- the power-off saving unit includes:
- the selection module is configured to select a scene image whose number of feature points exceeds a preset value range and save the image to the scene map.
- the positioning continuous scanning unit includes:
- a comparison module configured to compare feature points of the current scene image and the scene image in the scene map
- a matching module configured to select a frame image of the scene point in the scene map that matches a feature point of the current scene image
- a calculation module configured to calculate a relative position between the power-on position and the one-frame scene image according to a landmark in the one-frame scene image and a landmark of a power-on position of the sweeper.
- the positioning continuous scanning unit includes:
- a location calculation module configured to calculate a power-on position of the sweeper in the scene map and a corresponding cleaning position of the scene image with the highest matching degree in the scene map;
- a route generating module configured to calculate a relative position between the power-on position and the corresponding cleaning position, and calculate a moving route according to the relative position.
- the present invention also provides a sweeping machine comprising the above-described visual repositioning based sweeper power-off continuous sweeping device.
- the present invention has at least the following beneficial effects: the present invention saves the scene image collected during the sweeping process of the sweeper and the scene image at the power-off position to the scene map when the sweeper is powered off, and when the sweeper is powered on and restarted And acquiring a current scene image and selecting a scene image with the highest matching degree with the current scene image in the scene map, and then calculating a relative position between the sweeper and the corresponding position of the scene image with the highest matching degree, and further Calculating the movement route according to the relative position, driving the sweeping machine to move to the corresponding position of the scene image with the highest matching degree according to the movement route to continue cleaning, thereby implementing the power-off continuous scanning function, and preventing the sweeping machine from cleaning again from the beginning Improve cleaning efficiency.
- FIG. 1 is a schematic flow chart of an embodiment of a method for powering off and sweeping a sweeper based on visual repositioning according to the present invention.
- step S1 is a schematic diagram of a specific process of step S1 of an embodiment of a method for powering off and sweeping a sweeper according to the present invention.
- FIG. 3 is a schematic diagram of a specific process of step S1 of another embodiment of the method for powering off and sweeping the shovel according to the present invention.
- step S1 is a schematic diagram of a specific process of step S1 of another embodiment of the method according to the present invention.
- FIG. 5 is a schematic diagram of a specific process of step S2 of an embodiment of a method for powering off and sweeping a sweeper according to the present invention.
- FIG. 6 is a schematic diagram of a specific process of step S3 of an embodiment of a method for powering off and sweeping a sweeper according to the present invention.
- FIG. 7 is a schematic diagram of a specific process of step S3 of another embodiment of the method for powering off and sweeping the sweeper according to the present invention.
- FIG. 8 is a schematic structural view of an embodiment of a sweeping machine power-off continuous sweeping device based on visual repositioning according to the present invention.
- FIG. 9 is a schematic diagram showing the specific structure of an embodiment of a power sweep re-scanning device based on visual repositioning according to the present invention.
- the present invention provides a technical solution: a method for power-off continuous cleaning of a sweeper based on visual relocation, comprising:
- Step S1 when the sweeping machine is powered off, save the scene image collected during the sweeping process of the sweeper and the scene image at the power-off position to the scene map;
- Step S2 After the power is turned on, the current scene image is collected, and the scene image with the highest matching degree with the current scene image in the scene map is selected, and the corresponding position between the sweeper and the scene image with the highest matching degree is calculated. Relative position
- Step S3 Calculate the movement route according to the relative position, and drive the sweeper to move to the corresponding position of the scene image with the highest matching degree according to the movement route and start cleaning according to a preset strategy.
- the sweeper collects the scene image through the camera and saves it to the scene map during the cleaning process.
- the scene map stores a certain number of scene images.
- the sweeper When the sweeper is powered off, the sweeping opportunity saves the power-off position.
- the scene image is in the scene map.
- the sweeper may be moved to another location, for example, artificially carrying the sweeper to the charging pile for charging.
- the sweeper When the sweeper is powered on and restarted, the current scene image is collected by the camera, and then all the scene images in the scene map are traversed to select a scene image with the highest matching degree with the current scene image, and the sweeper can calculate the power of the sweeper.
- the sweeping position of the sweeping machine starts cleaning according to the preset strategy to realize the power-off continuous sweeping function, so as to prevent the sweeping machine from cleaning again from the beginning and improve the cleaning efficiency.
- saving the scene image collected during the sweeping process of the sweeper and the scene image at the power-off position to the scene map includes:
- Step S11 saving the scene image acquired during the cleaning process and the feature point of the scene image at the power-off position and the landmark corresponding to the feature point.
- the sweeper continuously collects the scene image through the camera during the cleaning process, and collects the scene image at the power-off position during the power-off, and extracts feature points, feature descriptors, and
- the landmark corresponding to the feature point is saved in the scene map, wherein the feature point is an object in an area to be cleaned, for example, an object such as a sofa, a coffee table, and a seat, and the feature descriptor is used for specific Positioning the object describing the feature point, for example, a teacup on a coffee table, a teapot, etc., by calculating the feature point and the feature descriptor, the landmark corresponding to the feature point can be obtained, the positioning function of the sweeper is realized, and the cleaning efficiency is improved.
- the scene image collected during the cleaning process of the sweeping machine and the scene image at the power-off position are included in the scene map:
- step S12 the scene image acquired during the cleaning process and the sweeper posture of the scene image at the power-off position are calculated and saved.
- the sweeper calculates the sweeper posture by visual positioning, gyroscope or code wheel, and the sweeper posture is used to determine information such as the direction of travel of the sweeper at the corresponding position and its shape, and the sweeper is
- the scene image is collected by the camera, and the feature points, the feature descriptors, the landmarks corresponding to the feature points, and the posture of the sweeper are calculated in the scene image, and then saved to the scene map, specifically,
- the feature points in each frame image are numbered to obtain a feature ID number
- the landmark is a coordinate of a pixel whose feature point is located on the scene image in the world coordinate system
- each scene image in the scene map is numbered.
- the sweeping machine can be positioned according to the scene map to plan a cleaning route, wherein the feature points and landmarks in the scene image are used Positioning the sweeper at the time of power failure or at a certain moment during the cleaning process, the posture of the sweeper is used to determine the direction of the sweeper and the state of the sweeper at a certain moment in the power-off position or during the cleaning process, avoiding the sweeper In the case of reverse cleaning, the cleaning efficiency is improved.
- saving the scene image collected during the sweeping process of the sweeper and the scene image at the power-off position to the scene map includes:
- step S13 the scene image whose number of feature points exceeds the preset value range is saved to the scene map.
- the sweeper continuously collects and saves the scene image through the camera during the cleaning process.
- the number of the scene images is relatively large, and the occupied storage space is large, and the image data of the scene whose number of feature points exceeds the preset value range is selected. Into the map data, can effectively reduce data redundancy.
- the scene image with the highest matching degree with the current scene image in the scene map is selected, and the sweeper is calculated to correspond to the scene image with the highest matching degree.
- the relative positions between locations include:
- Step S21 comparing feature points of the current scene image and the scene image in the scene map
- Step S22 selecting a frame scene image in which the feature points in the scene map match the feature points of the current scene image
- Step S23 Calculate a relative position between the one-frame scene image and the power-on position according to the landmark of the one-frame scene image and the landmark of the power-on position of the Sweeper.
- the sweeper collects the scene images and extracts the feature points in the scene image to perform training classification to obtain the words of the feature points and store them in the visual dictionary.
- the sweeper records the power-on position, and the sweeping machine collects the current position.
- the current scene image may be determined to match the one-frame scene image, according to the scene image of the one frame
- the image ID number can calculate the distribution of the scene image of the one frame in the scene map, so that the cleaning position corresponding to the image of the scene frame at a certain moment before the power-off of the sweeping machine can be calculated, thereby calculating the sweeping machine.
- the relative position between the power-on position and the cleaning position is a dictionary package obtained by offline training, and is obtained by acquiring a picture and extracting feature points in the picture.
- the visual dictionary may also be loaded by the sweeper from the network.
- the cleaning according to the preset policy includes:
- Step S31 determining whether the corresponding position of the scene image with the highest matching degree is the power-off position, and if not, driving the sweeper to move to the same according to the posture of the sweeper of the scene image with the highest matching degree
- the power-off position is continuously scanned.
- the sweeping machine moves to the cleaning position and determines whether the cleaning position is the same as the power-off position. If the same, the sweeping machine is located at the power-off position, and the sweeping machine is based on the power-off position in the scene map.
- the sweeper posture is adjusted until it is consistent with the sweeper posture of the power-off position, and then the sweeper starts to continue cleaning; if not, the sweeper is located at a cleaning position at a certain time in the area swept by the sweeper before the power-off,
- the posture of the sweeper at the cleaning position the posture of the sweeper is adjusted, and then cleaning is performed according to the cleaning route of the sweeper before the power-off from the cleaning position, thereby avoiding the cleaning from the beginning, and the cleaning efficiency can be effectively improved.
- the calculating the movement route according to the relative position comprises:
- Step S32 calculating a power-on position of the sweeper in the scene map and a corresponding cleaning position of the scene image with the highest matching degree in the scene map;
- Step S33 calculating a relative position between the power-on position and the corresponding cleaning position and calculating a movement route according to the relative position.
- the power-on position in the scene map is calculated, and then the screen is calculated according to the selected scene image in the highest scene map.
- the cleaning position at a certain moment before the machine is powered off the moving route is calculated according to the relative position between the power-on position and the cleaning position, and then the sweeping machine is driven to perform the obstacle-avoiding movement according to the moving route to optimize the walking of the sweeping machine. Route, reduce pathfinding time and improve cleaning efficiency.
- the present invention further provides a sweeping machine power-off continuous sweeping device based on visual repositioning, comprising:
- the power-off saving unit 1 is configured to save the scene image collected during the sweeping process of the sweeper and the scene image at the power-off position to the scene map when the sweeper is powered off;
- the power-on positioning unit 2 is configured to collect a current scene image after the power-on of the sweeping machine, and select a scene image with the highest matching degree with the current scene image in the scene map, and calculate that the sweeping machine has the highest matching degree.
- the scene image corresponds to the relative position between the positions;
- the continuation sweeping unit 3 is configured to calculate a moving route according to the relative position, and drive the sweeping machine to move to a corresponding position of the scene image with the highest matching degree according to the moving route and start cleaning according to a preset strategy.
- the power-off saving unit 1 saves the scene image collected by the sweeper during the cleaning process and the scene image at the power-off position to the scene map when the sweeper is powered off, and after the power-on and restart of the sweeper,
- the electric positioning unit 2 collects a current scene image and selects a scene image of the scene map that has the highest matching degree with the current scene image, and calculates a sweeping machine at a certain moment before the power failure according to the one-frame scene image.
- the cleaning position is further calculated by the positioning continuous scanning unit 3 to determine the cleaning position and the power-on position of the sweeping machine to determine the moving route of the sweeping machine, and the driving sweeper is moved to the cleaning position according to the moving route and the cleaning is started according to the preset strategy. , to achieve the power-off continuous sweep function, to avoid cleaning the cleaned area, improve cleaning efficiency.
- the power-off saving unit 1 includes:
- the feature point processing module 11 is configured to save a scene image acquired during the cleaning process and a feature point of the scene image at the power-off position and a landmark corresponding to the feature point.
- the sweeper continuously collects the scene image through the camera during the cleaning process, and collects the scene image at the power-off position during the power-off, and extracts the feature points in the scene image through the feature processing module 11
- the feature descriptor and the landmark corresponding to the feature point are saved in the scene map, and the sweeper can obtain the landmark corresponding to the feature point by calculating the feature point and the feature descriptor, thereby realizing the positioning function of the sweeper and improving Sweeper's power failure and continuous reliability.
- the power-off saving unit 1 further includes:
- the gesture processing module 12 is configured to calculate and save a scene image acquired during the cleaning process and a sweeper posture of the scene image at the power-off position.
- the sweeper calculates and saves the sweeper posture through the attitude processing module 12.
- the sweeper posture can be calculated by visual positioning, gyroscope or code wheel, and the sweeper posture is used to determine the sweeper.
- the sweeper collects the scene image through the camera during the cleaning process, and calculates feature points, feature descriptors, landmarks corresponding to the feature points, and sweeping the ground in the scene image.
- the machine pose is then saved to the scene map.
- the feature points in each frame image are numbered to obtain a feature ID number
- the landmark is a pixel whose feature point is located on the scene image in the world coordinate system.
- the sweeper can be based on the scene map Positioning to plan the cleaning route, wherein the feature points and landmarks in the scene image are used to locate the position of the sweeper when the power is turned off or at a certain time during the cleaning process, and the posture of the sweeper is used to determine whether the sweeper is in the power-off position or during the cleaning process.
- the direction of the sweeper and the state of the sweeper at a certain moment avoid the situation that the sweeper is in the reverse sweep or the cleaning direction is wrong, and the cleaning efficiency is improved.
- the power-off saving unit 1 includes:
- the selection module 13 is configured to select a scene image whose number of feature points exceeds a preset value range and save the image to the scene map.
- the sweeper continuously collects and saves the scene image through the camera during the cleaning process.
- the number of the scene images is relatively large, and the occupied storage space is large.
- the number of feature points is selected by the selection module 13 to exceed the preset.
- the scene image data of the value range is stored in the map data, which can effectively reduce data redundancy.
- the power-on positioning unit 2 includes:
- a comparison module 21 configured to compare feature points of the current scene image and the scene image in the scene map
- the matching module 22 is configured to select a frame image of the scene point in the scene map that matches a feature point of the current scene image
- the calculating module 23 is configured to calculate a relative position between the power-on position and the one-frame scene image according to a landmark in the one-frame scene image and a landmark of a power-on position of the Sweeper.
- the sweeping machine collects the feature image by collecting the scene image and extracting the feature points in the scene image, and the word of the feature point is stored in the visual dictionary.
- the sweeping opportunity collects the current scene image and calculates the current scene image.
- the feature points in the visual dictionary are searched for by the comparison module 21, and finally the distribution of the feature points of the current scene image in the visual dictionary is obtained, and then the matching module 22 is used according to the scene.
- the calculation module The distribution of the image of the one-frame scene in the scene map can be calculated according to the image ID number of the image of the scene image that matches the current scene image, so that the sweeper can be calculated at a certain moment before the power-off The cleaning position, thereby calculating the relative position between the power-on position of the sweeper and the cleaning position.
- the positioning and resuming unit 3 comprises:
- the continuation control module 31 is configured to determine whether the corresponding position of the scene image with the highest matching degree is the power-off position, and if not, drive the sweeping according to the posture of the sweeper of the scene image with the highest matching degree The machine moves to the power-off position for continuous scanning.
- the sweeping machine moves to the cleaning position and determines whether the cleaning position is the same as the power-off position by the continuous-scan control module 31. If the same, the sweeping machine is located at the power-off position, and the sweeping machine according to the scene The sweeper posture of the power-off position in the map is adjusted until it is consistent with the sweeper posture of the power-off position, and then the sweeper starts to continue cleaning; if not, the sweeper is located in the area swept by the sweeper before the power-off. At the cleaning position of the time, the sweeper posture of the cleaning position is extracted to adjust the posture of the sweeper, and then the sweeping position of the sweeper is cleaned from the cleaning position to avoid cleaning from the beginning. Effectively improve cleaning efficiency.
- the positioning and resuming unit 3 comprises:
- a position calculation module 32 configured to calculate a power-on position of the sweeper in the scene map and a corresponding cleaning position of the scene image with the highest matching degree in the scene map;
- the route generating module 33 is configured to calculate a relative position between the power-on position and the corresponding cleaning position and calculate a moving route according to the relative position.
- the power-on position in the scene map is calculated, and then calculated according to the selected one scene image in the highest scene map matching the current scene image.
- the sweeping machine calculates a relative position between the power-on position and the cleaning position at a cleaning position at a certain time before the power-off, calculates a moving route according to the relative position, and then drives the sweeper to perform obstacle avoidance according to the moving route. Move, optimize the walking route of the sweeper, reduce the pathfinding time, and improve efficiency.
- the present invention further provides a sweeping machine, comprising the above-mentioned visual repositioning based sweeping machine power-off continuous sweeping device;
- the sweeping machine of the embodiment includes the above-mentioned visual repositioning-based sweeping machine power-off continuous scanning device, which can automatically return to the cleaning position in the cleaned area before the power-off of the sweeping machine after the power-on restart to continue cleaning the unfinished cleaning.
- the area realizes the power-off continuous sweep function and improves the cleaning efficiency.
Landscapes
- Control Of Position, Course, Altitude, Or Attitude Of Moving Bodies (AREA)
Abstract
本发明涉及一种基于视觉重定位的扫地机断电续扫方法、装置及扫地机,所述方法包括:扫地机断电时,保存扫地机清扫过程中采集的场景图像和断电位置处的场景图像至场景地图中;上电后,采集当前场景图像,选取场景地图中与当前场景图像匹配度最高的场景图像,计算出扫地机与匹配度最高的场景图像对应位置之间的相对位置;根据相对位置计算移动路线,驱动扫地机根据移动路线移动至匹配度最高的场景图像的对应位置并按照预设策略开始清扫。本发明实施例能驱动扫地机重启后移动至场景地图中与当前场景图像匹配度最高的场景图像的对应位置并按照预设策略继续清扫,实现断电续扫功能,避免对清扫过的区域进行再次清扫,提高清扫效率。
Description
本发明涉及扫地机技术领域,特别涉及一种基于视觉重定位的扫地机断电续扫方法、装置及扫地机。
扫地机是指扫地机器人,是现代家庭中必不可少的清洁帮手,扫地机由于使用方便且经济耐用的特定逐步的取代人工清洁。扫地机器人在工作时是通过传感器感知环境和自身状态,进而实现在有障碍物的环境中自主运动。
在扫地机器人系统关闭或者断电的情况下,当扫地机器人的位置和姿态发生变化时,扫地机器人重新启动后无法定位其所在地图位置和自身姿态,需要人为将扫地机器人移动到初始位置重新启动系统获取初始位置和自身姿态后重新进行清扫,这样,会再次清扫之前已经清扫过的区域,降低了清扫效率。
发明内容
本发明要解决的技术问题在于针对上述现有技术中的不足之处,提供一种基于视觉重定位的扫地机断电续扫方法、装置及扫地机。
本发明解决技术问题采用的技术手段是提供一种基于视觉重定位的扫地机断电续扫方法,包括:
扫地机断电时,保存扫地机清扫过程中采集的场景图像和断电位置处的场景图像至场景地图中;
上电后,采集当前场景图像,选取所述场景地图中与所述当前场景图像匹配度最高的场景图像,计算出所述扫地机与所述匹配度最高的场景图像对应位置之间的相对位置;
根据所述相对位置计算移动路线,驱动扫地机根据所述移动路线移动至所述匹配度最高的场景图像的对应位置并按照预设策略开始清扫。
进一步地,所述扫地机断电时,保存扫地机清扫过程中采集的场景图像和断电位置处的场景图像至场景地图中包括:
保存清扫过程中采集的场景图像及所述断电位置处的场景图像的特征点及与所述特征点对应的地标。
进一步地,所述保存扫地机清扫过程中采集的场景图像和断电位置处的场景图像至场景地图中包括:
计算并保存清扫过程中采集的场景图像及所述断电位置处的场景图像的扫地机姿态。
进一步地,所述按照预设策略开始清扫包括:
判断所述匹配度最高的场景图像的对应位置是否为所述断电位置,若否,则根据所述所述匹配度最高的场景图像的扫地机的姿态驱动扫地机移动至所述断电位置进行续扫。
进一步地,所述扫地机断电时,保存扫地机清扫过程中采集的场景图像和断电位置处的场景图像至场景地图中包括:
保存特征点数量超出预设值范围的场景图像至所述场景地图中。
进一步地,所述选取所述场景地图中与所述当前场景图像匹配度最高的场景图像,计算出所述扫地机与该匹配度最高的场景图像对应位置之间的相对位置包括:
对比所述当前场景图像与所述场景地图中的场景图像的特征点;
选取所述场景地图中特征点与所述当前场景图像的特征点相匹配的一帧场景图像;
根据所述一帧场景图像的地标与扫地机的上电位置的地标计算出所述一帧场景图像与所述上电位置之间的相对位置。
进一步地,所述根据所述相对位置计算移动路线包括:
计算出所述扫地机在所述场景地图中的上电位置和所述匹配度最高的场景图像在所述场景地图中的对应清扫位置;
计算出所述上电位置和所述对应清扫位置之间相对位置并根据所述相对位置计算出移动路线。
另一方面,本发明还提供一种基于视觉重定位的扫地机断电续扫装置,包括:
断电保存单元,用于在扫地机断电时,保存扫地机清扫过程中采集的场景图像和断电位置处的场景图像至场景地图中;
上电定位单元,用于在扫地机上电后,采集当前场景图像,选取所述场景地图中与所述当前场景图像匹配度最高的场景图像,计算出所述扫地机与所述匹配度最高的场景图像对应位置之间的相对位置;
定位续扫单元,用于根据所述相对位置计算移动路线,驱动扫地机根据所述移动路线移动至所述匹配度最高的场景图像的对应位置并按照预设策略开始清扫。
进一步地,所述断电保存单元包括:
特征点处理模块,用于保存清扫过程中采集的场景图像及所述断电位置处的场景图像的特征点及与所述特征点对应的地标。
进一步地,所述断电保存单元包括:
姿态处理模块,用于计算并保存清扫过程中采集的场景图像及所述断电位置处的场景图像的扫地机姿态。
进一步地,所述定位续扫单元包括:
续扫控制模块,用于判断所述匹配度最高的场景图像的对应位置是否为所述断电位置,若否,则根据所述所述匹配度最高的场景图像的扫地机的姿态驱动扫地机移动至所述断电位置进行续扫。
进一步地,所述断电保存单元包括:
选择模块,用于选取特征点数量超出预设值范围的场景图像并保存至所述场景地图中。
进一步地,所述定位续扫单元包括:
比较模块,用于对比所述当前场景图像与所述场景地图中的场景图像的特征点;
匹配模块,用于选取所述场景地图中特征点与所述当前场景图像的特征点相匹配的一帧场景图像;
计算模块,用于根据所述一帧场景图像中的地标和扫地机的上电位置的地标计算出所述上电位置与所述一帧场景图像之间的相对位置。
进一步地,所述定位续扫单元包括:
位置计算模块,用于计算出所述扫地机在所述场景地图中的上电位置和所述匹配度最高的场景图像在所述场景地图中的对应清扫位置;
路线生成模块,用于计算出所述上电位置和所述对应清扫位置之间相对位置并根据所述相对位置计算出移动路线。
再一方面,本发明还提供一种扫地机,所述扫地机包括上述的基于视觉重定位的扫地机断电续扫装置。
采用上述技术方案,本发明至少具有以下有益效果:本发明通过扫地机断电时保存扫地机清扫过程中采集的场景图像和断电位置处的场景图像至场景地图中,当扫地机上电重启后,通过采集当前场景图像并选取所述场景地图中与所述当前场景图像匹配度最高的场景图像,然后计算出扫地机与所述匹配度最高的场景图像的对应位置之间的相对位置,进而根据所述相对位置计算出移动路线,驱动扫地机根据所述移动路线移动至所述匹配度最高的场景图像的对应位置以继续进行清扫,实现断电续扫功能,避免扫地机从头开始再次清扫,提高清扫效率。
图1是本发明基于视觉重定位的扫地机断电续扫方法一个实施例的流程示意图。
图2是本发明基于视觉重定位的扫地机断电续扫方法一个实施例步骤S1的具体流程示意图。
图3是本发明基于视觉重定位的扫地机断电续扫方法另一个实施例步骤S1的具体流程示意图。
图4是本发明基于视觉重定位的扫地机断电续扫方法再一个实施例步骤S1的具体流程示意图。
图5是本发明基于视觉重定位的扫地机断电续扫方法一个实施例步骤S2的具体流程示意图。
图6是本发明基于视觉重定位的扫地机断电续扫方法一个实施例步骤S3的具体流程示意图。
图7是本发明基于视觉重定位的扫地机断电续扫方法另一个实施例步骤S3的具体流程示意图。
图8是本发明基于视觉重定位的扫地机断电续扫装置一个实施例的结构示意图。
图9是本发明基于视觉重定位的扫地机断电续扫装置一个实施例的具体结构示意图。
本发明目的的实现、功能特点及优点将结合实施例,参照附图做进一步说明。
具体实施方式
下面将结合本发明实施例中的附图,对本发明实施例中的技术方案进行清楚、完整地描述,显然,所描述的实施例仅仅是本发明一部分实施例,而不是全部的实施例。基于本发明中的实施例,本领域普通技术人员在没有做出创造性劳动前提下所获得的所有其他实施例,都属于本发明保护的范围。
请参阅图1至图7所示,本发明提供一种技术方案:一种基于视觉重定位的扫地机断电续扫方法,包括:
步骤S1,扫地机断电时,保存扫地机清扫过程中采集的场景图像和断电位置处的场景图像至场景地图中;
步骤S2,上电后,采集当前场景图像,选取所述场景地图中与所述当前场景图像匹配度最高的场景图像,计算出所述扫地机与所述匹配度最高的场景图像对应位置之间的相对位置;
步骤S3,根据所述相对位置计算移动路线,驱动扫地机根据所述移动路线移动至所述匹配度最高的场景图像的对应位置并按照预设策略开始清扫。
在实施时,扫地机在清扫过程中会通过摄像头采集场景图像并保存至场景地图中,所述场景地图中保存着一定数量的场景图像,当扫地机断电时,扫地机会保存断电位置处的场景图像至场景地图中,此时,扫地机有可能被移动到其它的位置,例如:人为的搬运扫地机到充电桩进行充电。当扫地机上电重启时,会通过摄像头采集当前场景图像,然后遍历场景地图中所有的场景图像从而选出与所述当前场景图像匹配度最高的一帧场景图像,扫地机可以计算出扫地机上电时和所述一帧场景图像在所述场景地图中的对应位置之间的相对位置,再根据所述相对位置计算扫地机的移动路线并驱动扫地机进行避障移动,使扫地机移动至断电前某一时刻扫地机的清扫位置处并按照预设策略开始清扫,实现断电续扫功能,避免扫地机从头开始再次清扫,提高清扫效率。
在一个实施例中,如图2所示,所述扫地机断电时,保存扫地机清扫过程中采集的场景图像和断电位置处的场景图像至场景地图中包括:
步骤S11,保存清扫过程中采集的场景图像及所述断电位置处的场景图像的特征点及与所述特征点对应的地标。
在具体实施时,扫地机在清扫过程中通过摄像头持续采集场景图像,并在断电时采集及所述断电位置处的场景图像,提取所述场景图像中的特征点、特征描述子及与所述特征点对应的地标保存至所述场景地图中,其中,所述特征点为需要清扫的区域内的物体,例如:沙发、茶几以及座椅等物体,所述特征描述子为用于具体定位描述所述特征点的物体,例如:茶几上的茶杯、茶壶等,通过计算所述特征点和特征描述子可以得到所述特征点对应的地标,实现扫地机的定位功能,提高清扫效率。
在一个实施例中,如图3所示,所述保存扫地机清扫过程中采集的场景图像和断电位置处的场景图像至场景地图中包括:
步骤S12,计算并保存清扫过程中采集的场景图像及所述断电位置处的场景图像的扫地机姿态。
在实施时,扫地机通过视觉定位、陀螺仪或者码盘等计算得出扫地机姿态,所述扫地机姿态用于确定扫地机在对应位置的行进方向和自身形态等信息,所述扫地机在清扫过程中会通过摄像头采集场景图像,计算所述场景图像中的特征点、特征描述子、与所述特征点对应的地标以及扫地机姿态,然后保存至所述场景地图中,具体地,对每一帧场景图像中的特征点进行编号得到特征ID号,所述地标为特征点位于场景图像上的像素在世界坐标系下的坐标,再对所述场景地图中每一帧场景图像进行编号得到图像ID号,所述场景地图包括每一帧场景图像的图像ID号、特征点、特征描述子、与所述特征点对应的地标、对应的扫地机姿态以及特征点对应场景图像数据的索引,扫地机可以根据所述场景地图进行定位以规划清扫路线,其中场景图像中的特征点和地标用于定位扫地机断电时或者清扫过程中某一时刻的位置,扫地机的姿态用于确定扫地机在断电位置或者清扫过程中某一时刻的扫地机的行进方向和自身的状态,避免扫地机出现反向清扫的情况,提高清扫效率。
在一个可选实施例中,如图4所示,所述扫地机断电时,保存扫地机清扫过程中采集的场景图像和断电位置处的场景图像至场景地图中包括:
步骤S13,保存特征点数量超出预设值范围的场景图像至所述场景地图中。
在实施时,扫地机在清扫过程中通过摄像头持续采集场景图像并保存,所述场景图像的数量比较多,占用的存储空间大,而通过选取特征点数量超出预设值范围的场景图像数据存入地图数据中,能有效减少数据冗余。
在一个可选实施例中,如图5所示,所述选取所述场景地图中与所述当前场景图像匹配度最高的场景图像,计算出所述扫地机与该匹配度最高的场景图像对应位置之间的相对位置包括:
步骤S21,对比所述当前场景图像与所述场景地图中的场景图像的特征点;
步骤S22,选取所述场景地图中特征点与所述当前场景图像的特征点相匹配的一帧场景图像;
步骤S23,根据所述一帧场景图像的地标与扫地机的上电位置的地标计算出所述一帧场景图像与所述上电位置之间的相对位置。
在实施时,扫地机通过采集场景图像并提取场景图像中的特征点进行训练归类得到特征点的单词存入视觉词典中,当扫地机工作时,扫地机记录上电位置,扫地机会采集当前场景图像并计算当前场景图像中的特征点,然后在所述视觉词典中寻找对应的特征点的单词,最终得到所述当前场景图像的特征点在视觉词典中的分布,当所述当前场景图像中的特征点与场景地图中的一帧场景图像的特征点在视觉词典中的分布类似时,可以确定所述当前场景图像与所述一帧场景图像相匹配,根据所述一帧场景图像的图像ID号可以计算出所述一帧场景图像在所述场景地图中的分布,从而可以计算得到扫地机在断电前某一时刻对应所述一帧场景图像的清扫位置,从而计算出扫地机的上电位置和所述清扫位置之间的相对位置。在具体实施时,所述视觉词典是离线训练得到的一个词典包,通过采集图片并提取图片中的特征点训练得到的,当然所述视觉词典还可以是扫地机从网络中加载得到的。
在一个可选实施例中,如图6所示,所述按照预设策略开始清扫包括:
步骤S31,判断所述匹配度最高的场景图像的对应位置是否为所述断电位置,若否,则根据所述所述匹配度最高的场景图像的扫地机的姿态驱动扫地机移动至所述断电位置进行续扫。
在具体实施时,扫地机移动到所述清扫位置并判断所述清扫位置是否与所述断电位置相同,若相同,则扫地机刚好位于断电位置,扫地机根据场景地图中的断电位置的扫地机姿态进行调整直到与断电位置的扫地机姿态一致,然后扫地机开始继续清扫;若不相同,则扫地机位于断电前扫地机清扫过的区域中某一时刻的清扫位置上,通过提取所述清扫位置的扫地机姿态以调整扫地机的姿态,然后从所述清扫位置开始按照断电前扫地机的清扫路线进行清扫,避免从头开始进行再次清扫,能有效提高清扫效率。
在一个实施例中,如图7所示,所述根据所述相对位置计算移动路线包括:
步骤S32,计算出所述扫地机在所述场景地图中的上电位置和所述匹配度最高的场景图像在所述场景地图中的对应清扫位置;
步骤S33,计算出所述上电位置和所述对应清扫位置之间相对位置并根据所述相对位置计算出移动路线。
在具体实施时,扫地机上电重启时,计算出自身在所述场景地图中的上电位置,然后根据选取的与所述当前场景图像匹配对最高的场景地图中的一帧场景图像计算出扫地机断电前某一时刻的清扫位置,根据所述上电位置和清扫位置之间的相对位置计算出移动路线,然后驱动扫地机根据所述移动路线进行避障移动,优化优化扫地机的行走路线,减少寻路时间,提高清扫效率。
另一方面,如图8至图9所示,本发明还提供一种基于视觉重定位的扫地机断电续扫装置,包括:
断电保存单元1,用于在扫地机断电时,保存扫地机清扫过程中采集的场景图像和断电位置处的场景图像至场景地图中;
上电定位单元2,用于在扫地机上电后,采集当前场景图像,选取所述场景地图中与所述当前场景图像匹配度最高的场景图像,计算出所述扫地机与所述匹配度最高的场景图像对应位置之间的相对位置;
定位续扫单元3,用于根据所述相对位置计算移动路线,驱动扫地机根据所述移动路线移动至所述匹配度最高的场景图像的对应位置并按照预设策略开始清扫。
本实施例通过断电保存单元1在扫地机断电时将扫地机在清扫过程中采集的场景图像和断电位置处的场景图像保存至场景地图中,并在扫地机上电重启后,通过上电定位单元2采集当前场景图像并选出所述场景地图中与所述当前场景图像匹配度最高的一帧场景图像,根据所述一帧场景图像计算出扫地机在断电前某一时刻的清扫位置,再通过定位续扫单元3计算所述清扫位置和扫地机的上电位置以确定扫地机的移动路线,驱动扫地机按照所述移动路线移动至清扫位置并按照预设策略开始继续清扫,实现断电续扫功能,避免对清扫过的区域进行再次清扫,提高清扫效率。
在一个可选实施例中,所述断电保存单元1包括:
特征点处理模块11,用于保存清扫过程中采集的场景图像及所述断电位置处的场景图像的特征点及与所述特征点对应的地标。
在具体实施时,扫地机在清扫过程中通过摄像头持续采集场景图像,并在断电时采集及所述断电位置处的场景图像,通过特征处理模块11提取所述场景图像中的特征点、特征描述子及与所述特征点对应的地标保存至所述场景地图中,扫地机通过计算所述特征点和特征描述子可以得到所述特征点对应的地标,实现扫地机的定位功能,提高扫地机的断电续扫可靠性。
在一个可选实施例中,所述断电保存单元1还包括:
姿态处理模块12,用于计算并保存清扫过程中采集的场景图像及所述断电位置处的场景图像的扫地机姿态。
在实施时,扫地机通过姿态处理模块12计算并保存扫地机姿态,具体地,可以通过视觉定位、陀螺仪或者码盘等计算得出扫地机姿态,所述扫地机姿态用于确定扫地机在对应位置的行进方向和自身形态等信息,所述扫地机在清扫过程中会通过摄像头采集场景图像,计算所述场景图像中的特征点、特征描述子、与所述特征点对应的地标以及扫地机姿态,然后保存至所述场景地图中,具体地,对每一帧场景图像中的特征点进行编号得到特征ID号,所述地标为特征点位于场景图像上的像素在世界坐标系下的坐标,再对所述场景地图中每一帧场景图像进行编号得到图像ID号,所述场景地图包括每一帧场景图像的图像ID号、特征点、特征描述子、与所述特征点对应的地标、对应的扫地机姿态以及特征点对应场景图像数据的索引,扫地机可以根据所述场景地图进行定位以规划清扫路线,其中场景图像中的特征点和地标用于定位扫地机断电时或者清扫过程中某一时刻的位置,扫地机的姿态用于确定扫地机在断电位置或者清扫过程中某一时刻的扫地机的行进方向和自身的状态,避免扫地机出现反向清扫或者清扫方向错误的情况,提高清扫效率。
在一个可选实施例中,所述断电保存单元1包括:
选择模块13,用于选取特征点数量超出预设值范围的场景图像并保存至所述场景地图中。
在实施时,扫地机在清扫过程中通过摄像头持续采集场景图像并保存,所述场景图像的数量比较多,占用的存储空间大,本实施例通过设有选择模块13选取特征点数量超出预设值范围的场景图像数据存入地图数据中,能有效减少数据冗余。
在一个可选实施例中,所述上电定位单元2包括:
比较模块21,用于对比所述当前场景图像与所述场景地图中的场景图像的特征点;
匹配模块22,用于选取所述场景地图中特征点与所述当前场景图像的特征点相匹配的一帧场景图像;
计算模块23,用于根据所述一帧场景图像中的地标和扫地机的上电位置的地标计算出所述上电位置与所述一帧场景图像之间的相对位置。
具体地,扫地机通过采集场景图像并提取场景图像中的特征点进行训练归类得到特征点的单词存入视觉词典中,当扫地机上电工作时,扫地机会采集当前场景图像并计算当前场景图像中的特征点,然后通过比较模块21在所述视觉词典中寻找对应的特征点的单词,最终得到所述当前场景图像的特征点在视觉词典中的分布,再通过匹配模块22根据所述场景图像中的特征点与场景地图中的一帧场景图像的特征点在视觉词典中的分布类似时,确定所采集的当前场景图像与所述场景地图中的一帧场景图像相匹配,然后计算模块23根据与所述当前场景图像相匹配的一帧场景图像的图像ID号可以计算出所述一帧场景图像在所述场景地图中的分布,从而可以计算得到扫地机在断电前某一时刻的清扫位置,从而计算出扫地机的上电位置和所述清扫位置之间的相对位置。
在一个可选实施例中,所述定位续扫单元3包括:
续扫控制模块31,用于判断所述匹配度最高的场景图像的对应位置是否为所述断电位置,若否,则根据所述所述匹配度最高的场景图像的扫地机的姿态驱动扫地机移动至所述断电位置进行续扫。
在具体实施时,扫地机移动到所述清扫位置并通过续扫控制模块31判断所述清扫位置是否与所述断电位置相同,若相同,则扫地机刚好位于断电位置,扫地机根据场景地图中的断电位置的扫地机姿态进行调整直到与断电位置的扫地机姿态一致,然后扫地机开始继续清扫;若不相同,则扫地机位于断电前扫地机清扫过的区域中某一时刻的清扫位置上,通过提取所述清扫位置的扫地机姿态以调整扫地机的姿态,然后从所述清扫位置开始按照断电前扫地机的清扫路线进行清扫,避免从头开始进行再次清扫,能有效提高清扫效率。
在一个可选实施例中,所述定位续扫单元3包括:
位置计算模块32,用于计算出所述扫地机在所述场景地图中的上电位置和所述匹配度最高的场景图像在所述场景地图中的对应清扫位置;
路线生成模块33,用于计算出所述上电位置和所述对应清扫位置之间相对位置并根据所述相对位置计算出移动路线。
在具体实施时,扫地机上电重启时,计算出自身在所述场景地图中的上电位置,然后根据选取的与所述当前场景图像匹配对最高的场景地图中的一帧场景图像,计算出扫地机在断电前某一时刻的清扫位置,计算所述上电位置和清扫位置之间的相对位置并根据所述相对位置计算出移动路线,然后驱动扫地机根据所述移动路线进行避障移动,优化扫地机的行走路线,减少寻路时间,提高效率。
再一方面,本发明还提供一种扫地机,所述扫地机包括上述的基于视觉重定位的扫地机断电续扫装置;
本实施例的扫地机包括上述的基于视觉重定位的扫地机断电续扫装置,能在上电重启后自主回到扫地机断电前清扫过的区域中的清扫位置以继续清扫未完成清扫的区域,实现断电续扫功能,提高清扫效率。
以上所述仅为本发明的优选实施例,并非因此限制本发明的专利范围,凡是利用本发明说明书及附图内容所作的等效结构或等效流程变换,或直接或间接运用在其他相关的技术领域,均同理包括在本发明的专利保护范围内。
Claims (15)
- 一种基于视觉重定位的扫地机断电续扫方法,其特征在于,包括:扫地机断电时,保存扫地机清扫过程中采集的场景图像和断电位置处的场景图像至场景地图中;上电后,采集当前场景图像,选取所述场景地图中与所述当前场景图像匹配度最高的场景图像,计算出所述扫地机与所述匹配度最高的场景图像对应位置之间的相对位置;根据所述相对位置计算移动路线,驱动扫地机根据所述移动路线移动至所述匹配度最高的场景图像的对应位置并按照预设策略开始清扫。
- 如权利要求1所述的基于视觉重定位的扫地机断电续扫方法,其特征在于,所述扫地机断电时,保存扫地机清扫过程中采集的场景图像和断电位置处的场景图像至场景地图中包括:保存清扫过程中采集的场景图像及所述断电位置处的场景图像的特征点及与所述特征点对应的地标。
- 如权利要求1或2所述的基于视觉重定位的扫地机断电续扫方法,其特征在于:所述保存扫地机清扫过程中采集的场景图像和断电位置处的场景图像至场景地图中包括:计算并保存清扫过程中采集的场景图像及所述断电位置处的场景图像的扫地机姿态。
- 根据权利要求3所述的基于视觉重定位的扫地机断电续扫方法,其特征在于,所述按照预设策略开始清扫包括:判断所述匹配度最高的场景图像的对应位置是否为所述断电位置,若否,则根据所述所述匹配度最高的场景图像的扫地机的姿态驱动扫地机移动至所述断电位置进行续扫。
- 根据权利要求1所述的基于视觉重定位的扫地机断电续扫方法,其特征在于,所述扫地机断电时,保存扫地机清扫过程中采集的场景图像和断电位置处的场景图像至场景地图中包括:保存特征点数量超出预设值范围的场景图像至所述场景地图中。
- 根据权利要求2所述的基于视觉重定位的扫地机断电续扫方法,其特征在于,所述选取所述场景地图中与所述当前场景图像匹配度最高的场景图像,计算出所述扫地机与该匹配度最高的场景图像对应位置之间的相对位置包括:对比所述当前场景图像与所述场景地图中的场景图像的特征点;选取所述场景地图中特征点与所述当前场景图像的特征点相匹配的一帧场景图像;根据所述一帧场景图像的地标与扫地机的上电位置的地标计算出所述一帧场景图像与所述上电位置之间的相对位置。
- 根据权利要求1所述的基于视觉重定位的扫地机断电续扫方法,其特征在于,所述根据所述相对位置计算移动路线包括:计算出所述扫地机在所述场景地图中的上电位置和所述匹配度最高的场景图像在所述场景地图中的对应清扫位置;计算出所述上电位置和所述对应清扫位置之间相对位置并根据所述相对位置计算出移动路线。
- 一种基于视觉重定位的扫地机断电续扫装置,其特征在于,包括:断电保存单元,用于在扫地机断电时,保存扫地机清扫过程中采集的场景图像和断电位置处的场景图像至场景地图中;上电定位单元,用于在扫地机上电后,采集当前场景图像,选取所述场景地图中与所述当前场景图像匹配度最高的场景图像,计算出所述扫地机与所述匹配度最高的场景图像对应位置之间的相对位置;定位续扫单元,用于根据所述相对位置计算移动路线,驱动扫地机根据所述移动路线移动至所述匹配度最高的场景图像的对应位置并按照预设策略开始清扫。
- 根据权利要求8所述的基于视觉重定位的扫地机断电续扫装置,其特征在于,所述断电保存单元包括:特征点处理模块,用于保存清扫过程中采集的场景图像及所述断电位置处的场景图像的特征点及与所述特征点对应的地标。
- 根据权利要求8或9所述的基于视觉重定位的扫地机断电续扫装置,其特征在于,所述断电保存单元包括:姿态处理模块,用于计算并保存清扫过程中采集的场景图像及所述断电位置处的场景图像的扫地机姿态。
- 根据权利要求10所述的基于视觉重定位的扫地机断电续扫装置,其特征在于,所述定位续扫单元包括:续扫控制模块,用于判断所述匹配度最高的场景图像的对应位置是否为所述断电位置,若否,则根据所述所述匹配度最高的场景图像的扫地机的姿态驱动扫地机移动至所述断电位置进行续扫。
- 根据权利要求8所述的基于视觉重定位的扫地机断电续扫装置,其特征在于,所述断电保存单元包括:选择模块,用于选取特征点数量超出预设值范围的场景图像并保存至所述场景地图中。
- 根据权利要求9所述的基于视觉重定位的扫地机断电续扫装置,其特征在于,所述定位续扫单元包括:比较模块,用于对比所述当前场景图像与所述场景地图中的场景图像的特征点;匹配模块,用于选取所述场景地图中特征点与所述当前场景图像的特征点相匹配的一帧场景图像;计算模块,用于根据所述一帧场景图像中的地标和扫地机的上电位置的地标计算出所述上电位置与所述一帧场景图像之间的相对位置。
- 根据权利要求8所述的基于视觉重定位的扫地机断电续扫装置,其特征在于,所述定位续扫单元包括:位置计算模块,用于计算出所述扫地机在所述场景地图中的上电位置和所述匹配度最高的场景图像在所述场景地图中的对应清扫位置;路线生成模块,用于计算出所述上电位置和所述对应清扫位置之间相对位置并根据所述相对位置计算出移动路线。
- 一种扫地机,其特征在于:所述扫地机包括权利要求8所述的基于视觉重定位的扫地机断电续扫装置。
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
PCT/CN2017/114505 WO2019109228A1 (zh) | 2017-12-04 | 2017-12-04 | 基于视觉重定位的扫地机断电续扫方法、装置及扫地机 |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
PCT/CN2017/114505 WO2019109228A1 (zh) | 2017-12-04 | 2017-12-04 | 基于视觉重定位的扫地机断电续扫方法、装置及扫地机 |
Publications (1)
Publication Number | Publication Date |
---|---|
WO2019109228A1 true WO2019109228A1 (zh) | 2019-06-13 |
Family
ID=66750012
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
PCT/CN2017/114505 WO2019109228A1 (zh) | 2017-12-04 | 2017-12-04 | 基于视觉重定位的扫地机断电续扫方法、装置及扫地机 |
Country Status (1)
Country | Link |
---|---|
WO (1) | WO2019109228A1 (zh) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN114397928A (zh) * | 2021-12-31 | 2022-04-26 | 长沙金湘水泥制品有限公司 | 一种智能电杆养护系统及方法 |
Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101920498A (zh) * | 2009-06-16 | 2010-12-22 | 泰怡凯电器(苏州)有限公司 | 实现室内服务机器人同时定位和地图创建的装置及机器人 |
US20160144505A1 (en) * | 2014-11-26 | 2016-05-26 | Irobot Corporation | Systems and Methods for Performing Occlusion Detection |
CN105844260A (zh) * | 2016-04-14 | 2016-08-10 | 吴本刚 | 一种多功能智能清洁机器人装置 |
CN106537186A (zh) * | 2014-11-26 | 2017-03-22 | 艾罗伯特公司 | 用于使用机器视觉系统执行同时定位和映射的系统和方法 |
CN106569489A (zh) * | 2015-10-13 | 2017-04-19 | 录可系统公司 | 具有视觉导航功能的扫地机器人及其导航方法 |
CN107427177A (zh) * | 2015-02-13 | 2017-12-01 | 三星电子株式会社 | 清洁机器人及其控制方法 |
CN108021884A (zh) * | 2017-12-04 | 2018-05-11 | 深圳市沃特沃德股份有限公司 | 基于视觉重定位的扫地机断电续扫方法、装置及扫地机 |
-
2017
- 2017-12-04 WO PCT/CN2017/114505 patent/WO2019109228A1/zh active Application Filing
Patent Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101920498A (zh) * | 2009-06-16 | 2010-12-22 | 泰怡凯电器(苏州)有限公司 | 实现室内服务机器人同时定位和地图创建的装置及机器人 |
US20160144505A1 (en) * | 2014-11-26 | 2016-05-26 | Irobot Corporation | Systems and Methods for Performing Occlusion Detection |
CN106537186A (zh) * | 2014-11-26 | 2017-03-22 | 艾罗伯特公司 | 用于使用机器视觉系统执行同时定位和映射的系统和方法 |
CN107427177A (zh) * | 2015-02-13 | 2017-12-01 | 三星电子株式会社 | 清洁机器人及其控制方法 |
CN106569489A (zh) * | 2015-10-13 | 2017-04-19 | 录可系统公司 | 具有视觉导航功能的扫地机器人及其导航方法 |
CN105844260A (zh) * | 2016-04-14 | 2016-08-10 | 吴本刚 | 一种多功能智能清洁机器人装置 |
CN108021884A (zh) * | 2017-12-04 | 2018-05-11 | 深圳市沃特沃德股份有限公司 | 基于视觉重定位的扫地机断电续扫方法、装置及扫地机 |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN114397928A (zh) * | 2021-12-31 | 2022-04-26 | 长沙金湘水泥制品有限公司 | 一种智能电杆养护系统及方法 |
CN114397928B (zh) * | 2021-12-31 | 2023-04-11 | 长沙金湘水泥制品有限公司 | 一种智能电杆养护系统及方法 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN108021884B (zh) | 基于视觉重定位的扫地机断电续扫方法、装置及扫地机 | |
CN109074083B (zh) | 移动控制方法、移动机器人及计算机存储介质 | |
CN105302131B (zh) | 借助可自主运行的设备清洁或处理空间的方法 | |
US20200054185A1 (en) | Cleaning robot traveling using region-based human activity data and method of driving cleaning robot | |
WO2016050215A1 (zh) | 一种自移动表面行走机器人系统及回归主充电座的方法 | |
US7920941B2 (en) | Dust detection method and apparatus for cleaning robot | |
US20190254490A1 (en) | Vacuum cleaner and travel control method thereof | |
JP5898022B2 (ja) | 自走式機器 | |
CN110325938B (zh) | 电动吸尘器 | |
CN109843139B (zh) | 电动吸尘器 | |
US20190227566A1 (en) | Self-propelled vacuum cleaner | |
CN110968081B (zh) | 具有可伸缩摄像头的扫地机器人的控制方法及控制装置 | |
CN108634886A (zh) | 机器人清扫中断后的控制方法及芯片 | |
US20190184569A1 (en) | Robot based on artificial intelligence, and control method thereof | |
CN110650667B (zh) | 电动吸尘器 | |
CN111328017B (zh) | 一种地图传输方法和装置 | |
WO2023025028A1 (zh) | 充电方法、充电装置及机器人 | |
WO2019232804A1 (zh) | 软件更新方法、系统、移动机器人及服务器 | |
CN105919517B (zh) | 自动清扫机器人装置 | |
JP2019107400A (ja) | 自走式電気掃除機 | |
CN111700544A (zh) | 一种智能清洁机器人的防碰撞控制系统及其控制方法 | |
WO2022222345A1 (zh) | 移动机器人的定位修正方法和装置、存储介质、电子装置 | |
WO2019109228A1 (zh) | 基于视觉重定位的扫地机断电续扫方法、装置及扫地机 | |
CN114347040B (zh) | 目标物的拾取方法、装置、机器人以及存储介质 | |
CN111656138A (zh) | 构建地图及定位方法、客户端、移动机器人及存储介质 |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
121 | Ep: the epo has been informed by wipo that ep was designated in this application |
Ref document number: 17934248 Country of ref document: EP Kind code of ref document: A1 |
|
NENP | Non-entry into the national phase |
Ref country code: DE |
|
122 | Ep: pct application non-entry in european phase |
Ref document number: 17934248 Country of ref document: EP Kind code of ref document: A1 |