Indoor navigation method and device based on tactile stick, electronic equipment and storage medium

By combining lidar sensors and IMU on the smart blind rod, 3D point cloud maps are built and path planning are carried out, the shortcomings of existing smart blind rods positioning and path planning in indoor scenarios are solved, and more efficient blind rod navigation is achieved.

CN119022938AInactive Publication Date: 2024-11-26WUHAN YIXUN ELECTRONICS INFORMATION TECH

Patent Information

Application Number
CN202411514860.7
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-10-29
Publication Date
2024-11-26
Estimated Expiration
Not applicable · inactive patent

AI Technical Summary

Technical Problem

The existing smart blind stick cannot provide real-time and accurate positioning information and safe and reliable path planning solutions in indoor scenarios. It mainly relies on the sensor ranging function to remind users to avoid obstacles.

Method used

The combination of lidar sensor and inertial measurement unit (IMU) is used to construct a 3D point cloud map of indoor scenes, and globally positioned through the FAST LiDAR Inertial Odometry algorithm, and path planning is carried out based on the 2D map to ensure global and local paths without collisions.

Benefits of technology

It improves the real-time and accuracy of the positioning of the blind stick in indoor scenes, enhances the safety and reliability of path planning, and provides a collision-free navigation path.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119022938A_ABST
    Figure CN119022938A_ABST
Patent Text Reader

Abstract

The invention provides an indoor navigation method and device based on a tactile stick, electronic equipment and a storage medium, and belongs to the technical field of navigation, and the method comprises the following steps: constructing a 3D point cloud map of an indoor scene based on historical laser radar point cloud data and historical IMU data pre-collected by the tactile stick in the indoor scene; converting the 3D point cloud map to obtain a corresponding 2D map; acquiring current laser radar point cloud data and current IMU data in real time in an indoor scene, and performing global positioning on the tactile stick based on the current laser radar point cloud data and the current IMU data to obtain global positioning information of the tactile stick relative to the 3D point cloud map; and determining the target position of the tactile stick, and planning the path of the tactile stick in the indoor scene based on the 2D map, the current laser radar point cloud data, the global positioning information and the target position to obtain a collision-free global path and a collision-free local path. According to the invention, the safety and reliability of path planning in an indoor scene can be improved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the field of navigation technology, and in particular to an indoor navigation method, device, electronic equipment and storage medium based on a blind stick. Background Art

[0002] How to have a safe and convenient travel experience is one of the key issues to improve the quality of life and physical and mental health of visually impaired patients. Visually impaired patients mainly face three challenges in their daily actions: avoiding obstacles, real-time positioning, and finding ways indoors and outdoors. In recent years, the concept of smart canes has gradually emerged, but the main problem that most smart canes on the market solve is still how to detect and avoid obstacles faster. Such products usually rely on the ranging function of the sensor installed on the cane to detect nearby obstacles, and remind users to pay attention to safety through vibration or beeping, but they cannot provide users with real-time and accurate positioning information and safe and reliable path planning solutions in indoor scenes. Summary of the invention

[0003] The present invention provides an indoor navigation method, device, electronic device and storage medium based on a blind stick, so as to solve the defect that a blind stick in the prior art relies on the ranging function of a sensor to detect nearby obstacles and reminds the user to pay attention to safety by means of vibration or buzzing, but cannot provide the user with real-time and accurate positioning information and a safe and reliable path planning solution in an indoor scene.

[0004] In a first aspect, the present invention provides an indoor navigation method based on a blind stick, which is applied to a blind stick, wherein the blind stick includes a laser radar sensor and an inertial measurement unit (IMU) sensor, and the method includes: Based on the historical LiDAR point cloud data and historical IMU data pre-collected by the blind stick in the indoor scene, construct a 3D point cloud map of the indoor scene; Converting the 3D point cloud map to obtain a 2D map corresponding to the 3D point cloud map; Acquire current laser radar point cloud data and current IMU data in real time in the indoor scene, perform global positioning of the blind stick based on the current laser radar point cloud data and the current IMU data, and obtain global positioning information of the blind stick relative to the 3D point cloud map; The target position of the cane is determined, and based on the 2D map, the current laser radar point cloud data, the global positioning information and the target position, a path of the cane in the indoor scene is planned to obtain a collision-free global path and a local path.

[0005] In some embodiments, the planning of the path of the cane in the indoor scene based on the 2D map, the current laser radar point cloud data, the global positioning information and the target position to obtain a collision-free global path and a local path includes: Based on the obstacle information in the 2D map, construct an obstacle map and an expansion area; Performing global path planning according to the obstacle map and the expansion area, as well as the global positioning information and the target position, to obtain a collision-free global path; Detecting local obstacles according to the current laser radar point cloud data to obtain a detection result, and updating the obstacle map and the expansion area based on the detection result to obtain an updated obstacle map and the expansion area; The local positioning information of the cane is determined, and a local path planning is performed according to the updated obstacle map and the expansion area, as well as the local positioning information and the target position, to obtain a collision-free local path.

[0006] In some embodiments, performing global positioning of the cane based on the current laser radar point cloud data and the current IMU data to obtain global positioning information of the cane relative to the 3D point cloud map includes: Searching in the 3D point cloud map based on the current laser radar point cloud data, initializing the position and posture of the cane in the 3D point cloud map; Using an efficient FAST LiDAR Inertial Odometry (FAST-LIO) algorithm to process the current LiDAR point cloud data and the current IMU data in real time, to obtain a pose estimation result of the cane; Registering the current laser radar point cloud data with the 3D point cloud map to obtain a registration result; The position and posture estimation result of the cane is corrected based on the registration result to obtain global positioning information of the cane relative to the 3D point cloud map.

[0007] In some embodiments, constructing a 3D point cloud map of the indoor scene based on historical lidar point cloud data and historical IMU data pre-collected by the blind stick in the indoor scene includes: Obtain historical LiDAR point cloud data of indoor scenes and historical IMU data of blind sticks; Based on an efficient FAST LiDAR InertialOdometry Simultaneous Localization and Mapping (FAST-LIO-SLAM) algorithm, the historical LiDAR point cloud data and the historical IMU data are processed to construct a 3D point cloud map of the indoor scene.

[0008] In some embodiments, the FAST-LIO-SLAM algorithm based on efficient laser radar inertial odometer synchronous positioning and map construction processes the historical laser radar point cloud data and historical IMU data to construct a 3D point cloud map of the indoor scene, including: Synchronizing the historical LiDAR point cloud data and the historical IMU data; Performing denoising preprocessing on the historical IMU data to obtain preprocessed historical IMU data; Performing point cloud dedistortion preprocessing on the historical laser radar point cloud data to obtain preprocessed historical point cloud data; Performing feature extraction on the preprocessed historical point cloud data to obtain historical point cloud feature data; Using an Iterated Extended Kalman Filter (IEKF), based on the historical point cloud feature data and the pre-processed historical IMU data, the pose of the historical point cloud is iteratively estimated to obtain an optimal historical point cloud pose estimate; Based on the optimal historical point cloud pose estimation value, a 3D point cloud map of the indoor scene is constructed, and the 3D point cloud map is optimized.

[0009] In some embodiments, the optimizing the 3D point cloud map includes: Performing global loop detection on the 3D point cloud map to obtain a loop detection result; Determining adjacent pose constraints based on the historical lidar point cloud data and the historical IMU data; Based on the loop detection results and the adjacent pose constraints, the 3D point cloud map is globally optimized.

[0010] In some embodiments, converting the 3D point cloud map to obtain a 2D map corresponding to the 3D point cloud map includes: Performing point cloud filtering on the 3D point cloud map to obtain a filtered 3D point cloud map; Perform 2D projection on the filtered 3D point cloud map to obtain the 2D map.

[0011] In a second aspect, the present invention further provides an indoor navigation device based on a blind stick, comprising: An offline mapping module, used to construct a 3D point cloud map of the indoor scene based on historical lidar point cloud data and historical IMU data pre-collected by the blind stick in the indoor scene; A map conversion module, used to convert the 3D point cloud map to obtain a 2D map corresponding to the 3D point cloud map; A global positioning module, used for acquiring current laser radar point cloud data and current IMU data in real time in the indoor scene, and performing global positioning of the blind stick based on the current laser radar point cloud data and current IMU data to obtain global positioning information of the blind stick relative to the 3D point cloud map; The path planning module is used to determine the target position of the cane, and plan the path of the cane in the indoor scene based on the 2D map, the current laser radar point cloud data, the global positioning information and the target position to obtain a collision-free global path and a local path.

[0012] In a third aspect, the present invention further provides an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, the indoor navigation method based on a blind stick as described above is implemented.

[0013] In a fourth aspect, the present invention further provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements any of the above-described indoor navigation methods based on a blind stick.

[0014] The indoor navigation method, device, electronic device and storage medium based on the blind stick provided by the present invention construct a 3D point cloud map of the indoor scene based on historical laser radar point cloud data and historical IMU data, convert the 3D point cloud map to obtain a corresponding 2D map; obtain current laser radar point cloud data and current IMU data in real time in the indoor scene, combine them with the 3D point cloud map, perform global positioning of the blind stick, and improve the real-time and accuracy of positioning; perform path planning in the indoor scene based on the 2D map, current laser radar point cloud data, global positioning information and target position, obtain a collision-free global path and local path, and improve the safety and reliability of path planning in the indoor scene. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0016] Figure 1 is a flow chart of a method for indoor navigation based on a blind stick provided by an embodiment of the present invention; Figure 2 is a schematic diagram of a process for planning a path of a blind stick in an indoor scene provided by an embodiment of the present invention; Figure 3 is a schematic structural diagram of an indoor navigation device based on a blind stick provided in an embodiment of the present invention; Figure 4 It is a schematic diagram of the physical structure of an electronic device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0017] In order to make the purpose, technical solution and advantages of the present invention clearer, the technical solution of the present invention will be clearly and completely described below in conjunction with the drawings of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0018] Figure 1 The flowchart of the indoor navigation method based on the cane provided by the embodiment of the present invention is as follows. Figure 1 As shown, a method for indoor navigation based on a blind stick is provided, which is applied to a blind stick, and the blind stick includes a laser radar sensor and an inertial measurement unit IMU sensor, and includes the following steps: step 110, step 120, step 130 and step 140. The method flow steps are only used as a possible implementation of the present invention.

[0019] Step 110: construct a 3D point cloud map of the indoor scene based on the historical lidar point cloud data and historical IMU data pre-collected by the blind stick in the indoor scene.

[0020] Among them, the historical lidar point cloud data is the three-dimensional point cloud data of the indoor scene collected in advance by the lidar sensor, which is used to describe the spatial position and shape of objects in the indoor scene.

[0021] The historical IMU data is the historical motion data of the cane collected in advance by the IMU sensor, including historical acceleration and historical angular velocity, which is used to describe the historical motion state and posture changes of the cane.

[0022] In some embodiments, based on the historical LiDAR point cloud data and historical IMU data pre-collected by the cane in the indoor scene, a 3D point cloud map of the indoor scene is constructed, including: Obtain historical LiDAR point cloud data of indoor scenes and historical IMU data of blind sticks; Based on the efficient lidar inertial odometer synchronous positioning and map construction FAST-LIO-SLAM algorithm, historical lidar point cloud data and historical IMU data are processed to construct a 3D point cloud map of indoor scenes.

[0023] Optionally, the FAST-LIO-SLAM algorithm is used to fuse historical lidar point cloud data with historical IMU data, and a 3D point cloud map of the indoor scene is constructed based on the fused data to improve the accuracy of the 3D point cloud map.

[0024] In some embodiments, based on the efficient LiDAR inertial odometer synchronous positioning and map construction FAST-LIO-SLAM algorithm, historical LiDAR point cloud data and historical IMU data are processed to construct a 3D point cloud map of the indoor scene, including: Synchronize historical LiDAR point cloud data and historical IMU data; Perform denoising preprocessing on the historical IMU data to obtain preprocessed historical IMU data; Perform point cloud dedistortion preprocessing on historical laser radar point cloud data to obtain preprocessed historical point cloud data; Perform feature extraction on the preprocessed historical point cloud data to obtain historical point cloud feature data; The iterative extended Kalman filter (IEKF) is used to iteratively estimate the pose of the historical point cloud based on the historical point cloud feature data and the pre-processed historical IMU data to obtain the optimal historical point cloud pose estimate. Based on the optimal historical point cloud pose estimation value, a 3D point cloud map of the indoor scene is constructed and optimized.

[0025] Optionally, the historical point cloud feature data includes at least plane feature data and edge feature data.

[0026] Among them, the iterative extended Kalman filter IEKF is a filter used for state estimation, which is mainly used in nonlinear systems. Through iterative optimization, IEKF can better handle uncertainty and nonlinear problems and improve the stability and reliability of state estimation.

[0027] In some embodiments, optimizing the 3D point cloud map includes: Perform global loop detection on the 3D point cloud map to obtain loop detection results; Determine adjacent pose constraints based on historical LiDAR point cloud data and historical IMU data; Based on the loop detection results and adjacent pose constraints, the 3D point cloud map is globally optimized.

[0028] Optionally, features, such as local descriptors, are extracted from historical lidar point cloud data.

[0029] Optionally, a point cloud registration algorithm, such as the Iterative Closest Point (ICP) algorithm, is used to find the historical pose that is most similar to the current pose, and the loop detection result is confirmed by geometric detection.

[0030] Optionally, a point cloud registration algorithm (such as ICP) is used to calculate the relative pose transformation between adjacent frames, and the historical lidar point cloud data and historical IMU data are combined to accurately estimate the relative pose of adjacent frames, and the calculated relative pose is introduced as a constraint into the optimization problem.

[0031] It can be understood that loop closure detection can identify repeated positions, correct accumulated errors, and ensure the consistency of the 3D map. Adjacent pose constraints can provide constraints between adjacent frames, reduce sensor measurement errors and accumulated errors, and help to better adjust and correct pose estimates during the optimization process, thereby improving the accuracy of the 3D map.

[0032] Step 120: convert the 3D point cloud map to obtain a 2D map corresponding to the 3D point cloud map.

[0033] It can be understood that by converting the 3D point cloud map, a 2D map that is easy to understand and process can be obtained, thereby improving the efficiency of subsequent path navigation.

[0034] In some embodiments, converting the 3D point cloud map to obtain a 2D map corresponding to the 3D point cloud map includes: Perform point cloud filtering on the 3D point cloud map to obtain a filtered 3D point cloud map; Perform 2D projection on the filtered 3D point cloud map to obtain a 2D map.

[0035] Optionally, the filtered 3D point cloud map is projected onto a two-dimensional plane, the projected two-dimensional data is discretized into a grid, such as creating a raster map, and the discretized data is plotted into a 2D map to display information such as obstacles and paths.

[0036] Optionally, apply image processing techniques to optimize and clean up the 2D map, such as denoising, smoothing edges, etc.

[0037] Step 130: Acquire the current LiDAR point cloud data and the current IMU data in real time in the indoor scene, perform global positioning of the cane based on the current LiDAR point cloud data and the current IMU data, and obtain global positioning information of the cane relative to the 3D point cloud map.

[0038] Optionally, during the movement of the cane, the lidar sensor on the cane is used to collect current lidar point cloud data of the indoor scene in real time, and the IMU sensor on the cane is used to collect current IMU data of the cane in real time.

[0039] Among them, the current IMU data at least includes the current acceleration and angular velocity data.

[0040] Optionally, the current lidar point cloud data and the current IMU data are preprocessed to obtain preprocessed point cloud data and preprocessed IMU data; based on the preprocessed point cloud data and the preprocessed IMU data, the current position of the blind stick is estimated to obtain the global position of the blind stick in the 3D map.

[0041] In some embodiments, based on the current laser radar point cloud data and the current IMU data, the global positioning of the cane is performed to obtain the global positioning information of the cane relative to the 3D point cloud map, including: Search in the 3D point cloud map based on the current LiDAR point cloud data and initialize the position of the cane in the 3D point cloud map; The efficient LiDAR inertial odometer FAST-LIO algorithm is used to process the current LiDAR point cloud data and the current IMU data in real time to obtain the position estimation result of the blind stick. Register the current LiDAR point cloud data with the 3D point cloud map to obtain the registration result; Based on the registration result, the position estimation result of the cane is corrected to obtain the global positioning information of the cane relative to the 3D point cloud map.

[0042] It can be understood that by using the current lidar point cloud data to search in the 3D point cloud map, the position and posture of the cane are preliminarily estimated, and the FAST-LIO algorithm is used to combine the current lidar point cloud and the current IMU data to estimate the position and posture of the cane in real time. The current lidar point cloud data is aligned with the 3D point cloud map, the positioning information of the cane is optimized, and the position and posture estimation is adjusted based on the alignment result, thereby obtaining more accurate global positioning information, thereby improving the real-time and accuracy of positioning.

[0043] Step 140: determine the target position of the cane, and plan the path of the cane in the indoor scene based on the 2D map, the current lidar point cloud data, the global positioning information and the target position to obtain a collision-free global path and a local path.

[0044] Optionally, the target position in the indoor scene is determined based on user input or a preset target point.

[0045] Optionally, the current lidar point cloud data is fused with the 2D map to update the 2D map.

[0046] Optionally, obstacles are detected based on LiDAR point cloud data and marked on the 2D map.

[0047] Figure 2 A schematic diagram of a process for planning a path of a blind stick in an indoor scene provided by an embodiment of the present invention, such as Figure 2 As shown, in some embodiments, in step 140, based on the 2D map, the current laser radar point cloud data, the global positioning information and the target position, the path of the cane in the indoor scene is planned to obtain a collision-free global path and a local path, including: Step 141: construct an obstacle map and an expansion area based on obstacle information in the 2D map; Step 142: Perform global path planning based on the obstacle map and the expansion area, as well as the global positioning information and the target position, to obtain a collision-free global path; Step 143: Detect local obstacles according to the current laser radar point cloud data to obtain detection results, and update the obstacle map and expansion area based on the detection results to obtain an updated obstacle map and expansion area; Step 144: determine the local positioning information of the cane, perform local path planning based on the updated obstacle map and expansion area, as well as the local positioning information and the target position, and obtain a collision-free local path.

[0048] The expansion area refers to an area that is expanded outward by a certain distance based on the original obstacle. The expansion area is used to ensure that the cane user has a sufficient safety distance when walking in indoor scenes to avoid direct contact with obstacles.

[0049] Optionally, the expansion area is constructed by comprehensively considering factors such as the step length, motion error, and obstacle shape of the cane user.

[0050] Optionally, in the area surrounding the blind stick, a local 2D map is generated using current laser radar point cloud data, and obstacles in the local 2D map are detected in real time.

[0051] In an embodiment of the present invention, a 3D point cloud map of an indoor scene is constructed based on historical lidar point cloud data and historical IMU data, and the 3D point cloud map is converted to obtain a corresponding 2D map; the current lidar point cloud data and the current IMU data are acquired in real time in the indoor scene, and combined with the 3D point cloud map to perform global positioning of the cane, thereby improving the real-time and accuracy of positioning; based on the 2D map, the current lidar point cloud data, the global positioning information and the target position, path planning is performed in the indoor scene to obtain a collision-free global path and a local path, thereby improving the safety and reliability of path planning in the indoor scene.

[0052] The following describes an indoor navigation device based on a blind stick provided in an embodiment of the present invention. The indoor navigation device based on a blind stick described below and the indoor navigation method based on a blind stick described above can refer to each other.

[0053] Figure 3 A schematic diagram of the structure of an indoor navigation device based on a blind stick provided in an embodiment of the present invention is shown in FIG. Figure 3 As shown, the indoor navigation device 300 based on the blind stick includes: The offline mapping module 310 is used to construct a 3D point cloud map of the indoor scene based on the historical LiDAR point cloud data and historical IMU data pre-collected by the blind stick in the indoor scene; A map conversion module 320 is used to convert the 3D point cloud map to obtain a 2D map corresponding to the 3D point cloud map; The global positioning module 330 is used to obtain the current laser radar point cloud data and the current IMU data in real time in the indoor scene, and perform global positioning of the blind stick based on the current laser radar point cloud data and the current IMU data to obtain the global positioning information of the blind stick relative to the 3D point cloud map; The path planning module 340 is used to determine the target position of the cane, and plans the path of the cane in the indoor scene based on the 2D map, the current lidar point cloud data, the global positioning information and the target position to obtain a collision-free global path and local path.

[0054] Optionally, based on the 2D map, the current LiDAR point cloud data, the global positioning information and the target position, the path of the cane in the indoor scene is planned to obtain a collision-free global path and a local path, including: Based on the obstacle information in the 2D map, the obstacle map and expansion area are constructed; Based on the obstacle map and expansion area, as well as the global positioning information and target position, global path planning is performed to obtain a collision-free global path; Detect local obstacles according to the current lidar point cloud data to obtain detection results, and update the obstacle map and expansion area based on the detection results to obtain an updated obstacle map and expansion area; The local positioning information of the cane is determined, and a local path planning is performed according to the updated obstacle map and the expansion area, as well as the local positioning information and the target position, to obtain a collision-free local path.

[0055] Optionally, based on the current laser radar point cloud data and the current IMU data, the cane is globally positioned to obtain global positioning information of the cane relative to the 3D point cloud map, including: Search in the 3D point cloud map based on the current LiDAR point cloud data and initialize the position of the cane in the 3D point cloud map; The efficient LiDAR inertial odometer FAST-LIO algorithm is used to process the current LiDAR point cloud data and the current IMU data in real time to obtain the position estimation result of the blind stick. Register the current LiDAR point cloud data with the 3D point cloud map to obtain the registration result; Based on the registration result, the position estimation result of the cane is corrected to obtain the global positioning information of the cane relative to the 3D point cloud map.

[0056] Optionally, based on historical lidar point cloud data and historical IMU data pre-collected by the cane in the indoor scene, a 3D point cloud map of the indoor scene is constructed, including: Obtain historical LiDAR point cloud data of indoor scenes and historical IMU data of blind sticks; Based on the efficient lidar inertial odometer synchronous positioning and map construction FAST-LIO-SLAM algorithm, historical lidar point cloud data and historical IMU data are processed to construct a 3D point cloud map of indoor scenes.

[0057] Optionally, based on the efficient LiDAR-Inertial Odometry Synchronous Positioning and Mapping FAST-LIO-SLAM algorithm, the historical LiDAR point cloud data and historical IMU data are processed to construct a 3D point cloud map of the indoor scene, including: Synchronize historical LiDAR point cloud data and historical IMU data; Perform denoising preprocessing on the historical IMU data to obtain preprocessed historical IMU data; Perform point cloud dedistortion preprocessing on historical laser radar point cloud data to obtain preprocessed historical point cloud data; Perform feature extraction on the preprocessed historical point cloud data to obtain historical point cloud feature data; The iterative extended Kalman filter (IEKF) is used to iteratively estimate the pose of the historical point cloud based on the historical point cloud feature data and the pre-processed historical IMU data to obtain the optimal historical point cloud pose estimate. Based on the optimal historical point cloud pose estimation value, a 3D point cloud map of the indoor scene is constructed and optimized.

[0058] Optionally, the 3D point cloud map is optimized, including: Perform global loop detection on the 3D point cloud map to obtain loop detection results; Determine adjacent pose constraints based on historical LiDAR point cloud data and historical IMU data; Based on the loop detection results and adjacent pose constraints, the 3D point cloud map is globally optimized.

[0059] Optionally, the 3D point cloud map is converted to obtain a 2D map corresponding to the 3D point cloud map, including: Perform point cloud filtering on the 3D point cloud map to obtain a filtered 3D point cloud map; Perform 2D projection on the filtered 3D point cloud map to obtain a 2D map.

[0060] It should be noted here that the indoor navigation device based on the blind stick provided in the embodiment of the present invention can implement all the method steps implemented in the above-mentioned indoor navigation method embodiment based on the blind stick, and can achieve the same technical effect. The parts and beneficial effects that are the same as those in the method embodiment will not be specifically repeated here.

[0061] Figure 4 A schematic diagram of the physical structure of an electronic device provided by an embodiment of the present invention, such as Figure 4As shown, the electronic device may include: a processor (processor) 410 , a communication interface (Communications Interface) 420 , a memory (memory) 430 and a communication bus 440 , wherein the processor 410 , the communication interface 420 , and the memory 430 communicate with each other through the communication bus 440 . The processor 410 can call the logic instructions in the memory 430 to execute the indoor navigation method based on the blind stick, which includes: constructing a 3D point cloud map of the indoor scene based on the historical lidar point cloud data and the historical IMU data pre-collected by the blind stick in the indoor scene; converting the 3D point cloud map to obtain a 2D map corresponding to the 3D point cloud map; acquiring the current lidar point cloud data and the current IMU data in real time in the indoor scene, and globally positioning the blind stick based on the current lidar point cloud data and the current IMU data to obtain the global positioning information of the blind stick relative to the 3D point cloud map; determining the target position of the blind stick, and planning the path of the blind stick in the indoor scene based on the 2D map, the current lidar point cloud data, the global positioning information and the target position to obtain a collision-free global path and a local path.

[0062] In addition, the logic instructions in the above-mentioned memory 430 can be implemented in the form of a software functional unit and can be stored in a computer-readable storage medium when it is sold or used as an independent product. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art or the part of the technical solution, can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk, etc. Various media that can store program codes.

[0063] On the other hand, the present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, is implemented to execute the indoor navigation method based on the blind stick provided by the above-mentioned methods, the method comprising: constructing a 3D point cloud map of the indoor scene based on historical lidar point cloud data and historical IMU data pre-collected by the blind stick in the indoor scene; converting the 3D point cloud map to obtain a 2D map corresponding to the 3D point cloud map; acquiring current lidar point cloud data and current IMU data in real time in the indoor scene, and globally positioning the blind stick based on the current lidar point cloud data and current IMU data to obtain global positioning information of the blind stick relative to the 3D point cloud map; determining the target position of the blind stick, and planning the path of the blind stick in the indoor scene based on the 2D map, the current lidar point cloud data, the global positioning information and the target position to obtain a collision-free global path and local path.

[0064] The device embodiments described above are merely illustrative, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. Ordinary technicians in this field can understand and implement it without paying creative labor.

[0065] Through the description of the above implementation methods, those skilled in the art can clearly understand that each implementation method can be implemented by means of software plus a necessary general hardware platform, and of course, can also be implemented by hardware. Based on this understanding, the above technical solution is essentially or the part that contributes to the prior art can be embodied in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a disk, an optical disk, etc., including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.

[0066] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for indoor navigation based on a blind stick, characterized in that: Applied to a blind stick, the blind stick includes a laser radar sensor and an inertial measurement unit (IMU) sensor, and the method includes: Based on the historical LiDAR point cloud data and historical IMU data pre-collected by the blind stick in the indoor scene, construct a 3D point cloud map of the indoor scene; Converting the 3D point cloud map to obtain a 2D map corresponding to the 3D point cloud map; Acquire current laser radar point cloud data and current IMU data in real time in the indoor scene, perform global positioning of the blind stick based on the current laser radar point cloud data and the current IMU data, and obtain global positioning information of the blind stick relative to the 3D point cloud map; The target position of the cane is determined, and based on the 2D map, the current laser radar point cloud data, the global positioning information and the target position, a path of the cane in the indoor scene is planned to obtain a collision-free global path and a local path.

2. The indoor navigation method based on a blind stick according to claim 1, characterized in that: The method of planning the path of the cane in the indoor scene based on the 2D map, the current laser radar point cloud data, the global positioning information and the target position to obtain a collision-free global path and a local path includes: Based on the obstacle information in the 2D map, construct an obstacle map and an expansion area; Performing global path planning according to the obstacle map and the expansion area, as well as the global positioning information and the target position, to obtain a collision-free global path; Detecting local obstacles according to the current laser radar point cloud data to obtain a detection result, and updating the obstacle map and the expansion area based on the detection result to obtain an updated obstacle map and the expansion area; The local positioning information of the cane is determined, and a local path planning is performed according to the updated obstacle map and the expansion area, as well as the local positioning information and the target position, to obtain a collision-free local path.

3. The indoor navigation method based on a blind stick according to claim 1, characterized in that: The globally positioning the cane based on the current laser radar point cloud data and the current IMU data to obtain the global positioning information of the cane relative to the 3D point cloud map includes: Searching in the 3D point cloud map based on the current laser radar point cloud data, initializing the position and posture of the cane in the 3D point cloud map; Using the efficient LiDAR inertial odometer FAST-LIO algorithm to process the current LiDAR point cloud data and the current IMU data in real time, to obtain the position and posture estimation result of the cane; Registering the current laser radar point cloud data with the 3D point cloud map to obtain a registration result; The position and posture estimation result of the cane is corrected based on the registration result to obtain global positioning information of the cane relative to the 3D point cloud map.

4. The indoor navigation method based on a blind stick according to claim 1, characterized in that: The method of constructing a 3D point cloud map of the indoor scene based on the historical laser radar point cloud data and the historical IMU data pre-collected by the blind stick in the indoor scene includes: Obtain historical LiDAR point cloud data of indoor scenes and historical IMU data of blind sticks; Based on the efficient LiDAR inertial odometer synchronous positioning and map construction FAST-LIO-SLAM algorithm, the historical LiDAR point cloud data and historical IMU data are processed to construct a 3D point cloud map of the indoor scene.

5. The indoor navigation method based on a blind stick according to claim 4, characterized in that: The FAST-LIO-SLAM algorithm based on efficient laser radar inertial odometer synchronous positioning and map construction processes the historical laser radar point cloud data and historical IMU data to construct a 3D point cloud map of the indoor scene, including: Synchronizing the historical LiDAR point cloud data and the historical IMU data; Performing denoising preprocessing on the historical IMU data to obtain preprocessed historical IMU data; Performing point cloud dedistortion preprocessing on the historical laser radar point cloud data to obtain preprocessed historical point cloud data; Performing feature extraction on the preprocessed historical point cloud data to obtain historical point cloud feature data; Using an iterative extended Kalman filter (IEKF), based on the historical point cloud feature data and the pre-processed historical IMU data, the pose of the historical point cloud is iteratively estimated to obtain an optimal historical point cloud pose estimate; Based on the optimal historical point cloud pose estimation value, a 3D point cloud map of the indoor scene is constructed, and the 3D point cloud map is optimized.

6. The indoor navigation method based on a blind stick according to claim 5, characterized in that: The optimizing the 3D point cloud map includes: Performing global loop detection on the 3D point cloud map to obtain a loop detection result; Determining adjacent pose constraints based on the historical lidar point cloud data and the historical IMU data; Based on the loop detection results and the adjacent pose constraints, the 3D point cloud map is globally optimized.

7. The indoor navigation method based on a blind stick according to any one of claims 2 to 6, characterized in that: The converting the 3D point cloud map to obtain a 2D map corresponding to the 3D point cloud map includes: Performing point cloud filtering on the 3D point cloud map to obtain a filtered 3D point cloud map; Perform 2D projection on the filtered 3D point cloud map to obtain the 2D map.

8. An indoor navigation device based on a blind stick, characterized in that: include: An offline mapping module, used to construct a 3D point cloud map of the indoor scene based on historical lidar point cloud data and historical IMU data pre-collected by the blind stick in the indoor scene; A map conversion module, used to convert the 3D point cloud map to obtain a 2D map corresponding to the 3D point cloud map; A global positioning module, used for acquiring current laser radar point cloud data and current IMU data in real time in the indoor scene, and performing global positioning of the blind stick based on the current laser radar point cloud data and current IMU data to obtain global positioning information of the blind stick relative to the 3D point cloud map; The path planning module is used to determine the target position of the cane, and plan the path of the cane in the indoor scene based on the 2D map, the current laser radar point cloud data, the global positioning information and the target position to obtain a collision-free global path and a local path.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the computer program, the indoor navigation method based on the blind stick according to any one of claims 1 to 7 is implemented.

10. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the indoor navigation method based on a blind stick as claimed in any one of claims 1 to 7 is implemented.

Citation Information

Patent Citations

  • Indoor navigation method based on intelligent crutch and intelligent crutch

    CN109297488A

  • Positioning navigation method suitable for complex three-dimensional environment

    CN113269837A

  • Blind person obstacle avoidance navigation system based on visual SLAM

    CN114712181A

Cited By

  • Picking robot positioning method, device and equipment, medium and product

    CN119722788A

  • A positioning method, device, equipment, medium and product for a picking robot

    CN119722788B