WO2019062651A1 - 一种定位建图的方法及*** - Google Patents
一种定位建图的方法及*** Download PDFInfo
- Publication number
- WO2019062651A1 WO2019062651A1 PCT/CN2018/106812 CN2018106812W WO2019062651A1 WO 2019062651 A1 WO2019062651 A1 WO 2019062651A1 CN 2018106812 W CN2018106812 W CN 2018106812W WO 2019062651 A1 WO2019062651 A1 WO 2019062651A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- data
- mobile device
- pose
- map
- current
- Prior art date
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
Definitions
- the present application relates to the field of computers, and in particular, to a method and system for positioning a map.
- An object of the present application is to provide a method and system for positioning and constructing a map, which solves the problem that the conventional laser positioning and mapping system is difficult to globally relocate, difficult to navigate and avoid obstacles, and consumes huge resources.
- a system for locating a map comprising:
- a sensor module a drawing module, a map storage module, a navigation module, and a positioning module, wherein
- the sensor module is configured to collect odometer data, laser data and visual image data
- the mapping module is configured to perform mapping according to the odometer data, the laser data and the visual image data collected by the sensor, wherein the mapping includes generating a raster map of the current environment and a laser mapping of the current pose of the mobile device. And visual mapping that produces sparse three-dimensional maps;
- the map storage module is configured to store the raster map, the sparse three-dimensional map, and key frames of the image;
- the navigation module is configured to perform at least one of the following tasks according to the data collected by the sensor and the raster map: planning a global path, avoiding obstacles, and navigating tasks;
- the positioning module is configured to perform visual relocation according to the visual image data, a raster map and a key frame stored in the map storage module.
- the positioning module is configured to:
- the word bag data of the current image is extracted based on the relocation request, and the key frame corresponding to the historical word bag data whose similarity with the word bag data of the current image is within a preset range is retrieved through the dictionary tree.
- the positioning module is configured to:
- Performing feature matching based on the key frame to determine a pose of the current image in the world coordinate system determining the movement according to the pose of the current image in the world coordinate system and the installation pose of the image sensor in the mobile device The pose of the device in world coordinates.
- the drawing module is used to:
- the determined three-dimensional position is converted into word bag data and inserted into a corresponding node in the dictionary tree.
- the navigation module is configured to:
- the navigation module is configured to:
- a collision-free motion control decision is determined based on the current speed information of the mobile device and a dynamic window algorithm.
- a method for locating a map comprising:
- Performing feature matching based on the key frame to determine a pose of the current image in the world coordinate system determining the movement according to the pose of the current image in the world coordinate system and the installation pose of the image sensor in the mobile device The pose of the device in world coordinates to complete the relocation.
- the method includes:
- the converted word bag data is inserted into a corresponding node in the dictionary tree.
- determining a global optimal path from the start point to the end point according to the odometer data, the laser data, and the grid map to complete the navigation task of the mobile device including:
- a collision-free motion control decision is determined based on the current speed information of the mobile device and a dynamic window algorithm.
- a computer readable medium having stored thereon computer readable instructions executable by a processor to implement a method of locating a map.
- the present application determines a current map of the mobile device and a raster map of the current environment according to the odometer data and the laser data; determining the starting point according to the odometer data, the laser data, and the raster map.
- the global optimal path to the end point to complete the navigation task of the mobile device extract the word bag data of the current image based on the relocation request, and retrieve the historical word bag whose similarity with the current image word bag data is within the preset range through the dictionary tree Key frame corresponding to the data; performing feature matching based on the key frame to determine a pose of the current image in the world coordinate system, according to the pose of the current image in the world coordinate system and the installation position of the image sensor in the mobile device Position, determine the pose of the mobile device in world coordinates to complete the relocation.
- the laser and visual sensor information can be combined to enable the mobile device to perform mapping, positioning and navigation tasks in an unknown environment, and global relocation is easy to implement, which can be used for navigation and obstacle avoidance tasks of mobile devices, and reduces storage resources and Calculate the consumption of resources.
- FIG. 1 is a schematic structural diagram of a system for positioning and mapping according to an aspect of the present application
- FIG. 2 is a schematic flow chart of a method for positioning a drawing according to another aspect of the present application.
- the terminal, the device of the service network, and the trusted party each include one or more processors (CPUs), input/output interfaces, network interfaces, and memory.
- processors CPUs
- input/output interfaces network interfaces
- memory volatile and non-volatile memory
- the memory may include non-persistent memory, random access memory (RAM), and/or non-volatile memory in a computer readable medium, such as read only memory (ROM) or flash memory.
- RAM random access memory
- ROM read only memory
- Memory is an example of a computer readable medium.
- Computer readable media includes both permanent and non-persistent, removable and non-removable media.
- Information storage can be implemented by any method or technology.
- the information can be computer readable instructions, data structures, modules of programs, or other data.
- Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read only memory. (ROM), electrically erasable programmable read only memory (EEPROM), flash memory or other memory technology, compact disk read only memory (CD-ROM), digital versatile disk (DVD) or other optical storage,
- computer readable media does not include non-transitory computer readable media, such as modulated data signals and carrier waves.
- FIG. 1 is a schematic structural diagram of a system for positioning and mapping according to an aspect of the present application.
- the system includes: a sensor module 11, a mapping module 12, a map storage module 13, a navigation module 14, and a positioning module 15, wherein
- the sensor module 11 is configured to collect odometer data, laser data, and visual image data
- the mapping module 12 is configured to perform mapping according to odometer data, laser data, and visual image data collected by the sensor, wherein
- the map includes a laser map that generates a current map of the current environment and a current pose of the mobile device, and a visual map that produces a sparse three-dimensional map
- the map storage module 13 is configured to store the raster map, the sparse three-dimensional map, and a key frame of the image
- the navigation module 14 is configured to perform at least one of the following tasks according to the data collected by the sensor and the raster map: planning a global path, avoiding obstacles, and navigating tasks; Visual relocation is performed based on the visual image
- the sensor module 11 includes an odometer, a laser, and a vision sensor to provide a data foundation for the positioning and mapping system.
- the data collected by the odometer is the distance data of the device moved relative to the initial position by using the data obtained from the motion sensor to estimate the change of the position of the object with time.
- the mapping module 12 includes a laser mapping module and a visual mapping module, wherein the laser mapping module generates a current map of the current environment and a current pose of the mobile device, and the visual mapping module generates a sparse three-dimensional map for the mobile device.
- Global relocation task raster map divides the environment into a series of grids, each of which is given a possible value, indicating the probability that the grid is occupied.
- the laser mapping module obtains the odometer and laser data from the sensor, and uses the RBPF-SLAM (particle filter-based positioning and mapping) method to perform the positioning and mapping task, and obtains the current map of the mobile device and the current environment.
- the mapping module 12 the global path of the planned mobile device and the environment location information of the mobile device are determined, wherein the mobile device includes a mobile robot, such as a mobile sweeping device.
- the mapping module 12 is configured to extract a corner feature in the acquired visual image, and perform feature matching by using a violent search with a corner feature in a previous key frame to obtain a feature point; according to the current state of the mobile device
- the pose determines the three-dimensional position of the feature point in the world coordinate system, and inputs the determined three-dimensional position into the sparse three-dimensional map; converts the determined three-dimensional position into word bag data and inserts into the corresponding node in the dictionary tree .
- the ORB feature in the image output by the visual sensor is extracted, and the ORB feature of the previous key frame is matched by the violent search for feature matching, and the pose position output by the laser mapping module is used to calculate the feature point in the world by the triangulation method.
- ORB Oriented Fast and Rotated BRIEF
- the map storage module 13 includes a raster map storage module generated by the laser mapping and a sparse three-dimensional map module and a key frame storage module generated by the visual mapping.
- the grid map is mainly used for the navigation obstacle avoidance task of the mobile device, and is sparse.
- the 3D map module and key frame storage module are mainly used for global relocation tasks of mobile devices.
- the navigation module 14 mainly uses the data collected by the sensor and the information of the grid map to perform tasks such as destination path planning, obstacle avoidance, and navigation. When the mobile device loses its global location information due to some factors, the positioning module 15 is required to perform a global relocation task through the visual sensor, the sparse three-dimensional map, and the key frame storage module.
- the positioning module 15 is configured to: extract the word bag data of the current image based on the relocation request, and retrieve a history word bag whose similarity with the word bag data of the current image is within a preset range by using a dictionary tree.
- the key frame corresponding to the data When a global relocation task is required, the BOW (Bag Of Word) data of the current image is first extracted, and some key frames similar to the current BOW are retrieved by using the dictionary tree in the key frame storage module, for example, by setting a similarity threshold, according to The similarity threshold is judged whether it is similar, and the key frame for relocation is selected.
- BOW Bit Of Word
- the PNP algorithm can be used to calculate the pose of the current image in the world coordinate system.
- the pose of the mobile device in the world coordinate system is calculated and output to the laser.
- Diagram module for rebuilding and positioning Furthermore, relying on relatively inexpensive vision sensors, the problem of global relocation of the traditional two-dimensional laser positioning and mapping system is solved.
- the navigation module 14 is configured to: determine a global optimal path from a start point to an end point according to the odometer data, the laser data, and the grid map to complete a navigation task of the mobile device. Specifically, determining a global collision-free optimal path from a start point to an end point according to the odometer data, the laser data, and the grid map; determining a collision-free motion control decision according to the current speed information of the mobile device and a dynamic window algorithm .
- the heuristic search algorithm is used to search for the global collision-free optimal path from the start point to the end point.
- a smooth collision-free motion control decision is generated through the dynamic window algorithm, so that the specified navigation task is completed without collision.
- the positioning and mapping system described in the present application can combine laser and visual sensor information to enable the mobile device to perform mapping, positioning and navigation tasks in an unknown environment.
- Global relocation is easy to implement and can be used for navigation and obstacle avoidance tasks of mobile devices, and reduces the consumption of storage resources and computing resources.
- Step S11 determining a current bit of the mobile device according to the odometer data and the laser data.
- step S12 determining a global optimal path from the start point to the end point according to the odometer data, the laser data, and the grid map to complete the navigation task of the mobile device;
- step S13 extracting the word bag data of the current image based on the relocation request, and searching, by using the dictionary tree, a key frame corresponding to the history word bag data whose similarity with the word bag data of the current image is within a preset range;
- step S14 based on the Characterizing the key frame to determine the pose of the current image in the world coordinate system, determining the mobile device according to the pose of the current image in the world coordinate system and the installation pose of the image sensor in the mobile device Pose in world coordinates to complete relocation.
- a data foundation is provided for the positioning and mapping system.
- the data collected by the odometer is the distance data of the device moved relative to the initial position by using the data obtained from the motion sensor to estimate the change of the position of the object with time.
- the grid map of the current environment and the current pose of the mobile device and the sparse three-dimensional map are determined according to the odometer data and the laser data for the global relocation task of the mobile device.
- a raster map divides the environment into a series of rasters, each of which gives a possible value indicating the probability that the raster is occupied.
- the odometer and the laser data are obtained from the sensor, and the RBPF-SLAM (particle filter-based positioning and mapping) method is used to perform the positioning and mapping task, and the current map of the mobile device and the current environment are obtained.
- the method includes a step S11': extracting a corner feature in the acquired visual image, and performing feature matching by using a violent search with a corner feature in a previous key frame to obtain a feature point; Determining a current position of the mobile device to determine a three-dimensional position of the feature point in a world coordinate system, and inputting the determined three-dimensional position into the sparse three-dimensional map; converting the determined three-dimensional position into word bag data and inserting into the dictionary tree In the corresponding node.
- the ORB feature in the image output by the visual sensor is extracted, and the ORB feature of the previous key frame is matched by the violent search, and the feature point is calculated in the world coordinate system by using the pose determined in step S11.
- the next three-dimensional position determines a sparse three-dimensional map.
- ORB Oriented Fast and Rotated BRIEF
- ORB is an efficient image corner feature descriptor.
- the grid map is mainly used for the navigation obstacle avoidance task of the mobile device, and the sparse three-dimensional map and the key frame are mainly used for the global relocation task of the mobile device.
- steps S12 tasks such as destination path planning, obstacle avoidance, and navigation are performed using data collected by the sensor and information of the grid map.
- step S13 the word bag data of the current image is extracted based on the relocation request, and the key frame corresponding to the historical word bag data whose similarity with the word bag data of the current image is within the preset range is retrieved through the dictionary tree.
- BOW Bit Of Word
- step S14 feature matching is performed based on the key frame to determine a pose of the current image in the world coordinate system, according to the pose of the current image in the world coordinate system and the installation of the image sensor in the mobile device.
- the pose determines the pose of the mobile device in world coordinates to complete the relocation.
- the PNP algorithm can be used to calculate the pose of the current image in the world coordinate system, and at the same time, combined with the installation pose of the image sensor in the mobile device, the pose of the mobile device in the world coordinate system is calculated, and the laser mapping is performed. In order to rebuild the map and locate.
- the problem of global relocation of the traditional two-dimensional laser positioning and mapping system is solved.
- a global optimal path from a start point to an end point is determined according to the odometer data, the laser data, and the grid map to complete a navigation task of the mobile device. Specifically, determining a global collision-free optimal path from a start point to an end point according to the odometer data, the laser data, and the grid map; determining a collision-free motion control decision according to the current speed information of the mobile device and a dynamic window algorithm .
- the heuristic search algorithm is used to search for the global collision-free optimal path from the start point to the end point.
- the smooth windowless motion control decision is generated by the dynamic window algorithm, so that the specified navigation task is completed without collision.
- the laser and visual sensor information can be combined to enable the mobile device to perform mapping, positioning and navigation tasks in an unknown environment.
- Global relocation is easy to implement and can be used for navigation and obstacle avoidance tasks of mobile devices, and reduces the consumption of storage resources and computing resources.
- a computer readable medium having stored thereon computer readable instructions executable by a processor to implement a method of locating a map.
- computer readable instructions are executed by a processor:
- Performing feature matching based on the key frame to determine a pose of the current image in the world coordinate system determining the movement according to the pose of the current image in the world coordinate system and the installation pose of the image sensor in the mobile device The pose of the device in world coordinates to complete the relocation.
- the present application can be implemented in software and/or a combination of software and hardware, for example, using an application specific integrated circuit (ASIC), a general purpose computer, or any other similar hardware device.
- the software program of the present application can be executed by a processor to implement the steps or functions described above.
- the software programs (including related data structures) of the present application can be stored in a computer readable recording medium such as a RAM memory, a magnetic or optical drive or a floppy disk and the like.
- some of the steps or functions of the present application may be implemented in hardware, for example, as a circuit that cooperates with a processor to perform various steps or functions.
- a portion of the present application can be applied as a computer program product, such as computer program instructions, which, when executed by a computer, can invoke or provide a method and/or technical solution in accordance with the present application.
- the program instructions for invoking the method of the present application may be stored in a fixed or removable recording medium, and/or transmitted by a data stream in a broadcast or other signal bearing medium, and/or stored in a The working memory of the computer device in which the program instructions are run.
- an embodiment in accordance with the present application includes a device including a memory for storing computer program instructions and a processor for executing program instructions, wherein when the computer program instructions are executed by the processor, triggering
- the apparatus operates based on the aforementioned methods and/or technical solutions in accordance with various embodiments of the present application.
Landscapes
- Engineering & Computer Science (AREA)
- Radar, Positioning & Navigation (AREA)
- Remote Sensing (AREA)
- Automation & Control Theory (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Navigation (AREA)
- Traffic Control Systems (AREA)
- Control Of Position, Course, Altitude, Or Attitude Of Moving Bodies (AREA)
Abstract
一种定位建图的方法,根据里程计数据及激光数据确定移动设备的当前位姿和当前环境的栅格地图(S11);根据里程计数据、激光数据及栅格地图确定从起点到终点的全局最优路径,以完成移动设备的导航任务(S12);基于重定位请求提取当前图像的词袋数据,通过词典树检索与当前图像的词袋数据相似度在预设范围内的历史词袋数据对应的关键帧(S13);基于关键帧进行特征匹配以确定当前图像在世界坐标系下的位姿,根据当前图像在世界坐标系下的位姿及图像传感器在移动设备中的安装位姿,确定移动设备在世界坐标下的位姿,以完成重定位(S14)。还涉及一种定位建图的***。
Description
本申请涉及计算机领域,尤其涉及一种定位建图的方法及***。
随着计算机技术、微电子技术、网络技术等快速的发展,机器人技术也得到了飞速发展,除了工业机器人水平不断提高之外,各种用于服务行业的先进机器人***也有了长足的发展,各类移动服务机器人越来越普及,比如移动设备、导购机、自动柜员机等等。
目前,移动设备定位建图***的实现主要有以下两种方式:
第一,基于激光雷达传感器的2D定位建图***,这种***大部分使用栅格地图作为地图存储方法,但具有一个比较大的缺点和难点,全局重定位比较困难。
第二,基于视觉传感器的3D定位建图***,这种***大部分使用稀疏三维特征点作为地图格式,搭配上词袋模型(Bag Of Word)非常适合于全局重定位任务。然而该方案中,所建的稀疏三维特征点地图难以用于移动机器人导航和避障任务。虽然利用稠密三维点云地图也能够进行导航与避障,但是这种方法会消耗巨大存储资源和计算资源。
发明内容
本申请的一个目的是提供一种定位建图的方法及***,解决传统激光定位建图***全局重定位困难、难以导航和避障任务及消耗巨大资源的问题。
根据本申请的一个方面,提供了一种定位建图的***,该***包括:
传感器模块、建图模块、地图存储模块、导航模块及定位模块,其中,
所述传感器模块用于采集里程计数据、激光数据及视觉图像数据;
所述建图模块用于根据所述传感器采集的里程计数据、激光数据及视觉图像数据进行建图,其中,建图包括产生当前环境的栅格地图和移动设备的当前位姿的激光建图,以及产生稀疏三维地图的视觉建图;
所述地图存储模块用于存储所述栅格地图、稀疏三维地图及图像的关键帧;
所述导航模块用于根据所述传感器采集的数据及所述栅格地图执行以下至少任一项任务:规划全局路径、避障及导航任务;
所述定位模块用于根据所述视觉图像数据、所述地图存储模块中存储的栅格地图和关键帧进行视觉重定位。
进一步地,上述***中,所述定位模块用于:
基于重定位请求提取当前图像的词袋数据,通过词典树检索与当前图像的词袋数据相似度在预设范围内的历史词袋数据对应的关键帧。
进一步地,上述***中,所述定位模块用于:
基于所述关键帧进行特征匹配以确定当前图像在世界坐标系下的位姿,根据所述当前图像在世界坐标系下的位姿及图像传感器在移动设备中的安装位姿,确定所述移动设备在世界坐标下的位姿。
进一步地,上述***中,所述建图模块用于:
提取所获取的视觉图像中的角点特征,与上一关键帧中角点特征通过暴力搜索进行特征匹配,得到特征点;
根据所述移动设备的当前位姿确定所述特征点在世界坐标系下的三维位置,并将确定的三维位置输入至稀疏三维地图中;
将所确定的三维位置转换为词袋数据后***至词典树中对应的节点中。
进一步地,上述***中,所述导航模块用于:
根据所述里程计数据、激光数据及所述栅格地图确定从起点到终点的全局最优路径,以完成移动设备的导航任务。
进一步地,上述***中,所述导航模块用于:
根据所述里程计数据、激光数据及所述栅格地图确定从起点到终点的全局无碰撞最优路径;
根据所述移动设备的当前速度信息及动态窗口算法确定无碰撞运动控制决策。
根据本申请另一个方面,还提供了一种定位建图的方法,该方法包括:
根据里程计数据及激光数据确定移动设备的当前位姿和当前环境的栅格地图;
根据所述里程计数据、激光数据及所述栅格地图确定从起点到终点的全局最优路径,以完成移动设备的导航任务;
基于重定位请求提取当前图像的词袋数据,通过词典树检索与当前图像的词袋数据相似度在预设范围内的历史词袋数据对应的关键帧;
基于所述关键帧进行特征匹配以确定当前图像在世界坐标系下的位姿,根据所述当前图像在世界坐标系下的位姿及图像传感器在移动设备中的安装位姿,确定所述移动设备在世界坐标下的位姿,以完成重定位。
进一步地,根据里程计数据及激光数据确定移动设备的当前位姿和当前环境的栅格地图之后,包括:
提取所获取的视觉图像中的角点特征,与上一关键帧中角点特征通过暴力搜索进行特征匹配,得到特征点;
根据所述移动设备的当前位姿确定所述特征点在世界坐标系下的三维位置,并将所确定的三维位置转换为词袋数据;
将所述转换后的词袋数据***至词典树中对应的节点中。
进一步地,根据所述里程计数据、激光数据及所述栅格地图确定从起点到终点的全局最优路径,以完成移动设备的导航任务,包括:
根据所述里程计数据、激光数据及所述栅格地图确定从起点到终点的全局无碰撞最优路径;
根据所述移动设备的当前速度信息及动态窗口算法确定无碰撞运动控制 决策。
根据本申请再一个方面,还提供了一种计算机可读介质,其上存储有计算机可读指令,所述计算机可读指令可被处理器执行以实现一种定位建图的方法。
与现有技术相比,本申请通过根据里程计数据及激光数据确定移动设备的当前位姿和当前环境的栅格地图;根据所述里程计数据、激光数据及所述栅格地图确定从起点到终点的全局最优路径,以完成移动设备的导航任务;基于重定位请求提取当前图像的词袋数据,通过词典树检索与当前图像的词袋数据相似度在预设范围内的历史词袋数据对应的关键帧;基于所述关键帧进行特征匹配以确定当前图像在世界坐标系下的位姿,根据所述当前图像在世界坐标系下的位姿及图像传感器在移动设备中的安装位姿,确定所述移动设备在世界坐标下的位姿,以完成重定位。从而能够结合激光和视觉传感器信息,使得移动设备能够在未知环境下进行建图,定位和导航任务,全局重定位易实现,可以用于移动设备的导航和避障任务,并降低了存储资源和计算资源的消耗。
通过阅读参照以下附图所作的对非限制性实施例所作的详细描述,本申请的其它特征、目的和优点将会变得更明显:
图1示出根据本申请的一个方面提供的一种定位建图的***结构示意图;
图2示出根据本申请另一个方面提供的一种定位建图的方法流程示意图。
附图中相同或相似的附图标记代表相同或相似的部件。
下面结合附图对本申请作进一步详细描述。
在本申请一个典型的配置中,终端、服务网络的设备和可信方均包括一个或多个处理器(CPU)、输入/输出接口、网络接口和内存。
内存可能包括计算机可读介质中的非永久性存储器,随机存取存储器(RAM)和/或非易失性内存等形式,如只读存储器(ROM)或闪存(flash RAM)。内存是计算机可读介质的示例。
计算机可读介质包括永久性和非永久性、可移动和非可移动媒体可以由任何方法或技术来实现信息存储。信息可以是计算机可读指令、数据结构、程序的模块或其他数据。计算机的存储介质的例子包括,但不限于相变内存(PRAM)、静态随机存取存储器(SRAM)、动态随机存取存储器(DRAM)、其他类型的随机存取存储器(RAM)、只读存储器(ROM)、电可擦除可编程只读存储器(EEPROM)、快闪记忆体或其他内存技术、只读光盘只读存储器(CD-ROM)、数字多功能光盘(DVD)或其他光学存储、磁盒式磁带,磁带磁盘存储或其他磁性存储设备或任何其他非传输介质,可用于存储可以被计算设备访问的信息。按照本文中的界定,计算机可读介质不包括非暂存电脑可读媒体(transitory media),如调制的数据信号和载波。
图1示出根据本申请的一个方面提供的一种定位建图的***结构示意图,该***包括:传感器模块11、建图模块12、地图存储模块13、导航模块14及定位模块15,其中,所述传感器模块11用于采集里程计数据、激光数据及视觉图像数据;所述建图模块12用于根据所述传感器采集的里程计数据、激光数据及视觉图像数据进行建图,其中,建图包括产生当前环境的栅格地图和移动设备的当前位姿的激光建图,以及产生稀疏三维地图的视觉建图;所述地图存储模块13用于存储所述栅格地图、稀疏三维地图及图像的关键帧;所述导航模块14用于根据所述传感器采集的数据及所述栅格地图执行以下至少任一项任务:规划全局路径、避障及导航任务;所述定位模块15用 于根据所述视觉图像数据、所述地图存储模块中存储的栅格地图和关键帧进行视觉重定位。从而能够结合激光和视觉传感器信息,使得移动设备能够在未知环境下进行建图,定位和导航任务。
具体地,传感器模块11包括里程计、激光和视觉传感器,为定位建图***提供数据基础。其中,里程计采集的数据是利用从移动传感器获得的数据来估计物***置随时间变化而改变的设备相对于初始位置移动的距离数据。
建图模块12包括激光建图模块和视觉建图模块,其中,激光建图模块产生当前环境的栅格地图和移动设备的当前位姿,视觉建图模块产生稀疏三维地图,用于移动设备的全局重定位任务,栅格地图把环境划分成一系列栅格,其中每一栅格给定一个可能值,表示该栅格被占据的概率。激光建图模块从传感器中获取里程计与激光数据,利用RBPF-SLAM(基于粒子滤波器的定位建图)方法进行定位建图任务,得到移动设备当前位姿和当前环境的栅格地图。通过建图模块12,完成规划移动设备的全局路径及确定移动设备所在的环境位置信息,其中,移动设备包括移动机器人,如移动扫地设备。
进一步地,所述建图模块12用于提取所获取的视觉图像中的角点特征,与上一关键帧中角点特征通过暴力搜索进行特征匹配,得到特征点;根据所述移动设备的当前位姿确定所述特征点在世界坐标系下的三维位置,并将确定的三维位置输入至稀疏三维地图中;将所确定的三维位置转换为词袋数据后***至词典树中对应的节点中。在此,提取视觉传感器所输出图像中的ORB特征,与上一关键帧的ORB特征通过暴力搜索进行特征匹配,利用激光建图模块所输出的位姿进行三角法测距法算出特征点在世界坐标系下的三维位置,并输入到稀疏三维地图模块。全局重定位中只需用历史关键帧的BOW数据,因此,将当前输入图像转换成BOW数据,***至词典树中,以便通过词典树的数据结构快速检索相近BOW数据。在上述实施例中,ORB(Oriented Fast and Rotated BRIEF)为一种高效的图像角点特征描述子。
地图存储模块13,包括激光建图所产生的栅格地图存储模块和视觉建图 所生成的稀疏三维地图模块与关键帧存储模块,栅格地图主要用于移动设备的导航避障任务,而稀疏三维地图模块与关键帧存储模块主要用于移动设备的全局重定位任务。导航模块14主要利用传感器所采集的数据和栅格地图的信息进行目的地路径规划、避障及导航等任务。当移动设备因某些因素丢失其全局位置信息时,需要定位模块15通过视觉传感器、稀疏三维地图和关键帧存储模块进行全局重定位任务。
在本申请一实施例中,所述定位模块15用于:基于重定位请求提取当前图像的词袋数据,通过词典树检索与当前图像的词袋数据相似度在预设范围内的历史词袋数据对应的关键帧。当需要全局重定位任务时,首先提取当前图像的BOW(Bag Of Word)数据,在关键帧存储模块中利用词典树检索出与当前BOW相似的一些关键帧,如通过设定相似度阈值,根据相似度阈值判断是否相似,筛选出用于重定位的关键帧。接着,基于所述关键帧进行特征匹配以确定当前图像在世界坐标系下的位姿,根据所述当前图像在世界坐标系下的位姿及图像传感器在移动设备中的安装位姿,确定所述移动设备在世界坐标下的位姿。在此,可以利用PNP算法计算当前图像在世界坐标系下的位姿,同时结合图像传感器在移动设备中的安装位姿,计算出移动设备在世界坐标系下的位姿,并输出至激光建图模块,以便重新建图和定位。进而依靠比较廉价的视觉传感器,解决传统二维激光定位建图***全局重定位难题。
在本申请一实施例中,所述导航模块14用于:根据所述里程计数据、激光数据及所述栅格地图确定从起点到终点的全局最优路径,以完成移动设备的导航任务。具体地,根据所述里程计数据、激光数据及所述栅格地图确定从起点到终点的全局无碰撞最优路径;根据所述移动设备的当前速度信息及动态窗口算法确定无碰撞运动控制决策。在此,利用里程计、激光数据和当前环境栅格地图,利用启发式搜索算法,搜索起点到终点全局无碰撞最优路径。利用规划好的全局最优路径,结合移动设备当前速度信息,通过动态窗 口算法产生平滑的无碰撞运动控制决策,使得无碰撞完成指定的导航任务。
通过本申请所述定位建图的***,能够结合激光和视觉传感器信息,使得移动设备在未知环境下进行建图,定位和导航任务。全局重定位容易实现,可以用于移动设备的导航和避障任务,并降低了存储资源和计算资源的消耗。
图2示出根据本申请另一个方面提供的一种定位建图的方法流程示意图,该方法包括:步骤S11~步骤S14,在步骤S11中,根据里程计数据及激光数据确定移动设备的当前位姿和当前环境的栅格地图;在步骤S12中,根据所述里程计数据、激光数据及所述栅格地图确定从起点到终点的全局最优路径,以完成移动设备的导航任务;在步骤S13中,基于重定位请求提取当前图像的词袋数据,通过词典树检索与当前图像的词袋数据相似度在预设范围内的历史词袋数据对应的关键帧;在步骤S14中,基于所述关键帧进行特征匹配以确定当前图像在世界坐标系下的位姿,根据所述当前图像在世界坐标系下的位姿及图像传感器在移动设备中的安装位姿,确定所述移动设备在世界坐标下的位姿,以完成重定位。从而能够结合激光和视觉传感器信息,使得移动设备能够在未知环境下进行建图,定位和导航任务。
具体地,在步骤S11中,为定位建图***提供数据基础。其中,里程计采集的数据是利用从移动传感器获得的数据来估计物***置随时间变化而改变的设备相对于初始位置移动的距离数据。根据里程计数据及激光数据确定当前环境的栅格地图和移动设备的当前位姿以及稀疏三维地图,以便进行移动设备的全局重定位任务。栅格地图把环境划分成一系列栅格,其中每一栅格给定一个可能值,表示该栅格被占据的概率。具体地,从传感器中获取里程计与激光数据,利用RBPF-SLAM(基于粒子滤波器的定位建图)方法进行定位建图任务,得到移动设备当前位姿和当前环境的栅格地图。
进一步地,在步骤S11之后,所述方法包括步骤S11’:提取所获取的视觉图像中的角点特征,与上一关键帧中角点特征通过暴力搜索进行特征匹配,得到特征点;根据所述移动设备的当前位姿确定所述特征点在世界坐标系下 的三维位置,并将确定的三维位置输入至稀疏三维地图中;将所确定的三维位置转换为词袋数据后***至词典树中对应的节点中。在此,提取视觉传感器所输出图像中的ORB特征,与上一关键帧的ORB特征通过暴力搜索进行特征匹配,利用步骤S11中确定的位姿进行三角法测距法算出特征点在世界坐标系下的三维位置,进而确定稀疏三维地图。全局重定位中只需用历史关键帧的BOW数据,因此,将当前输入图像转换成BOW数据,***至词典树中,以便通过词典树的数据结构快速检索相近BOW数据。在上述实施例中,ORB(Oriented Fast and Rotated BRIEF)为一种高效的图像角点特征描述子。
在本申请一实施例中,栅格地图主要用于移动设备的导航避障任务,而稀疏三维地图与关键帧主要用于移动设备的全局重定位任务。在步骤S12中,利用传感器所采集的数据和栅格地图的信息进行目的地路径规划、避障及导航等任务。通过确定栅格地图、移动设备的当前位姿及稀疏三维地图,完成规划移动设备的全局路径及确定移动设备所在的环境位置信息,其中,移动设备包括移动机器人,如移动扫地设备。
在步骤S13中,基于重定位请求提取当前图像的词袋数据,通过词典树检索与当前图像的词袋数据相似度在预设范围内的历史词袋数据对应的关键帧。当需要全局重定位任务时,首先提取当前图像的BOW(Bag Of Word)数据,利用词典树检索出与当前BOW相似的一些关键帧,如通过设定相似度阈值,根据相似度阈值判断是否相似,筛选出用于重定位的关键帧。接着,在步骤S14中,基于所述关键帧进行特征匹配以确定当前图像在世界坐标系下的位姿,根据所述当前图像在世界坐标系下的位姿及图像传感器在移动设备中的安装位姿,确定所述移动设备在世界坐标下的位姿,以完成重定位。在此,可以利用PNP算法计算当前图像在世界坐标系下的位姿,同时结合图像传感器在移动设备中的安装位姿,计算出移动设备在世界坐标系下的位姿,进行激光建图,以便重新建图和定位。进而依靠比较廉价的视觉传感器,解决传统二维激光定位建图***全局重定位难题。
在本申请一实施例中,在步骤S12中,根据所述里程计数据、激光数据及所述栅格地图确定从起点到终点的全局最优路径,以完成移动设备的导航任务。具体地,根据所述里程计数据、激光数据及所述栅格地图确定从起点到终点的全局无碰撞最优路径;根据所述移动设备的当前速度信息及动态窗口算法确定无碰撞运动控制决策。在此,利用里程计、激光数据和当前环境栅格地图,利用启发式搜索算法,搜索起点到终点全局无碰撞最优路径。利用规划好的全局最优路径,结合移动设备当前速度信息,通过动态窗口算法产生平滑的无碰撞运动控制决策,使得无碰撞完成指定的导航任务。
通过本申请所述定位建图的方法,能够结合激光和视觉传感器信息,使得移动设备在未知环境下进行建图,定位和导航任务。全局重定位容易实现,可以用于移动设备的导航和避障任务,并降低了存储资源和计算资源的消耗。
根据本申请再一个方面,还提供了一种计算机可读介质,其上存储有计算机可读指令,所述计算机可读指令可被处理器执行以实现一种定位建图的方法。例如,计算机可读指令被处理器执行:
根据里程计数据及激光数据确定移动设备的当前位姿和当前环境的栅格地图;
根据所述里程计数据、激光数据及所述栅格地图确定从起点到终点的全局最优路径,以完成移动设备的导航任务;
基于重定位请求提取当前图像的词袋数据,通过词典树检索与当前图像的词袋数据相似度在预设范围内的历史词袋数据对应的关键帧;
基于所述关键帧进行特征匹配以确定当前图像在世界坐标系下的位姿,根据所述当前图像在世界坐标系下的位姿及图像传感器在移动设备中的安装位姿,确定所述移动设备在世界坐标下的位姿,以完成重定位。
显然,本领域的技术人员可以对本申请进行各种改动和变型而不脱离本申请的精神和范围。这样,倘若本申请的这些修改和变型属于本申请权利要求及其等同技术的范围之内,则本申请也意图包含这些改动和变型在 内。
需要注意的是,本申请可在软件和/或软件与硬件的组合体中被实施,例如,可采用专用集成电路(ASIC)、通用目的计算机或任何其他类似硬件设备来实现。在一个实施例中,本申请的软件程序可以通过处理器执行以实现上文所述步骤或功能。同样地,本申请的软件程序(包括相关的数据结构)可以被存储到计算机可读记录介质中,例如,RAM存储器,磁或光驱动器或软磁盘及类似设备。另外,本申请的一些步骤或功能可采用硬件来实现,例如,作为与处理器配合从而执行各个步骤或功能的电路。
另外,本申请的一部分可被应用为计算机程序产品,例如计算机程序指令,当其被计算机执行时,通过该计算机的操作,可以调用或提供根据本申请的方法和/或技术方案。而调用本申请的方法的程序指令,可能被存储在固定的或可移动的记录介质中,和/或通过广播或其他信号承载媒体中的数据流而被传输,和/或被存储在根据所述程序指令运行的计算机设备的工作存储器中。在此,根据本申请的一个实施例包括一个装置,该装置包括用于存储计算机程序指令的存储器和用于执行程序指令的处理器,其中,当该计算机程序指令被该处理器执行时,触发该装置运行基于前述根据本申请的多个实施例的方法和/或技术方案。
对于本领域技术人员而言,显然本申请不限于上述示范性实施例的细节,而且在不背离本申请的精神或基本特征的情况下,能够以其他的具体形式实现本申请。因此,无论从哪一点来看,均应将实施例看作是示范性的,而且是非限制性的,本申请的范围由所附权利要求而不是上述说明限定,因此旨在将落在权利要求的等同要件的含义和范围内的所有变化涵括在本申请内。不应将权利要求中的任何附图标记视为限制所涉及的权利要求。此外,显然“包括”一词不排除其他单元或步骤,单数不排除复数。装置权利要求中陈述的多个单元或装置也可以由一个单元或装置通过软件或者硬件来实现。第一,第二等词语用来表示名称,而并不表示任何特定 的顺序。
Claims (10)
- 一种定位建图的***,其中,所述***包括:传感器模块、建图模块、地图存储模块、导航模块及定位模块,其中,所述传感器模块用于采集里程计数据、激光数据及视觉图像数据;所述建图模块用于根据所述传感器采集的里程计数据、激光数据及视觉图像数据进行建图,其中,建图包括产生当前环境的栅格地图和移动设备的当前位姿的激光建图,以及产生稀疏三维地图的视觉建图;所述地图存储模块用于存储所述栅格地图、稀疏三维地图及图像的关键帧;所述导航模块用于根据所述传感器采集的数据及所述栅格地图执行以下至少任一项任务:规划全局路径、避障及导航任务;所述定位模块用于根据所述视觉图像数据、所述地图存储模块中存储的栅格地图和关键帧进行视觉重定位。
- 根据权利要求1所述的***,其中,所述定位模块用于:基于重定位请求提取当前图像的词袋数据,通过词典树检索与当前图像的词袋数据相似度在预设范围内的历史词袋数据对应的关键帧。
- 根据权利要求2所述的***,其中,所述定位模块用于:基于所述关键帧进行特征匹配以确定当前图像在世界坐标系下的位姿,根据所述当前图像在世界坐标系下的位姿及图像传感器在移动设备中的安装位姿,确定所述移动设备在世界坐标下的位姿。
- 根据权利要求1所述的***,其中,所述建图模块用于:提取所获取的视觉图像中的角点特征,与上一关键帧中角点特征通过暴力搜索进行特征匹配,得到特征点;根据所述移动设备的当前位姿确定所述特征点在世界坐标系下的三维位置,并将确定的三维位置输入至稀疏三维地图中;将所确定的三维位置转换为词袋数据后***至词典树中对应的节点中。
- 根据权利要求1所述的***,其中,所述导航模块用于:根据所述里程计数据、激光数据及所述栅格地图确定从起点到终点的全局最优路径,以完成移动设备的导航任务。
- 根据权利要求1所述的***,其中,所述导航模块用于:根据所述里程计数据、激光数据及所述栅格地图确定从起点到终点的全局无碰撞最优路径;根据所述移动设备的当前速度信息及动态窗口算法确定无碰撞运动控制决策。
- 一种定位建图的方法,其中,所述方法包括:根据里程计数据及激光数据确定移动设备的当前位姿和当前环境的栅格地图;根据所述里程计数据、激光数据及所述栅格地图确定从起点到终点的全局最优路径,以完成移动设备的导航任务;基于重定位请求提取当前图像的词袋数据,通过词典树检索与当前图像的词袋数据相似度在预设范围内的历史词袋数据对应的关键帧;基于所述关键帧进行特征匹配以确定当前图像在世界坐标系下的位姿,根据所述当前图像在世界坐标系下的位姿及图像传感器在移动设备中的安装位姿,确定所述移动设备在世界坐标下的位姿,以完成重定位。
- 根据权利要求7所述的方法,其中,根据里程计数据及激光数据确定 移动设备的当前位姿和当前环境的栅格地图之后,包括:提取所获取的视觉图像中的角点特征,与上一关键帧中角点特征通过暴力搜索进行特征匹配,得到特征点;根据所述移动设备的当前位姿确定所述特征点在世界坐标系下的三维位置,并将所确定的三维位置转换为词袋数据;将所述转换后的词袋数据***至词典树中对应的节点中。
- 根据权利要求8所述的方法,其中,根据所述里程计数据、激光数据及所述栅格地图确定从起点到终点的全局最优路径,以完成移动设备的导航任务,包括:根据所述里程计数据、激光数据及所述栅格地图确定从起点到终点的全局无碰撞最优路径;根据所述移动设备的当前速度信息及动态窗口算法确定无碰撞运动控制决策。
- 一种计算机可读介质,其上存储有计算机可读指令,所述计算机可读指令可被处理器执行以实现如权利要求7至9中任一项所述的方法。
Applications Claiming Priority (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710927028.3 | 2017-09-26 | ||
CN201710927028.3A CN107677279B (zh) | 2017-09-26 | 2017-09-26 | 一种定位建图的方法及*** |
Publications (1)
Publication Number | Publication Date |
---|---|
WO2019062651A1 true WO2019062651A1 (zh) | 2019-04-04 |
Family
ID=61138807
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
PCT/CN2018/106812 WO2019062651A1 (zh) | 2017-09-26 | 2018-09-20 | 一种定位建图的方法及*** |
Country Status (2)
Country | Link |
---|---|
CN (1) | CN107677279B (zh) |
WO (1) | WO2019062651A1 (zh) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2022012229A1 (zh) * | 2020-07-15 | 2022-01-20 | 京东科技信息技术有限公司 | 机器人的重定位方法及装置、存储介质、电子设备 |
Families Citing this family (28)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107677279B (zh) * | 2017-09-26 | 2020-04-24 | 上海思岚科技有限公司 | 一种定位建图的方法及*** |
CN110360999B (zh) * | 2018-03-26 | 2021-08-27 | 京东方科技集团股份有限公司 | 室内定位方法、室内定位***和计算机可读介质 |
CN108692720B (zh) | 2018-04-09 | 2021-01-22 | 京东方科技集团股份有限公司 | 定位方法、定位服务器及定位*** |
CN108717710B (zh) * | 2018-05-18 | 2022-04-22 | 京东方科技集团股份有限公司 | 室内环境下的定位方法、装置及*** |
CN110169741B (zh) * | 2018-06-25 | 2021-07-09 | 宁波洒哇地咔电器有限公司 | 一种清洁处理方法及设备 |
CN109141393B (zh) * | 2018-07-02 | 2020-12-08 | 北京百度网讯科技有限公司 | 重定位方法、设备及存储介质 |
CN110728711B (zh) * | 2018-07-17 | 2021-11-12 | 北京三快在线科技有限公司 | 定位建图方法及装置、定位方法、装置和*** |
US10953545B2 (en) * | 2018-08-13 | 2021-03-23 | Beijing Jingdong Shangke Information Technology Co., Ltd. | System and method for autonomous navigation using visual sparse map |
CN109087359B (zh) * | 2018-08-30 | 2020-12-08 | 杭州易现先进科技有限公司 | 位姿确定方法、位姿确定装置、介质和计算设备 |
CN109489660A (zh) * | 2018-10-09 | 2019-03-19 | 上海岚豹智能科技有限公司 | 机器人定位方法及设备 |
CN111368860B (zh) * | 2018-12-25 | 2024-04-26 | 深圳市优必选科技有限公司 | 重定位方法及终端设备 |
US10782137B2 (en) | 2019-01-28 | 2020-09-22 | Qfeeltech (Beijing) Co., Ltd. | Methods, apparatus, and systems for localization and mapping |
CN111489393B (zh) * | 2019-01-28 | 2023-06-02 | 速感科技(北京)有限公司 | Vslam方法、控制器和可移动设备 |
CN109992690B (zh) * | 2019-03-11 | 2021-04-13 | 中国华戎科技集团有限公司 | 一种图像检索方法和*** |
CN111829532B (zh) * | 2019-04-18 | 2022-05-17 | 丰翼科技(深圳)有限公司 | 一种飞行器重定位***和重定位方法 |
CN112393719B (zh) * | 2019-08-12 | 2023-04-14 | 科沃斯商用机器人有限公司 | 一种栅格语义地图生成方法、装置以及存储设备 |
CN110515382A (zh) * | 2019-08-28 | 2019-11-29 | 锐捷网络股份有限公司 | 一种智能设备及其定位方法 |
CN110866953B (zh) * | 2019-10-31 | 2023-12-29 | Oppo广东移动通信有限公司 | 地图构建方法及装置、定位方法及装置 |
CN111157005A (zh) * | 2020-01-07 | 2020-05-15 | 深圳市锐曼智能装备有限公司 | 基于反光片定位方法和装置 |
CN113465600B (zh) * | 2020-03-30 | 2024-05-28 | 浙江宇视科技有限公司 | 一种导航方法、装置及电子设备和存储介质 |
CN114199243B (zh) * | 2020-09-18 | 2024-05-24 | 浙江舜宇智能光学技术有限公司 | 用于机器人的位姿估计和运动规划方法、装置和机器人 |
CN112630745A (zh) * | 2020-12-24 | 2021-04-09 | 深圳市大道智创科技有限公司 | 一种基于激光雷达的环境建图方法和装置 |
CN112697151B (zh) * | 2020-12-24 | 2023-02-21 | 北京百度网讯科技有限公司 | 移动机器人的初始点的确定方法、设备和存储介质 |
CN113191412A (zh) * | 2021-04-25 | 2021-07-30 | 珠海格力电器股份有限公司 | 设备重定位方法、装置、计算机设备和存储介质 |
CN113375683A (zh) * | 2021-06-10 | 2021-09-10 | 亿嘉和科技股份有限公司 | 机器人环境地图实时更新方法 |
CN113624223B (zh) * | 2021-07-30 | 2024-05-24 | 中汽创智科技有限公司 | 一种室内停车场地图构建方法及装置 |
CN114577214B (zh) * | 2022-03-02 | 2022-09-20 | 哈尔滨工业大学 | 一种应用于跨异构多层空间的轮式机器人路径规划方法 |
CN114742884B (zh) * | 2022-06-09 | 2022-11-22 | 杭州迦智科技有限公司 | 一种基于纹理的建图、里程计算、定位方法及*** |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN104732518A (zh) * | 2015-01-19 | 2015-06-24 | 北京工业大学 | 一种基于智能机器人地面特征的ptam改进方法 |
CN104848851A (zh) * | 2015-05-29 | 2015-08-19 | 山东鲁能智能技术有限公司 | 基于多传感器数据融合构图的变电站巡检机器人及其方法 |
CN105865449A (zh) * | 2016-04-01 | 2016-08-17 | 深圳杉川科技有限公司 | 基于激光和视觉的移动机器人的混合定位方法 |
CN106092104A (zh) * | 2016-08-26 | 2016-11-09 | 深圳微服机器人科技有限公司 | 一种室内机器人的重定位方法及装置 |
CN107677279A (zh) * | 2017-09-26 | 2018-02-09 | 上海思岚科技有限公司 | 一种定位建图的方法及*** |
Family Cites Families (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN100449444C (zh) * | 2006-09-29 | 2009-01-07 | 浙江大学 | 移动机器人在未知环境中同时定位与地图构建的方法 |
CN104865578B (zh) * | 2015-05-12 | 2018-05-04 | 上海交通大学 | 一种室内停车场高精细地图生成装置及方法 |
CN105843223B (zh) * | 2016-03-23 | 2018-11-20 | 东南大学 | 一种基于空间词袋模型的移动机器人三维建图与避障方法 |
CN106599108B (zh) * | 2016-11-30 | 2019-12-31 | 浙江大学 | 一种三维环境中多模态环境地图构建方法 |
CN107193279A (zh) * | 2017-05-09 | 2017-09-22 | 复旦大学 | 基于单目视觉和imu信息的机器人定位与地图构建*** |
-
2017
- 2017-09-26 CN CN201710927028.3A patent/CN107677279B/zh active Active
-
2018
- 2018-09-20 WO PCT/CN2018/106812 patent/WO2019062651A1/zh active Application Filing
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN104732518A (zh) * | 2015-01-19 | 2015-06-24 | 北京工业大学 | 一种基于智能机器人地面特征的ptam改进方法 |
CN104848851A (zh) * | 2015-05-29 | 2015-08-19 | 山东鲁能智能技术有限公司 | 基于多传感器数据融合构图的变电站巡检机器人及其方法 |
CN105865449A (zh) * | 2016-04-01 | 2016-08-17 | 深圳杉川科技有限公司 | 基于激光和视觉的移动机器人的混合定位方法 |
CN106092104A (zh) * | 2016-08-26 | 2016-11-09 | 深圳微服机器人科技有限公司 | 一种室内机器人的重定位方法及装置 |
CN107677279A (zh) * | 2017-09-26 | 2018-02-09 | 上海思岚科技有限公司 | 一种定位建图的方法及*** |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2022012229A1 (zh) * | 2020-07-15 | 2022-01-20 | 京东科技信息技术有限公司 | 机器人的重定位方法及装置、存储介质、电子设备 |
Also Published As
Publication number | Publication date |
---|---|
CN107677279B (zh) | 2020-04-24 |
CN107677279A (zh) | 2018-02-09 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
WO2019062651A1 (zh) | 一种定位建图的方法及*** | |
JP6745328B2 (ja) | 点群データを復旧するための方法及び装置 | |
WO2019062649A1 (zh) | 一种自适应区域划分的方法及*** | |
JP7282080B2 (ja) | 二次元画像および点群データからの三次元境界ボックス | |
CN109084746B (zh) | 用于具有辅助传感器的自主平台引导***的单目模式 | |
CN109540142B (zh) | 一种机器人定位导航的方法、装置、计算设备 | |
CN113376650B (zh) | 移动机器人定位方法及装置、电子设备及存储介质 | |
US11549817B2 (en) | Distributed device mapping | |
US10614324B2 (en) | Method and apparatus for identifying static obstacle | |
WO2018112926A1 (zh) | 定位方法、终端和服务器 | |
JP2015148601A (ja) | マッピング、位置特定、及び姿勢補正のためのシステム及び方法 | |
US11475591B2 (en) | Hybrid metric-topological camera-based localization | |
KR20220066325A (ko) | 이동 로봇용 장애물 정보 감지 방법, 장치 | |
WO2020063878A1 (zh) | 一种处理数据的方法和装置 | |
US11880931B2 (en) | High-definition city mapping | |
CN113902801A (zh) | 移动机器人重定位方法、装置、设备及存储介质 | |
WO2024119776A1 (zh) | 一种基于votenet模型的三维目标检测方法和*** | |
CN112923938A (zh) | 一种地图优化方法、装置、存储介质及*** | |
Yu et al. | On designing computing systems for autonomous vehicles: A perceptin case study | |
US20220292290A1 (en) | Distributed computing network to perform simultaneous localization and mapping | |
Wang | Investigation of SLAMs applicable on computational-resource-constrained cars in urban environments | |
CN117830397A (zh) | 重定位方法、装置、电子设备、介质和车辆 | |
Tian | Research And Design of A 3D Reconstruction Algorithm with Lio Algorithm | |
CN118189948A (zh) | 局部地图构建方法、装置、存储介质和机器人 | |
CN112639502A (zh) | 机器人位姿估计 |
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: 18863623 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: 18863623 Country of ref document: EP Kind code of ref document: A1 |