Robot indoor environment sensing method and device

Abstract

The invention discloses a robot indoor environment sensing method and device, and relates to the technical field of robot indoor environment sensing. The invention aims to effectively improve the efficiency of sensing the indoor environment. Comprising the following steps that S1, a control module is used for controlling a robot, and the control module executes a decision command based on node type distribution and conducts decision control on movement of the robot; s2, an induction module installed on the robot obtains indoor environment parameters such as height and size information of an obstacle and feeds back the indoor environment parameters to all nodes on the control module, and support is provided for walking of the robot in the indoor environment. According to the invention, a system framework is established, a plurality of processing nodes are installed on the framework, information transmission and mutual interaction among data can be realized conveniently, and the visual sensor is integrated in the depth camera, so that in the process of collecting indoor environment data, the energy consumption of the system is reduced, and meanwhile, the depth collection of the environment data is realized.

Description

Technical Field

[0001] The present invention relates to the technical field of robot indoor environment perception, and in particular to a robot indoor environment perception method and device. Background Art

[0002] In the prior art, currently, artificial intelligence autonomous mobile robots are usually equipped with a variety of different sensors and perception data processing algorithms to meet the complex obstacle detection, boundary detection, moving object detection and other perception requirements in the application scenarios, so that the robot can perceive the environment in which it is located during the movement in the indoor environment, thereby achieving safe and stable walking;

[0003] The existing robots used for indoor environment perception have insufficient coordination between their control modules during the perception process. This results in the robot being unable to accurately perceive the indoor environment when it needs to deal with various complex situations in the indoor environment, such as obstacle avoidance and speed control, resulting in errors in the final perception accuracy.

[0004] In summary, this application proposes a robot indoor environment perception method and device, which effectively overcomes the above-mentioned defects and improves its efficiency in perceiving the indoor environment. Summary of the invention

[0005] The purpose of the present invention is to solve the shortcomings of the prior art and to propose a robot indoor environment perception method and device.

[0006] In order to achieve the above object, the present invention adopts the following technical solutions:

[0007] A robot indoor environment perception method comprises the following steps:

[0008] S1: The robot is controlled by the control module. The control module is based on node distribution, executes decision commands, and makes decisions and controls the movement of the robot;

[0009] S2: The sensing module installed on the robot obtains indoor environmental parameters such as obstacle height and size information and feeds them back to each node on the control module to support the robot's walking in the indoor environment;

[0010] S3: Use the execution module installed on the robot to control the walking mechanism at the bottom of the robot to move in the direction of the generated path.

[0011] Preferably, the control module is used to establish a system framework, on which a plurality of processing nodes are installed, and each processing node is connected in series.

[0012] Further: the control module establishes a connection with the terminal based on a distributed communication method, and uses a map construction algorithm, a path planning algorithm, and a positioning algorithm to realize the control of the sensing module and the execution module. The specific control method of the control module includes the following steps:

[0013] S11: using the sensing module to obtain the global map size and posture data information of the indoor environment;

[0014] S12: After acquiring the data information, the data information is fed back to the terminal, and the terminal generates a path plan after processing, which specifically includes global planning and local planning;

[0015] S13: Use the execution module to control the walking mechanism to achieve autonomous navigation.

[0016] Further: the sensing module includes an image processing unit, which is based on a depth camera, an odometer and a laser radar. The depth camera is integrated with a visual sensor, which uses structured light technology. The sensing module is also integrated with a laser radar, which is used to effectively identify obstacles. The laser radar detects obstacles based on a threshold method, and its detection method includes the following steps:

[0017] C1: Process and collect the data detected by the LiDAR and transmit it to the cloud for storage;

[0018] C2: Clean up the information in the cloud to improve the accuracy of information collection and reduce the interference of redundant and misplaced information;

[0019] C3: According to the specific perception of the indoor environment, set the corresponding threshold value, and perceive the obstacles in the indoor environment by identifying the threshold value.

[0020] As a preferred solution of the present invention: the image processing unit processes the image captured by the depth camera based on the image processing algorithm, and performs edge detection, feature extraction and object recognition on the wall in the indoor environment during the processing. The edge detection algorithm is based on Canny, and the process of edge detection includes the following steps:

[0021] A1: Reduce noise in edge detection process through Gaussian filtering;

[0022] A2: Detect the gradient of the image and complete the edge recognition;

[0023] A3: Suppress non-maximum values ​​to reduce noise in the gradient calculation process;

[0024] A4: Determine the edge intensity level by detecting the double peaks;

[0025] A5: After tracking the edge, connect each edge point to form a complete edge line.

[0026] As a further solution of the present invention: the feature extraction is based on the SIFT algorithm, and the extraction method thereof comprises the following steps:

[0027] B1: Reduce image noise through image filtering and improve the accuracy of feature extraction;

[0028] B2: Identify feature points by calculating image gradients;

[0029] B3: Calculate the split gradient of the image to identify feature points;

[0030] B4: Provide description information for feature point description for subsequent feature matching.

[0031] A robot indoor environment sensing device comprises a control module, a walking mechanism, a sensing module and a buffer mechanism. The walking mechanism comprises a base, a McHim wheel and a drive motor. The drive motor is fixed to an outer wall of one side of the base, and the McHim wheel is connected to an output end of the drive motor through a coupling.

[0032] On the basis of the above-mentioned solution: a main control box is fixed to the outer wall of the top of the base by bolts, and the control module and the sensing module are integrated in the main control box.

[0033] On the basis of the above-mentioned scheme: the top outer wall of the base is welded with a push rod, and a top plate is welded on the top of the push rod, and a buffer mechanism is installed on the top of the top plate, the buffer mechanism includes a sliding column and a sliding seat, the sliding column is fixed to the top outer wall of the top plate, and the sliding seat is movably connected to the outer wall of the sliding column, and a spring is clamped between the outer wall on one side of the sliding seat and the inner wall on one side of the sliding column, and the inner wall of the sliding seat is movably connected to a swing arm, and the two swing arms are movably connected to a buffer plate at one end away from the sliding seat.

[0034] Based on the above solution: a laser radar and a depth camera are fixed to the top outer wall of the base by bolts.

[0035] The beneficial effects of the present invention are:

[0036] 1. A robot indoor environment perception method and device, establish a system framework, and install a number of processing nodes on the framework. The processing nodes are connected in series. Each node can be used alone, and can also work together to facilitate the transmission of information and the interaction between data.

[0037] 2. A robot indoor environment perception method and device, which uses a visual sensor integrated in a depth camera to effectively reduce its own energy consumption while realizing in-depth collection of environmental data during the process of collecting indoor environmental data. The sensing module is also integrated with a laser radar, which is used to effectively identify obstacles and has strong anti-interference capabilities.

[0038] 3. A robot indoor environment perception method and device, which is provided with a drive motor to drive the Mechheim wheels to rotate. Since the four Mechheim wheels are independently driven by the drive motors, the drive motors are used to control the Mechheim wheels to rotate independently, so that the four Mechheim wheels can realize the steering and walking of the entire device.

[0039] 4. A robot indoor environment perception method and device. When the entire perception device is walking indoors and an obstacle falls, a buffer plate is used to support the obstacle. After the buffer plate is compressed, the slide seat slides along the sliding column under the connection of the swing arm. A spring is provided to play a buffering role, thereby effectively reducing the impact of the obstacle on the entire device. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] Figure 1 It is a flow chart of a robot indoor environment perception method proposed by the present invention;

[0041] Figure 2 It is a schematic diagram of the overall structure of a robot indoor environment perception device proposed by the present invention.

[0042] In the figure: base 1; Macheim wheel 2; main control box 3; top plate 4; buffer plate 5; sliding column 6; swing arm 7; slide seat 8; spring 9; push rod 10; laser radar 11; depth camera 12; drive motor 13. DETAILED DESCRIPTION

[0043] The technical solution of this patent is further described in detail below in conjunction with specific implementation methods.

[0044] The embodiments of the present invention are described in detail below, and examples of the embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present invention, and cannot be understood as limiting the present invention.

[0045] Embodiment 1:

[0046] A robot indoor environment perception method, such as Figure 1 As shown, the following steps are included:

[0047] S1: The robot is controlled by the control module. The control module is based on node distribution, executes decision commands, and makes decisions and controls the movement of the robot;

[0048] S2: The sensing module installed on the robot obtains indoor environmental parameters such as obstacle height and size information and feeds them back to each node on the control module to support the robot's walking in the indoor environment;

[0049] S3: Using the execution module installed on the robot to control the walking mechanism at the bottom of the robot to move in the direction of the generated path;

[0050] The control module is used to establish a system framework, on which a number of processing nodes are installed. The processing nodes are connected in series, and each node is used alone or in collaboration to achieve information transmission and data interaction.

[0051] The control module establishes a connection with the terminal based on a distributed communication method, and uses a map construction algorithm, a path planning algorithm, and a positioning algorithm to control the sensing module and the execution module. The specific control method of the control module includes the following steps:

[0052] S11: using the sensing module to obtain the global map size and posture data information of the indoor environment;

[0053] S12: After acquiring the data information, the data information is fed back to the terminal, and the terminal generates a path plan after processing, which specifically includes global planning and local planning;

[0054] S13: Using the execution module to control the walking mechanism to achieve autonomous navigation;

[0055] The sensing module of the robot in this embodiment includes an image processing unit, which is based on a depth camera, an odometer and a laser radar. The depth camera is integrated with a visual sensor, which uses structured light technology to effectively reduce its own energy consumption while realizing the deep collection of environmental data during the collection of indoor environmental data. The sensing module is also integrated with a laser radar, which is used to effectively identify obstacles and has strong anti-interference ability.

[0056] The image processing unit processes the image captured by the depth camera based on the image processing algorithm. During the processing, the wall in the indoor environment is edge detected, feature extracted and object recognized. The edge detection algorithm is based on Canny, and the process of edge detection includes the following steps:

[0057] A1: Reduce noise in edge detection process through Gaussian filtering;

[0058] A2: Detect the gradient of the image and complete the edge recognition;

[0059] A3: Suppress non-maximum values ​​to reduce noise in the gradient calculation process;

[0060] A4: Determine the edge intensity level by detecting the double peaks;

[0061] A5: After tracking the edge, connect each edge point to form a complete edge line;

[0062] The feature extraction is based on the SIFT algorithm, and the extraction method includes the following steps:

[0063] B1: Reduce image noise through image filtering and improve the accuracy of feature extraction;

[0064] B2: Identify feature points by calculating image gradients;

[0065] B3: Calculate the split gradient of the image to identify feature points;

[0066] B4: Provide description information for feature point description and subsequent feature matching;

[0067] The laser radar detects obstacles based on a threshold method, and the detection method includes the following steps:

[0068] C1: Process and collect the data detected by the LiDAR and transmit it to the cloud for storage;

[0069] C2: Clean up the information in the cloud to improve the accuracy of information collection and reduce the interference of redundant and misplaced information;

[0070] C3: According to the specific perception of the indoor environment, set the corresponding threshold value, and perceive the obstacles in the indoor environment by identifying the threshold value;

[0071] The walking mechanism adopts the Mechheim wheel walking mode to control the movement of the robot, and the Mechheim wheel controls the robot to realize turning and steering in situ during walking;

[0072] The execution module, the sensing module and the control module are connected via ports. The odometer of the sensing module calculates the angular velocity and linear velocity information of the Macheim wheels of the walking mechanism during the robot's perception of the indoor environment. The Macheim wheels are controlled by servo motors, and the execution module controls the direction control, braking and travel, i.e., reversing speed of the Macheim wheels.

[0073] Embodiment 2:

[0074] A robot indoor environment sensing device, such as Figure 2 As shown, it includes a control module, a walking mechanism, a sensing module and a buffer mechanism, the walking mechanism includes a base 1, a Macheim wheel 2 and a drive motor 13, the drive motor 13 is fixed to the outer wall of one side of the base 1 by screws, and the Macheim wheel 2 is connected to the output end of the drive motor 13 by a coupling;

[0075] The drive motor 13 can be provided to drive the Macheim wheels 2 to rotate. Since the four Macheim wheels 2 are independently driven by the corresponding drive motors 13, the drive motors 13 are used to control the Macheim wheels 2 to rotate independently, so that the four Macheim wheels 2 can realize the steering and walking of the entire device.

[0076] A main control box 3 is fixed to the top outer wall of the base 1 by bolts, and the control module and the sensing module are integrated in the main control box 3;

[0077] By integrating the control module and the sensing module into the master control box 3, the control module and the sensing module are used to control the entire device to sense the indoor environment.

[0078] The top outer wall of the base 1 is fixed with a laser radar 11 and a depth camera 12 by bolts;

[0079] By setting up a depth camera 12, the indoor environment data can be collected, effectively reducing its own energy consumption while realizing the deep collection of environmental data. The laser radar 11 is used to effectively identify obstacles and has strong anti-interference ability.

[0080] The top outer wall of the base 1 is welded with a push rod 10, and the top of the push rod 10 is welded with a top plate 4, and a buffer mechanism is installed on the top of the top plate 4, and the buffer mechanism includes a sliding column 6 and a sliding seat 8, the sliding column 6 is fixed to the top outer wall of the top plate 4 by screws, and the sliding seat 8 is slidably connected to the outer wall of the sliding column 6, and a spring 9 is clamped between the outer wall of one side of the sliding seat 8 and the inner wall of one side of the sliding column 6, and the inner wall of the sliding seat 8 is rotatably connected with a swing arm 7, and the ends of the two swing arms 7 away from the sliding seat 8 are rotatably connected with a buffer plate 5;

[0081] When the entire sensing device is moving indoors and an obstacle topples over, the obstacle is received by the buffer plate 5. The buffer plate 5 is compressed so that the slide seat 8 slides along the slide column 6 under the connection of the swing arm 7. The spring 9 plays a buffering role, thereby effectively reducing the impact of the obstacle on the entire device.

[0082] The above is a preferred specific implementation manner of the present invention, and the protection scope of the present invention is not limited thereto. Any modification, equivalent substitution and improvement made by any technician familiar with the field within the technical scope disclosed by the present invention in combination with the prior art or public common sense, within the spirit and principle of the present invention, shall be covered by the protection scope of the present invention.

Claims

1. A robot indoor environment perception method, characterized in that: The following steps are involved: S1: The robot is controlled by the control module. The control module is based on node distribution, executes decision commands, and makes decisions and controls the movement of the robot; S2: The sensing module installed on the robot obtains indoor environmental parameters such as obstacle height and size information and feeds them back to each node on the control module to support the robot's walking in the indoor environment; S3: Use the execution module installed on the robot to control the walking mechanism at the bottom of the robot to move in the direction of the generated path.

2. A robot indoor environment perception method according to claim 1, characterized in that: The control module is used to establish a system framework, on which a number of processing nodes are installed, and each processing node is connected in series.

3. A robot indoor environment perception method according to claim 2, characterized in that: The control module establishes a connection with the terminal based on a distributed communication method, and uses a map construction algorithm, a path planning algorithm, and a positioning algorithm to control the sensing module and the execution module. The specific control method of the control module includes the following steps: S11: using the sensing module to obtain the global map size and posture data information of the indoor environment; S12: After acquiring the data information, the data information is fed back to the terminal, and the terminal generates a path plan after processing, which specifically includes global planning and local planning; S13: Use the execution module to control the walking mechanism to achieve autonomous navigation.

4. A robot indoor environment perception method according to claim 3, characterized in that: The sensing module includes an image processing unit, which is based on a depth camera, an odometer and a laser radar. The depth camera is integrated with a visual sensor, which uses structured light technology. The sensing module is also integrated with a laser radar, which is used to effectively identify obstacles. The laser radar detects obstacles based on a threshold method, and its detection method includes the following steps: C1: Process and collect the data detected by the LiDAR and transmit it to the cloud for storage; C2: Clean up the information in the cloud to improve the accuracy of information collection and reduce the interference of redundant and misplaced information; C3: According to the specific perception of the indoor environment, set the corresponding threshold value, and perceive the obstacles in the indoor environment by identifying the threshold value.

5. A robot indoor environment perception method according to claim 4, characterized in that: The image processing unit processes the image captured by the depth camera based on the image processing algorithm. During the processing, the wall in the indoor environment is edge detected, feature extracted and object recognized. The edge detection algorithm is based on Canny, and the process of edge detection includes the following steps: A1: Reduce noise in edge detection process through Gaussian filtering; A2: Detect the gradient of the image and complete the edge recognition; A3: Suppress non-maximum values ​​to reduce noise in the gradient calculation process; A4: Determine the edge intensity level by detecting the double peaks; A5: After tracking the edge, connect each edge point to form a complete edge line.

6. A robot indoor environment perception method according to claim 5, characterized in that: The feature extraction is based on the SIFT algorithm, and the extraction method includes the following steps: B1: Reduce image noise through image filtering and improve the accuracy of feature extraction; B2: Identify feature points by calculating image gradients; B3: Calculate the split gradient of the image to identify feature points; B4: Provide description information for feature point description for subsequent feature matching.

7. A robot indoor environment perception device, which uses the perception method described in any one of claims 1 to 6 to perform environment perception, characterized in that: The invention comprises a control module, a walking mechanism, a sensing module and a buffer mechanism, wherein the walking mechanism comprises a base (1), a Macheim wheel (2) and a drive motor (13), wherein the drive motor (13) is fixed to an outer wall of one side of the base (1), and the Macheim wheel (2) is connected to the output end of the drive motor (13) via a coupling.

8. A robot indoor environment perception method and device according to claim 7, characterized in that: A main control box (3) is fixed to the top outer wall of the base (1) by means of bolts, and a control module and a sensing module are integrated in the main control box (3).

9. A robot indoor environment perception method and device according to claim 8, characterized in that: The top outer wall of the base (1) is welded with a push rod (10), and the top of the push rod (10) is welded with a top plate (4), and a buffer mechanism is installed on the top of the top plate (4), and the buffer mechanism comprises a sliding column (6) and a sliding seat (8), the sliding column (6) is fixed to the top outer wall of the top plate (4), and the sliding seat (8) is movably connected to the outer wall of the sliding column (6), and a spring (9) is clamped between the outer wall of one side of the sliding seat (8) and the inner wall of one side of the sliding column (6), and the inner wall of the sliding seat (8) is movably connected to a swing arm (7), and the ends of the two swing arms (7) away from the sliding seat (8) are movably connected to the buffer plate (5).

10. A robot indoor environment perception method and device according to claim 9, characterized in that: A laser radar (11) and a depth camera (12) are fixed to the top outer wall of the base (1) by means of bolts.

Images