Method for detecting negative obstacle, method for creating obstacle avoidance map, and electronic device

Abstract

The present application relates to the technical field of autonomous navigation, and in particular to a negative obstacle detection method, an obstacle avoidance map creation method and an electronic device, the detection method comprising: obtaining a current depth map collected by a collection device at a current position, and determining a pixel point with a height lower than a reference surface in the current depth map as a negative obstacle point, the reference surface being a working surface of the collection device; updating a historical negative obstacle occupancy state of a corresponding position based on a position of the negative obstacle point; obtaining a collection range of the collection device at the current position to determine a clearance range with a height lower than the reference surface; updating the historical negative obstacle occupancy state of the corresponding position based on an updated state of the historical negative obstacle occupancy state of each position in the clearance range, to obtain a current negative obstacle occupancy state corresponding to the current position, and to determine the negative obstacle. Meanwhile, the historical occupancy states of the negative obstacles detected and not detected are updated, thereby ensuring the accuracy of the current negative obstacle occupancy state.

Description

Technical Field

[0001] This invention relates to the field of autonomous navigation technology, specifically to a method for detecting negative obstacles, a method for creating obstacle avoidance maps, and an electronic device. Background Technology

[0002] When a robot operates in an environment with negative obstacles, such as steps, fractures, or stairs, it faces the risk of falling while performing autonomous navigation tasks, which can easily cause damage to the machine or even injury or death to personnel. Figure 1 As shown, the detection of negative obstacles mainly relies on infrared sensors installed under the robot to detect the ground. If the detection distance is greater than the installation height of the sensor or exceeds the detection range, it is considered that a negative obstacle has been detected.

[0003] For example, such as Figure 2 As shown, the robot has multiple infrared sensors installed on its outer edge. When a sensor is triggered, the robot brakes and performs a pre-set action based on the sensor's location (e.g., retreating when triggered in front, rotating and retreating in the opposite direction when triggered on a side surface) to leave the danger zone and mark the corresponding point as an obstacle. However, this method relies on the robot's braking distance being short enough so that it can stop in time when it detects a negative obstacle, thus preventing a fall. Summary of the Invention

[0004] In view of this, embodiments of the present invention provide a method for detecting negative obstacles, a method for creating obstacle avoidance maps, and an electronic device to solve the problem of negative obstacle detection.

[0005] According to a first aspect, embodiments of the present invention provide a method for detecting negative obstacles, comprising:

[0006] The current depth map acquired by the acquisition device at the current location is obtained, and pixels in the current depth map whose height is lower than the reference plane are identified as negative obstacle points, where the reference plane is the working surface of the acquisition device.

[0007] Based on the location of the negative obstacle point, the historical occupancy status of the corresponding negative obstacle is updated;

[0008] The acquisition range of the acquisition device at the current position is obtained to determine the clearance range where the acquisition range is below the reference plane;

[0009] Based on the updated status of the historical negative obstacle occupancy status at each location within the clearing range, the historical negative obstacle occupancy status at the corresponding location is updated to obtain the current negative obstacle occupancy status corresponding to the current location, thereby identifying the negative obstacle.

[0010] The negative obstacle detection method provided in this invention updates the historical obstacle occupancy state after detecting a negative obstacle point, indicating the presence of a negative obstacle point at the corresponding location. Simultaneously, for the acquisition device, the clearing range below the reference plane constitutes the effective detection range for negative obstacles. Within this effective detection range, the historical negative obstacle occupancy state of any previously undetected negative obstacles is updated. This update, combined with the acquisition device's acquisition range, ensures accuracy. This method updates both the historical occupancy states of detected and undetected negative obstacles simultaneously, guaranteeing the accuracy of the current negative obstacle occupancy state and enabling accurate early detection of negative obstacles.

[0011] In conjunction with the first aspect, in the first embodiment of the first aspect, updating the historical occupancy state of the corresponding negative obstacle point based on its location includes:

[0012] Acquire the external parameters of the acquisition device and the positioning information of the acquisition device on the obstacle avoidance map, wherein the obstacle avoidance map includes the historical negative obstacle occupancy status at each location;

[0013] Based on the positioning information and the extrinsic parameters, the negative obstacle point is mapped onto the obstacle avoidance map to obtain the corresponding position on the obstacle avoidance map;

[0014] Update the historical negative obstacle occupancy status of the corresponding location.

[0015] The negative obstacle detection method provided in this embodiment of the invention maps negative obstacle points onto an obstacle avoidance map by combining the positioning information of the acquisition device. This ensures that subsequent updates are performed on the corresponding locations on the obstacle avoidance map, thereby improving the accuracy of the updates.

[0016] In conjunction with the first embodiment of the first aspect, in the second embodiment of the first aspect, updating the historical negative obstacle occupancy state at the corresponding location includes:

[0017] Get the update probability;

[0018] The historical negative obstacle occupancy status of the corresponding location is updated based on the update probability.

[0019] The negative obstacle detection method provided in this invention uses update probability to update the historical negative obstacle occupancy state, which can improve the growth of the historical negative obstacle occupancy state and thus update the historical negative obstacle occupancy state more accurately.

[0020] In conjunction with the first aspect, in the third embodiment of the first aspect, obtaining the acquisition range of the acquisition device at the current position to determine the clearing range below the reference plane includes:

[0021] Obtain the attribute parameters of the acquisition device, including minimum detection distance, maximum detection distance, horizontal detection angle, and vertical detection angle;

[0022] The actual detection range of the acquisition device at the current location is determined using the attribute parameters.

[0023] The acquisition range is determined based on the area below the reference plane within the actual detection range;

[0024] Based on the external parameters and the positioning information, the collection range is mapped to the obstacle avoidance map to determine the clearing range. The obstacle avoidance map includes the historical negative obstacle occupancy status of each location.

[0025] The method for detecting negative obstacles provided in this invention uses the attribute parameters of the acquisition device to determine the actual detection range, and then compares the actual detection range with the reference plane to determine the area below the reference plane as the acquisition range. That is, the acquisition range is determined by the attribute parameters of the acquisition device, which ensures the accuracy of the subsequently determined clearing range.

[0026] In conjunction with the third embodiment of the first aspect, in the fourth embodiment of the first aspect, the step of mapping the collection range to an obstacle avoidance map based on the extrinsic parameters and the positioning information to determine the clearing range includes:

[0027] Obtain the intersection point between the acquisition range and the reference plane;

[0028] Based on the extrinsic parameters and the positioning information, each intersection point is mapped to the obstacle avoidance map to determine the intersection point position of each intersection point in the obstacle avoidance map.

[0029] The clearing range is determined based on the size relationship of each of the intersection points.

[0030] The negative obstacle detection method provided in this invention determines the clearing range by utilizing the size relationship of each intersection point, which can reduce the number of data processing steps and improve the efficiency of negative obstacle detection while ensuring the accuracy of the clearing range.

[0031] In conjunction with the first aspect, in the fifth embodiment of the first aspect, the step of updating the historical negative obstacle occupancy status of the corresponding position based on the updated status of the historical negative obstacle occupancy status of each position within the clearing range to obtain the current negative obstacle occupancy status corresponding to the current position, in order to determine the negative obstacle, includes:

[0032] For each location within the clearing range, determine whether the historical obstacle occupancy status of that location has been updated at the current location;

[0033] When the historical obstacle occupancy status of the location has not been updated at the current location, the historical obstacle occupancy status of the location is updated to obtain the current negative obstacle occupancy status corresponding to the current location, so as to identify the negative obstacle.

[0034] The negative obstacle detection method provided in this embodiment of the invention updates the historical negative obstacle occupancy state for the second time. This update can be understood as the clearing of negative obstacles. At this time, only obstacles within the clearing range are cleared, not all historical negative obstacles. The clearing range here is the effective detection range of negative obstacles, avoiding the erroneous clearing of historical negative obstacles that are not within the effective detection range, and ensuring the accuracy of the identified negative obstacles.

[0035] In conjunction with the fifth embodiment of the first aspect, in the sixth embodiment of the first aspect, when the historical obstacle occupancy state of the location has not been updated at the current location, updating the historical obstacle occupancy state of the location to obtain the current negative obstacle occupancy state corresponding to the current location, in order to determine the negative obstacle, includes:

[0036] When the historical obstacle occupancy status of the location has not been updated at the current location, obtain the clearance probability;

[0037] The historical obstacle occupancy status of the location is updated based on the clearance probability to obtain the current negative obstacle occupancy status corresponding to the current location, so as to determine the negative obstacle.

[0038] According to a second aspect, embodiments of the present invention also provide a method for creating an obstacle avoidance map, comprising:

[0039] Obtain the current negative obstacle occupancy state corresponding to the current position in the obstacle avoidance map, wherein the current negative obstacle occupancy state is determined by the negative obstacle detection method according to the first aspect of the present invention, or any embodiment of the first aspect;

[0040] A target depth map is obtained, which is determined by performing coordinate system transformation on the initial depth map acquired by the acquisition device using the intrinsic and extrinsic parameters of the acquisition device, and based on the transformation result and the height of the reference surface acquired by the acquisition device.

[0041] Based on the relationship between the pixel values ​​of each pixel in the current depth map and the corresponding pixel values ​​in the target depth map, the pixel points of the reference plane in the current depth map are determined.

[0042] Remove the pixels of the reference plane from the current depth map to determine the processed depth map;

[0043] Obstacles are detected based on the processed depth map to update the obstacle avoidance map and determine the target obstacle avoidance map.

[0044] The obstacle avoidance map creation method provided in this embodiment of the invention determines the target depth map based on the intrinsic and extrinsic parameters of the target depth acquisition device and the pixel height of the reference plane. Therefore, each pixel in the target depth map is considered to be a pixel of the reference plane. Thus, after determining the target depth map, the current depth map only needs to be compared with the target depth map to determine the pixels of the reference plane, improving the efficiency of removing pixels of the reference plane. Subsequently, the processed depth map is used to detect obstacles, detecting obstacles above the reference plane. The final obstacle avoidance map includes both negative obstacles and obstacles above the reference plane, improving the accuracy and comprehensiveness of the target obstacle avoidance map.

[0045] According to a third aspect, embodiments of the present invention provide an electronic device, including: a memory and a processor, wherein the memory and the processor are communicatively connected to each other, the memory stores computer instructions, and the processor executes the computer instructions to perform the negative obstacle detection method described in the first aspect or any embodiment of the first aspect, or to perform the obstacle avoidance map creation method described in the second aspect.

[0046] According to a fourth aspect, embodiments of the present invention provide a computer-readable storage medium storing computer instructions for causing the computer to perform the negative obstacle detection method described in the first aspect or any embodiment of the first aspect, or to perform the obstacle avoidance map creation method described in the second aspect.

[0047] It should be noted that the corresponding beneficial effects of the electronic device and computer-readable storage medium provided in the embodiments of the present invention can be found in the description of the corresponding beneficial effects of the negative obstacle detection method or the obstacle avoidance map creation method above, and will not be repeated here. Attached Figure Description

[0048] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0049] Figure 1 A schematic diagram of existing negative obstacle detection is shown;

[0050] Figure 2A schematic diagram of existing negative obstacle detection is shown;

[0051] Figure 3 This is a flowchart of a method for detecting negative obstacles according to an embodiment of the present invention;

[0052] Figure 4 This is a schematic diagram of negative obstacle detection according to an embodiment of the present invention;

[0053] Figure 5 This is a flowchart of a method for detecting negative obstacles according to an embodiment of the present invention;

[0054] Figure 6 This is a flowchart of a method for detecting negative obstacles according to an embodiment of the present invention;

[0055] Figure 7 This is a schematic diagram of the clearing range according to an embodiment of the present invention;

[0056] Figure 8 This is a flowchart of a method for creating an obstacle avoidance map according to an embodiment of the present invention;

[0057] Figure 9 This is a structural block diagram of a negative obstacle detection device according to an embodiment of the present invention;

[0058] Figure 10 This is a structural block diagram of an obstacle avoidance map creation device according to an embodiment of the present invention;

[0059] Figure 11 This is a schematic diagram of the hardware structure of the electronic device provided in an embodiment of the present invention. Detailed Implementation

[0060] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0061] The negative obstacle detection method provided in this invention updates the historical negative obstacle occupancy state using the current depth map. This involves fusing multiple frames of the current depth map, ensuring that the updated current negative obstacle occupancy state retains both historical and current depth map information. This guarantees the accuracy of the obtained current negative obstacle occupancy state, enabling early detection of negative obstacles and providing accurate negative obstacle information for the robot's autonomous obstacle avoidance. The historical negative obstacle occupancy state represents the occupancy status of each location point by a negative obstacle. This can be represented by probability values, identifier values, etc. No specific limitations are placed on the specific representation of the obstacle occupancy state.

[0062] Furthermore, this embodiment of the invention also provides a method for creating an obstacle avoidance map. The obstacle avoidance map created by this method includes not only negative obstacle information but also obstacle information above the reference plane. The detection of negative obstacle information is determined using the aforementioned negative obstacle detection method. For obstacle information above the reference plane, the pixel values ​​of each pixel in the current depth map and the target depth map are compared. First, pixels corresponding to the reference plane in the current depth map are removed. Then, the depth map with the reference plane pixels removed is used for obstacle identification. Here, obstacle identification refers to the identification of obstacles above the reference plane. All pixels in the target depth map are pixels of the reference plane. The resulting obstacle avoidance map includes both negative obstacle information and obstacle information above the reference plane. Electronic devices, such as self-propelled robots or robotic vacuum cleaners, can use this obstacle avoidance map to achieve autonomous navigation, collision avoidance, and fall protection.

[0063] As an optional application scenario of this invention, the negative obstacle detection method is implemented in real-time by the robot during operation. Specifically, the robot includes a data acquisition device, a target carrier, and a processing device, with the data acquisition device fixedly mounted on the target carrier. During operation, the data acquisition device acquires the current depth map, and the processing device analyzes the current depth map by executing the negative obstacle detection method described in this invention to identify negative obstacles. After identifying the negative obstacles, the movement of the target carrier is controlled to perform obstacle avoidance.

[0064] As another optional application scenario of this invention, unlike the first application scenario, the robot is connected to other backend terminals. The robot includes a data acquisition device and a target carrier. The data acquisition device sends the acquired depth map to the other backend terminals. The other backend terminals analyze the depth map by executing the negative obstacle detection method described in this embodiment of the invention, determine the information of the negative obstacles, and send the information of the negative obstacles to the robot. Subsequently, the robot performs obstacle avoidance processing based on the information of the negative obstacles.

[0065] Of course, the above application scenarios are merely examples and do not limit the scope of protection of this invention. Specific application scenarios are set according to actual needs, and no limitations are made here.

[0066] According to embodiments of the present invention, a method for detecting negative obstacles and an embodiment of a method for creating obstacle avoidance maps are provided. It should be noted that the steps shown in the flowcharts in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions. Furthermore, although a logical order is shown in the flowcharts, in some cases, the steps shown or described may be executed in a different order than that shown here.

[0067] This embodiment provides a method for detecting negative obstacles, which can be used in the aforementioned electronic devices, such as mobile robots, robotic vacuum cleaners, mobile terminals, servers, etc. Figure 3 This is a flowchart of a method for detecting negative obstacles according to an embodiment of the present invention, such as... Figure 3 As shown, the process includes the following steps:

[0068] S11, acquire the current depth map acquired by the acquisition device at the current location, and determine the pixels in the current depth map whose height is lower than the reference plane as negative obstacle points.

[0069] The reference plane is the working surface of the acquisition device.

[0070] The depth maps acquired by the acquisition device differ depending on the location. At the current location, the acquisition device acquires a depth map for that location, thus obtaining the current depth map. The timing of this depth map acquisition can be at regular intervals, or when the device moves to a specific location, etc.; no specific restrictions are imposed here.

[0071] The reference plane is the working surface of the data acquisition device. For example, if the device operates on the ground, the ground is the reference plane; if it operates on a desktop, the desktop is the reference plane, and so on. The electronic device analyzes the acquired current depth map and identifies pixels whose height is lower than the reference plane, marking them as negative obstacle points. The pixel value of each pixel in the current depth map represents the height of its corresponding location, while the pixels on the reference plane are predetermined by the data acquisition device. By comparing the pixel values ​​of each individual pixel with those on the reference plane, negative obstacle points can be identified.

[0072] For example, the current depth map can be converted into a point cloud in the camera coordinate system. By performing planar fitting on the point cloud, pixels in areas below the reference plane can be identified as negative obstacle points. Alternatively, negative obstacle points below the reference plane can be segmented based on methods such as height and planar distance from the reference plane. No restrictions are placed on the method for determining negative obstacle points.

[0073] S12, based on the location of the negative obstacle point, update the historical occupancy status of the corresponding negative obstacle.

[0074] As mentioned above, the historical negative obstacle occupancy status indicates the situation where a corresponding location point is occupied by a negative obstacle. When an electronic device detects the existence of a negative obstacle point at a certain location, it needs to update the historical negative obstacle occupancy status corresponding to that location. For example, for location point A, its historical negative obstacle occupancy status is 9. Then, when it is determined that location point A has a negative obstacle point, the updated negative obstacle occupancy status is 9 + 2 = 11. It should be noted that if multiple negative obstacle points correspond to the same location at the current position, then the historical negative obstacle occupancy status of that location is only updated once.

[0075] The update method for the historical negative obstacle occupancy status could be to add a certain value to the historical obstacle occupancy status, or to update the historical obstacle occupancy status with a certain probability, and so on. No specific restrictions are placed on the specific update method here.

[0076] The specifics of this step will be described in detail below.

[0077] S13, obtain the acquisition range of the acquisition device at the current position to determine the clearing range where the acquisition range is lower than the reference plane.

[0078] Each data acquisition device has its own corresponding acquisition range, or field of view. Only within the field of view can the presence of obstacles be determined; outside the field of view, it is difficult to identify obstacles. Furthermore, the effective detection range for negative obstacles is the area within the acquisition range that is below the reference plane; only detection data within this range is valid.

[0079] The acquisition range of a data acquisition device is determined by its attributes; once these attributes are determined, its actual field of view is also determined. Consequently, the clearing range at different locations is also determined.

[0080] like Figure 4 As shown, the clearing range below the reference plane is... Figure 4The region is defined by A, B, C, D, and the two vertices below the reference plane. Specifically, after acquiring the acquisition range of the current position, the electronic device compares the acquisition range with the reference plane to determine the clearing range below the reference plane.

[0081] S14. Based on the updated status of the historical negative obstacle occupancy status of each location within the clearing range, update the historical negative obstacle occupancy status of the corresponding location to obtain the current negative obstacle occupancy status corresponding to the current location, so as to determine the negative obstacle.

[0082] At the same location, the historical negative obstacle occupancy status needs to be updated twice. The first update occurs after a negative obstacle is detected, by increasing the occupancy value or probability. The second update, based on the acquisition range of the data acquisition device, clears the historical negative obstacle occupancy status corresponding to non-negative obstacles within the clearing range, by decreasing the occupancy value or probability. Therefore, the update described in this step can also be understood as a clearing update. It should be noted that a clearing update does not completely remove the historical negative obstacle occupancy status, but rather reduces it with a certain probability or a certain value.

[0083] As mentioned above, the historical negative obstacle occupancy status needs to be updated twice. However, the two updates are not for the same historical negative obstacle occupancy status, but for the historical negative obstacle occupancy status at different locations.

[0084] After obtaining the current state of negative obstacle occupancy at its current location, the electronic device can determine the detectable negative obstacles at that location to form an obstacle avoidance map, or perform obstacle avoidance processing based on this map, and so on. The specific applications after detecting negative obstacles are not limited here; they can be configured according to actual needs.

[0085] The specifics of this step will be described in detail below.

[0086] The negative obstacle detection method provided in this embodiment updates the historical obstacle occupancy status after detecting a negative obstacle point, indicating the presence of a negative obstacle point at the corresponding location. Simultaneously, for the acquisition device, the clearing range below the reference plane constitutes the effective detection range for negative obstacles. Within this effective detection range, the historical negative obstacle occupancy status of previously undetected obstacles is updated. This update, combined with the acquisition device's acquisition range, ensures accuracy. This method updates both the historical occupancy status of detected and undetected negative obstacles simultaneously, guaranteeing the accuracy of the current negative obstacle occupancy status and enabling accurate early detection of negative obstacles.

[0087] This embodiment provides a method for detecting negative obstacles, which can be used in the aforementioned electronic devices, such as mobile robots, robotic vacuum cleaners, mobile terminals, servers, etc. Figure 5 This is a flowchart of a method for detecting negative obstacles according to an embodiment of the present invention, such as... Figure 5 As shown, the process includes the following steps:

[0088] S21, acquire the current depth map acquired by the acquisition device at the current location, and determine the pixels in the current depth map whose height is lower than the reference plane as negative obstacle points.

[0089] The reference plane is the working surface of the acquisition device.

[0090] Please see details Figure 3 S11 of the illustrated embodiment will not be described again here.

[0091] S22, based on the location of the negative obstacle point, update the historical occupancy status of the corresponding negative obstacle.

[0092] Specifically, S22 includes:

[0093] S221, acquire the external parameters of the acquisition device and the positioning information of the acquisition device on the obstacle avoidance map.

[0094] The obstacle avoidance map includes the historical negative obstacle occupancy status at each location.

[0095] For the data acquisition device, its location information on the obstacle avoidance map can be determined using the device's laser positioning algorithm, visual positioning algorithm, or other positioning algorithms. The extrinsic parameters of the data acquisition device are obtained through extrinsic parameter calibration, and no limitations are imposed on them here.

[0096] The obstacle avoidance map can be a grid map, where each grid cell represents an actual location point. The map also includes the historical negative obstacle occupancy status for each location, determined using historical depth maps.

[0097] S222: Based on the positioning information and extrinsic parameters, the negative obstacle points are mapped onto the obstacle avoidance map to obtain the corresponding positions on the obstacle avoidance map.

[0098] The electronic device uses positioning information and extrinsic parameters from the acquisition device to map negative obstacle points to an obstacle avoidance map. For example, the electronic device uses the extrinsic and intrinsic parameters of the acquisition device to determine the negative obstacle point cloud corresponding to each negative obstacle point. Then, it uses the positioning information and extrinsic parameters to perform an intermediate transformation on the negative obstacle point cloud, obtaining the transformation points corresponding to each point in the negative obstacle point cloud. Finally, it uses the resolution of the obstacle avoidance map to map these transformation points onto the obstacle avoidance map, obtaining the corresponding positions on the map. The resolution of the obstacle avoidance map is determined based on the obstacle avoidance accuracy, for example, set to 0.05m or 0.1m.

[0099] For example, the location information is T i The external parameter is T c Each point cloud P in the negative obstacle point cloud P i The corresponding conversion point pm i For: pm i =T i T c p i .

[0100] Let (x0, y0, z0) be the positioning coordinates corresponding to the first grid cell of the obstacle avoidance map, r be the resolution of the obstacle avoidance map, and let the transition point pm be... i If the coordinates are (x, y, z), then mapping them onto the obstacle avoidance map will yield the corresponding position (x, y, z) on the obstacle avoidance map. i y i , z i )for:

[0101] x i = (x-x0) / r;

[0102] y i = (y-y0) / r;

[0103] z i = (z-z0) / r.

[0104] S223, update the historical negative obstacle occupancy status of the corresponding position.

[0105] Once the corresponding location is determined, it means that a negative obstacle exists at that location on the obstacle avoidance map. Therefore, it is necessary to update the historical negative obstacle occupancy status of the corresponding location on the obstacle avoidance map. For example, the negative obstacle occupancy status can be represented numerically. Whenever a negative obstacle is detected, the historical negative obstacle occupancy value of the corresponding location is increased by C1, where the specific value of C1 is set according to actual needs. Alternatively, the historical negative obstacle occupancy status of the corresponding location can be updated with a certain probability P1, and so on.

[0106] In some alternative implementations, S223 includes:

[0107] (1) Obtain the update probability.

[0108] (2) Update the historical negative obstacle occupancy status of the corresponding position based on the update probability.

[0109] Electronic devices use updated probabilities to update the corresponding position (x) i y i , z i The historical negative obstacle occupancy status is updated, for example, by using Bayes' theorem to update the obstacle avoidance map at the corresponding location (x). i y i , z i The historical negative obstacle occupancy status is updated with update probability.

[0110] Updating the historical negative obstacle occupancy state using update probability can improve the growth of the historical negative obstacle occupancy state, thus enabling a more accurate update of the historical negative obstacle occupancy state.

[0111] S23, obtain the acquisition range of the acquisition device at the current position to determine the clearing range where the acquisition range is lower than the reference plane.

[0112] Please see details Figure 3 S13 of the illustrated embodiment will not be described again here.

[0113] S24, based on the updated status of the historical negative obstacle occupancy status of each location within the clearing range, update the historical negative obstacle occupancy status of the corresponding location to obtain the current negative obstacle occupancy status corresponding to the current location, so as to determine the negative obstacle.

[0114] Please see details Figure 3 S14 of the illustrated embodiment will not be described again here.

[0115] The negative obstacle detection method provided in this embodiment maps negative obstacle points onto the obstacle avoidance map by combining the positioning information of the acquisition device. This ensures that subsequent updates are performed on the corresponding locations on the obstacle avoidance map, thus improving the accuracy of the updates.

[0116] This embodiment provides a method for detecting negative obstacles, which can be used in the aforementioned electronic devices, such as mobile robots, robotic vacuum cleaners, mobile terminals, servers, etc. Figure 6 This is a flowchart of a method for detecting negative obstacles according to an embodiment of the present invention, such as... Figure 6 As shown, the process includes the following steps:

[0117] S31, acquire the current depth map acquired by the acquisition device at the current location, and determine the pixels in the current depth map whose height is lower than the reference plane as negative obstacle points.

[0118] The reference plane is the working surface of the acquisition device.

[0119] Please see details Figure 3 S11 of the illustrated embodiment will not be described again here.

[0120] S32, based on the location of the negative obstacle point, update the historical occupancy status of the corresponding negative obstacle.

[0121] Please see details Figure 5 S22 of the illustrated embodiment will not be described again here.

[0122] S33, Obtain the acquisition range of the acquisition device at the current location to determine the clearing range where the acquisition range is lower than the reference plane.

[0123] Specifically, S33 includes:

[0124] S331, Obtain the attribute parameters of the acquisition device.

[0125] The attribute parameters include minimum detection distance, maximum detection distance, horizontal detection angle, and vertical detection angle.

[0126] The attribute parameters of the data acquisition device are all determined at the time of manufacture of the target device and can be stored in the electronic device, etc.

[0127] S332 uses attribute parameters to determine the actual detection range of the acquisition device at the current location.

[0128] The actual detection range of the acquisition device at the current location can be expressed as follows: Figure 4 The hexahedron shown can be represented by its eight vertices. That is, the actual detection range includes the vertices of the actual detection area.

[0129] S333 determines the acquisition range based on the area below the reference plane within the actual detection range.

[0130] The actual detection range is compared with a reference plane, and the area below the reference plane within the actual detection range is determined as the acquisition range. For example, Figure 4 The area enclosed by A, B, C, D, and the two vertices below the reference plane.

[0131] S334, based on external parameters and positioning information, maps the collection range to the obstacle avoidance map to determine the clearing range.

[0132] The obstacle avoidance map includes the historical negative obstacle occupancy status of each location.

[0133] Electronic devices utilize the external parameters T of the acquisition device c and location information T iTake the four intersection points, for example, Figure 4 Mapping A, B, C, and D onto the obstacle avoidance map determines the clearance area; for example, we get: Figure 7 The area to be cleared is shown. Among them, Figure 7 Each grid in the image represents a grid in the obstacle avoidance map, and the trapezoidal area represents the defined clearance range.

[0134] In some alternative implementations, S334 includes:

[0135] (1) Obtain the intersection of the acquisition range and the reference plane.

[0136] (2) Based on the external parameters and positioning information, each intersection point is mapped to the obstacle avoidance map to determine the intersection point position of each intersection point in the obstacle avoidance map.

[0137] (3) Determine the clearing range based on the size relationship of each intersection point.

[0138] The intersection of the acquisition range and the reference plane is determined by comparing the hexahedron corresponding to the actual detection range with the reference plane. The acquisition device maps the intersection of the acquisition range and the reference plane onto the obstacle avoidance map using extrinsic parameters and positioning information, thus obtaining the intersection positions of each point on the obstacle avoidance map. The specific mapping method is similar to the method of mapping negative obstacle points to the obstacle avoidance map described in S222 above.

[0139] After obtaining the intersection points on the obstacle avoidance map, the electronic device finds the maximum and minimum positions among each intersection point. For example, the data acquisition device finds the maximum x-coordinate among the four intersection points on the obstacle avoidance map. max y max z max and the smallest x min y min z min , the largest x max y max z max And the smallest x min y min z min The enclosed area is defined as the clearance zone.

[0140] By using the size relationship of each intersection point to determine the clearing range, the number of data processing operations can be reduced while ensuring the accuracy of the clearing range, thus improving the efficiency of negative obstacle detection.

[0141] S34, based on the updated state of the historical negative obstacle occupancy state of each location within the clearing range, update the historical negative obstacle occupancy state of the corresponding location to obtain the current negative obstacle occupancy state corresponding to the current location, so as to determine the negative obstacle.

[0142] Specifically, S34 includes:

[0143] S341, For each location within the clearing range, determine whether the historical obstacle occupancy status of the location has been updated at the current location.

[0144] The electronic device determines whether the historical negative obstacle occupancy status of each location within the clearing range has been updated. That is, it determines whether the historical negative obstacle occupancy status of each location has been updated in the above S32. If it has been updated, it means that the location is occupied by a negative obstacle; if it has not been updated, it means that the location is not occupied by an obstacle.

[0145] S342, when the historical obstacle occupancy status of the location has not been updated at the current location, update the historical obstacle occupancy status of the location to obtain the current negative obstacle occupancy status corresponding to the current location, so as to determine the negative obstacle.

[0146] As mentioned above, the clearing range represents the effective detection range of negative obstacle points. If the historical negative obstacle occupancy state of a certain position within this range is not updated in S32, it means that the position is currently not occupied by an obstacle, and therefore the historical obstacle occupancy state of that position needs to be cleared and updated. That is, all positions are traversed within the clearing range, and if the position is not updated in S32, the historical negative obstacle occupancy state of that position is cleared and updated to obtain the current negative obstacle occupancy state corresponding to the current position, thereby identifying the obstacle.

[0147] In some alternative implementations, S342 includes:

[0148] (1) Obtain the clearance probability when the historical obstacle occupancy status of the location has not been updated at the current location.

[0149] (2) Update the historical obstacle occupancy status of the location based on the clearance probability to obtain the current negative obstacle occupancy status corresponding to the current location, so as to determine the negative obstacle.

[0150] The erasure probability can be determined based on the data quality of the acquisition device and the actual experimental results; no specific numerical limit is imposed on it here. Of course, the erasure probability can be updated according to different application scenarios.

[0151] Electronic devices use the clearing probability to determine the corresponding position (x) j y j , z j The historical negative obstacle occupancy status is updated, for example, by using Bayes' theorem to update the corresponding position (x) on the obstacle avoidance map. j y j , zj The historical negative obstacle occupancy status is updated with the clearance probability.

[0152] The negative obstacle detection method provided in this embodiment uses the attribute parameters of the acquisition device to determine the actual detection range, and then compares the actual detection range with a reference plane. The area below the reference plane is determined as the acquisition range. That is, the acquisition range is determined using the attribute parameters of the acquisition device, ensuring the accuracy of the subsequently determined clearing range. When updating the historical negative obstacle occupancy status for the second time, this update can be understood as the clearing of negative obstacles. At this time, only obstacles within the clearing range are cleared, not all historical negative obstacles. The clearing range here is the effective detection range of negative obstacles, avoiding the erroneous clearing of historical negative obstacles that are not within the effective detection range, thus ensuring the accuracy of the determined negative obstacles.

[0153] This embodiment provides a method for detecting negative obstacles, which can be used in the aforementioned electronic devices, such as mobile robots, robotic vacuum cleaners, mobile terminals, servers, etc. Figure 8 This is a flowchart of a method for detecting negative obstacles according to an embodiment of the present invention, such as... Figure 8 As shown, the process includes the following steps:

[0154] S41, obtain the current negative obstacle occupancy status corresponding to the current position in the obstacle avoidance map.

[0155] The current negative obstacle occupancy state is determined according to the aforementioned negative obstacle detection method.

[0156] The current state of negative obstacle occupancy corresponds to each position of the current location. Each position of the current location has a corresponding definite location in the obstacle avoidance map. Therefore, the current state of negative obstacle occupancy corresponding to the current position in the obstacle avoidance map can be determined by using the above-mentioned negative obstacle detection method.

[0157] In some alternative implementations, S41 includes:

[0158] S411, Obtain the initial depth map.

[0159] The initial depth map is represented as I(u, v) = d, where u and v are image coordinates and d is the pixel value, i.e., the measured distance value.

[0160] S412 uses intrinsic parameters to convert the initial depth map into the first point cloud data in the camera coordinate system.

[0161] The internal parameters of the target depth camera: P = (fx, fy, cx, cy), where fx and fy are the focal lengths of the camera in the X and Y directions, and cx and cy are the offsets of the camera optical axis from the center of the projection plane coordinates. The electronic device uses the internal parameters to convert the initial depth map into the first point cloud data p = [x, y, z] in the camera coordinate system. T ,

[0162]

[0163] S413. Use the external parameters to convert the first point cloud data into the second point cloud data in the target carrier coordinate system.

[0164] The first point cloud data in the camera coordinate system can be converted through the external parameters (i.e., the relative position of the installation position with respect to the origin of the target carrier coordinate system) to obtain the point cloud in the target carrier coordinate system, that is, the second point cloud data.

[0165] Among them, the external parameter T:

[0166]

[0167] The external parameters can be obtained through camera external parameter calibration, and the external parameter calibration is the rotation and translation matrix of the relationship between the camera coordinate system and the system world coordinate system.

[0168] The second point cloud data in the target carrier coordinate system is represented as P b = TP, that is, the relationship between the second point cloud data in the target carrier coordinate system and the initial depth map I(u, v) = d can be obtained:

[0169]

[0170] S414. Extract the height point cloud data from the second point cloud data.

[0171] Among them, the height point cloud data in the second point cloud data is the relationship between the height and the internal parameters, external parameters, and pixel values of the initial depth map, that is,

[0172]

[0173] S415. Determine the target depth map based on the relationship between the height point cloud data and the height of the reference plane.

[0174] According to the data characteristics of the acquisition device, set that when the point cloud height < the threshold height h, it is considered that the point belongs to the point of the reference plane, that is, a point is removed when z b < h. For this reason, the value corresponding to each pixel when all points in the depth map are ground points can be calculated, so as to obtain the target depth map. Among them, h is the height of the reference plane.

[0175] In some alternative implementations, S415 may include:

[0176] (1) Set the height in the height point cloud data to the height of the reference plane, and obtain the relationship between the height of the reference plane and the intrinsic parameters, extrinsic parameters and the pixel values ​​of the initial depth map.

[0177] (2) The target depth map is calculated based on the relationship between the height of the reference plane and the intrinsic, extrinsic parameters and the pixel values ​​of the initial depth map.

[0178] That is, by using the relationship between the height of the reference plane and the intrinsic and extrinsic parameters, as well as the pixel values ​​of the initial depth map, the relationship between the pixel values ​​of the target depth map and the height of the reference plane and the intrinsic and extrinsic parameters is calculated, so as to determine the target depth map.

[0179] In the above formula, the height in the elevation point cloud data is set to h, which is the reference plane height, thus obtaining:

[0180]

[0181] This leads to the following relationship for d:

[0182]

[0183] As mentioned above, since each pixel in the target depth map is a pixel on the reference plane, its pixel value is the value of d, thus obtaining the target depth map. :

[0184]

[0185] As can be seen from the target depth map, it is a threshold map with h as the height threshold, where h is the height of the reference plane.

[0186] Because of the error in the target depth camera measurement, the height scanned to the reference plane will usually fluctuate within the range of 0. The magnitude of the camera's error on the reference plane is used as the basis for conversion. Points with heights within this range are removed to obtain an accurate target depth map.

[0187] S42, Obtain the target depth map.

[0188] The target depth map is determined by transforming the initial depth map acquired by the acquisition device using its intrinsic and extrinsic parameters, and then using the transformation result and the height of the reference surface acquired by the acquisition device.

[0189] As long as the installation position of the acquisition device remains unchanged, the target depth map will remain unchanged and does not need to be updated. Specifically, it is sufficient that the z, roll, and pitch values ​​in the acquisition device's attitude (x, y, z, roll, pitch, yaw) in the target carrier coordinate system remain unchanged.

[0190] The target depth map can be acquired and stored by the electronic device from external sources, or it can be determined by the electronic device when an obstacle avoidance map needs to be created, etc. After the data acquisition device is deployed, it is controlled to acquire an initial depth map. The acquisition orientation and timing of the initial depth map are set according to actual needs. After acquiring the initial depth map, the coordinate system of the initial depth map is transformed using the intrinsic and extrinsic parameters of the data acquisition device. Based on the transformation result and the height of the reference plane, the target depth map is determined.

[0191] S43, Based on the relationship between the pixel values ​​of each pixel in the current depth map and the corresponding pixel values ​​in the target depth map, determine the pixel points of the reference plane in the current depth map.

[0192] After acquiring the target depth map, the electronic device compares the pixel values ​​of each pixel in the current depth map with the corresponding pixels in the target depth map to determine the pixels of the reference plane in the current depth map. In other words, a height threshold is used to remove pixels from the reference plane.

[0193] For example, S43 above may include:

[0194] (1) Compare the pixel values ​​of each pixel in the current depth map with the pixel values ​​of the corresponding pixels in the target depth map in turn.

[0195] (2) When the pixel value of the target pixel in the current depth map is less than the pixel value of the corresponding pixel in the target depth map, the target pixel is determined as the pixel of the reference plane.

[0196] As mentioned above, all pixels in the target depth image are considered to be pixels of the reference plane. If the pixel value of a target pixel in the current depth image is less than the pixel value of the corresponding pixel in the target depth image, it means that the target pixel is below the reference plane, i.e., the target pixel is considered to be a pixel of the reference plane. It should be noted that pixels of the reference plane include pixels on the reference plane as well as pixels below the reference plane.

[0197] S44 removes the pixels of the reference plane from the current depth map and determines the processed depth map.

[0198] After identifying the pixels of a reference surface, the electronic device uses the pixel's position in the depth map to remove those pixels from the depth map. For example, the pixel value of the reference surface can be set to zero or another pixel value in the depth map. The specific removal method is set according to actual needs.

[0199] The processed depth map is obtained by removing the pixels of the reference plane from the depth map. This processed depth map can be used for obstacle recognition.

[0200] S45, based on the processed depth map, performs obstacle detection to update the obstacle avoidance map and determine the target obstacle avoidance map.

[0201] After obtaining the processed depth map, the electronic device can use target recognition, target detection, or image processing to detect obstacles, thereby updating the obstacle avoidance map for obstacles above the reference plane and thus determining the target obstacle avoidance map.

[0202] Furthermore, to further ensure the accuracy of the target obstacle avoidance map, the electronic device can fuse the obstacle avoidance maps determined by the acquisition results from different angles at the same location to obtain the target obstacle avoidance map.

[0203] The obstacle avoidance map creation method provided in this embodiment determines the target depth map based on the intrinsic and extrinsic parameters of the target depth acquisition device and the pixel height of the reference plane. Therefore, each pixel in the target depth map is considered to be a pixel of the reference plane. Thus, after determining the target depth map, the current depth map only needs to be compared with the target depth map to determine the pixels of the reference plane, improving the efficiency of removing reference plane pixels. Subsequently, the processed depth map is used to detect obstacles, detecting obstacles above the reference plane. The final obstacle avoidance map includes both negative obstacles and obstacles above the reference plane, improving the accuracy and comprehensiveness of the target obstacle avoidance map.

[0204] This embodiment also provides a negative obstacle detection device and an obstacle avoidance map creation device, which are used to implement the above embodiments and preferred embodiments, and will not be repeated as already described. As used below, the term "module" can be a combination of software and / or hardware that implements a predetermined function. Although the devices described in the following embodiments are preferably implemented in software, hardware implementation, or a combination of software and hardware, is also possible and contemplated.

[0205] This embodiment provides a device for detecting negative obstacles, such as... Figure 9 As shown, it includes:

[0206] The first acquisition module 51 is used to acquire the current depth map acquired by the acquisition device at the current position, and to determine the pixels in the current depth map whose height is lower than the reference plane as negative obstacle points, wherein the reference plane is the working surface of the acquisition device;

[0207] The first update module 52 is used to update the historical occupancy status of the corresponding negative obstacle point based on the location of the negative obstacle point;

[0208] The second acquisition module 53 is used to acquire the acquisition range of the acquisition device at the current position, so as to determine the clearing range where the acquisition range is lower than the reference plane;

[0209] The second update module 54 is used to update the historical negative obstacle occupancy status of the corresponding position based on the update status of the historical negative obstacle occupancy status of each position within the clearing range, so as to obtain the current negative obstacle occupancy status corresponding to the current position and thus determine the negative obstacle.

[0210] In some alternative implementations, the first update module 52 includes:

[0211] The first acquisition unit is used to acquire the external parameters of the acquisition device and the positioning information of the acquisition device on the obstacle avoidance map, wherein the obstacle avoidance map includes the historical negative obstacle occupancy status at each location;

[0212] The first mapping unit is used to map the negative obstacle point onto the obstacle avoidance map based on the positioning information and the extrinsic parameters, so as to obtain the corresponding position on the obstacle avoidance map.

[0213] The first update unit is used to update the historical negative obstacle occupancy status of the corresponding location.

[0214] In some alternative implementations, the first update unit includes:

[0215] The first acquisition subunit is used to acquire the update probability;

[0216] The first update subunit is used to update the historical negative obstacle occupancy state of the corresponding position based on the update probability.

[0217] In some alternative implementations, the second acquisition module 53 includes:

[0218] The second acquisition unit is used to acquire the attribute parameters of the acquisition device, including the minimum detection distance, the maximum detection distance, the horizontal detection angle, and the vertical detection angle.

[0219] The first determining unit is used to determine the actual detection range of the acquisition device at the current location using the attribute parameters;

[0220] The second determining unit is used to determine the acquisition range based on the area below the reference plane within the actual detection range;

[0221] The second mapping unit is used to map the collection range to an obstacle avoidance map based on the external parameters and the positioning information, and to determine the clearing range. The obstacle avoidance map includes the historical negative obstacle occupancy status of each location.

[0222] In some alternative implementations, the second mapping unit includes:

[0223] The second acquisition subunit is used to acquire the intersection point of the acquisition range and the reference plane;

[0224] The mapping subunit is used to map each intersection point to the obstacle avoidance map based on the extrinsic parameters and the positioning information, and to determine the intersection point position of each intersection point in the obstacle avoidance map.

[0225] A subunit is determined to determine the clearing range based on the size relationship of each intersection point.

[0226] In some alternative implementations, the second update module 54 includes:

[0227] The third determining unit is used to determine, for each location within the clearing range, whether the historical obstacle occupancy status of the location has been updated at the current location;

[0228] The second update unit is used to update the historical obstacle occupancy status of the location when the historical obstacle occupancy status of the location has not been updated at the current location, so as to obtain the current negative obstacle occupancy status corresponding to the current location and to determine the negative obstacle.

[0229] In some alternative implementations, the second updating unit includes:

[0230] The third acquisition subunit is used to acquire the clearance probability when the historical obstacle occupancy status of the location has not been updated at the current location;

[0231] The second update subunit is used to update the historical obstacle occupancy status of the location based on the clearance probability, so as to obtain the current negative obstacle occupancy status corresponding to the current location and thus determine the negative obstacle.

[0232] This embodiment provides a device for creating an obstacle avoidance map, such as... Figure 10 As shown, it includes:

[0233] The third acquisition module 61 is used to acquire the current negative obstacle occupancy status corresponding to the current position in the obstacle avoidance map. The current negative obstacle occupancy status is determined by the negative obstacle detection method according to the first aspect of the present invention or any embodiment of the first aspect.

[0234] The fourth acquisition module 62 is used to acquire a target depth map, which is determined by performing coordinate system transformation on the initial depth map acquired by the acquisition device using the intrinsic and extrinsic parameters of the acquisition device, and based on the transformation result and the height of the reference surface acquired by the acquisition device.

[0235] The determining module 63 is used to determine the pixel points of the reference plane in the current depth map based on the relationship between the pixel values ​​of each pixel point in the current depth map and the pixel values ​​of the corresponding pixel points in the target depth map;

[0236] The removal module 64 is used to remove the pixels of the reference plane from the current depth map and determine the processed depth map;

[0237] The detection module 65 is used to detect obstacles based on the processed depth map in order to update the obstacle avoidance map and determine the target obstacle avoidance map.

[0238] In this embodiment, the obstacle detection device and the obstacle avoidance map creation device are presented in the form of functional units. Here, a unit refers to an ASIC circuit, a processor and memory that execute one or more software or fixed programs, and / or other devices that can provide the above functions.

[0239] Further functional descriptions of the above modules are the same as those in the corresponding embodiments described above, and will not be repeated here.

[0240] This invention also provides an electronic device having the above-described features. Figure 9 The device for detecting negative obstacles shown, or Figure 10 The device shown is for creating obstacle avoidance maps.

[0241] Please see Figure 11 , Figure 11 This is a schematic diagram of the structure of an electronic device provided in an optional embodiment of the present invention, such as... Figure 11 As shown, the electronic device may include: at least one processor 71, such as a CPU (Central Processing Unit), at least one communication interface 73, memory 74, and at least one communication bus 72. The communication bus 72 is used to enable communication between these components. The communication interface 73 may include a display screen and a keyboard; optionally, the communication interface 73 may also include a standard wired interface or a wireless interface. The memory 74 may be high-speed RAM (Random Access Memory) or non-volatile memory, such as at least one disk storage device. Optionally, the memory 74 may also be at least one storage device located remotely from the aforementioned processor 71. The processor 71 may be combined with... Figure 9 or Figure 10The described apparatus has an application program stored in memory 74, and the processor 71 calls the program code stored in memory 74 to perform any of the above method steps.

[0242] The communication bus 72 can be a peripheral component interconnect (PCI) bus or an extended industry standard architecture (EISA) bus, etc. The communication bus 72 can be divided into an address bus, a data bus, and a control bus, etc. For ease of representation, Figure 11 The bus is represented by a single thick line, but this does not mean that there is only one bus or one type of bus.

[0243] The memory 74 may include volatile memory, such as random-access memory (RAM); the memory may also include non-volatile memory, such as flash memory, hard disk drive (HDD) or solid-state drive (SSD); the memory 74 may also include a combination of the above types of memory.

[0244] The processor 71 can be a central processing unit (CPU), a network processor (NP), or a combination of CPU and NP.

[0245] The processor 71 may further include a hardware chip. This hardware chip may be an application-specific integrated circuit (ASIC), a programmable logic device (PLD), or a combination thereof. The PLD may be a complex programmable logic device (CPLD), a field-programmable gate array (FPGA), a generic array logic (GAL), or any combination thereof.

[0246] Optionally, the memory 74 is also used to store program instructions. The processor 71 can invoke the program instructions to implement the negative obstacle detection method or the obstacle avoidance map creation method as shown in any embodiment of this application.

[0247] This invention also provides a non-transitory computer storage medium storing computer-executable instructions that can execute the negative obstacle detection method or the obstacle avoidance map creation method in any of the above method embodiments. The storage medium can be a magnetic disk, optical disk, read-only memory (ROM), random access memory (RAM), flash memory, hard disk drive (HDD), or solid-state drive (SSD), etc.; the storage medium may also include combinations of the above types of memory.

[0248] Although embodiments of the invention have been described in conjunction with the accompanying drawings, those skilled in the art can make various modifications and variations without departing from the spirit and scope of the invention, and such modifications and variations all fall within the scope defined by the appended claims.

Claims

1. A method for detecting negative obstacles, characterized in that, include: The current depth map acquired by the acquisition device at the current location is obtained, and pixels in the current depth map whose height is lower than the reference plane are identified as negative obstacle points, where the reference plane is the working surface of the acquisition device. Based on the location of the negative obstacle point, the historical occupancy status of the corresponding negative obstacle is updated; The acquisition range of the acquisition device at the current position is obtained to determine the clearance range where the acquisition range is below the reference plane; Based on the updated status of the historical negative obstacle occupancy status at each location within the clearing range, the historical negative obstacle occupancy status at the corresponding location of the non-negative obstacle is updated to obtain the current negative obstacle occupancy status corresponding to the current location, thereby determining the negative obstacle. The step of updating the historical negative obstacle occupancy status of the corresponding positions of non-negative obstacles based on the updated status of the historical negative obstacle occupancy status of each position within the clearing range, to obtain the current negative obstacle occupancy status corresponding to the current position, and to determine the negative obstacle, includes: For each location within the clearing range, determine whether the historical obstacle occupancy status of that location has been updated at the current location; When the historical obstacle occupancy status of the location has not been updated at the current location, the historical obstacle occupancy status of the location is updated to obtain the current negative obstacle occupancy status corresponding to the current location, so as to determine the negative obstacle, and the location is the location corresponding to the non-negative obstacle.

2. The method according to claim 1, characterized in that, The step of updating the historical occupancy status of the corresponding negative obstacle point based on its location includes: Acquire the external parameters of the acquisition device and the positioning information of the acquisition device on the obstacle avoidance map, wherein the obstacle avoidance map includes the historical negative obstacle occupancy status at each location; Based on the positioning information and the extrinsic parameters, the negative obstacle point is mapped onto the obstacle avoidance map to obtain the corresponding position on the obstacle avoidance map; Update the historical negative obstacle occupancy status of the corresponding location.

3. The method according to claim 2, characterized in that, The step of updating the historical negative obstacle occupancy status at the corresponding location includes: Get the update probability; The historical negative obstacle occupancy status of the corresponding location is updated based on the update probability.

4. The method according to claim 2, characterized in that, The step of obtaining the acquisition range of the acquisition device at the current position to determine the clearing range below the reference plane includes: Obtain the attribute parameters of the acquisition device, including minimum detection distance, maximum detection distance, horizontal detection angle, and vertical detection angle; The actual detection range of the acquisition device at the current location is determined using the attribute parameters. The acquisition range is determined based on the area below the reference plane within the actual detection range; Based on the external parameters and the positioning information, the collection range is mapped to the obstacle avoidance map to determine the clearing range. The obstacle avoidance map includes the historical negative obstacle occupancy status of each location.

5. The method according to claim 4, characterized in that, The step of mapping the collection range to an obstacle avoidance map based on the extrinsic parameters and the positioning information to determine the clearing range includes: Obtain the intersection point between the acquisition range and the reference plane; Based on the extrinsic parameters and the positioning information, each intersection point is mapped to the obstacle avoidance map to determine the intersection point position of each intersection point in the obstacle avoidance map. The clearing range is determined based on the size relationship of each of the intersection points.

6. The method according to claim 1, characterized in that, When the historical obstacle occupancy status of the location has not been updated at the current location, the historical obstacle occupancy status of the location is updated to obtain the current negative obstacle occupancy status corresponding to the current location, so as to determine the negative obstacle, including: When the historical obstacle occupancy status of the location has not been updated at the current location, obtain the clearance probability; The historical obstacle occupancy status of the location is updated based on the clearance probability to obtain the current negative obstacle occupancy status corresponding to the current location, so as to determine the negative obstacle.

7. A method for creating an obstacle avoidance map, characterized in that, include: Obtain the current negative obstacle occupancy state corresponding to the current position in the obstacle avoidance map, wherein the current negative obstacle occupancy state is determined by the negative obstacle detection method according to any one of claims 1-6; A target depth map is obtained, which is determined by performing coordinate system transformation on the initial depth map acquired by the acquisition device using the intrinsic and extrinsic parameters of the acquisition device, and based on the transformation result and the height of the reference surface acquired by the acquisition device. Based on the relationship between the pixel values ​​of each pixel in the current depth map and the corresponding pixel values ​​in the target depth map, the pixel points of the reference plane in the current depth map are determined. Remove the pixels of the reference plane from the current depth map to determine the processed depth map; Obstacles are detected based on the processed depth map to update the obstacle avoidance map and determine the target obstacle avoidance map.

8. An electronic device, characterized in that, include: The memory and the processor are communicatively connected to each other. The memory stores computer instructions, and the processor executes the computer instructions to perform the method for detecting negative obstacles according to any one of claims 1-6, or the method for creating an obstacle avoidance map according to claim 7.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer instructions for causing a computer to perform the method for detecting negative obstacles according to any one of claims 1-6, or the method for creating an obstacle avoidance map according to claim 7.

Images