Map generation method for self-moving device, self-moving device, and storage medium

Abstract

The application belongs to the technical field of automatic control, and particularly relates to a map generation method of a self-moving device, the self-moving device and a medium. The method comprises the following steps: constructing a regional map of a target region during movement of the self-moving device in the target region, the regional map comprising an impassable region where obstacles in the target region are located, and the impassable region being identified by using a preset representation method; obtaining obstacle information obtained during the movement; correcting the regional structure and / or the preset representation method of the impassable region based on the obstacle information to obtain a corrected regional map; the problem that the matching degree of the regional map and the actual environment of the target region is not high can be solved; since the obstacle information is not used when the regional map is constructed, the matching degree of the regional map and the actual environment of the target region can be improved by correcting the regional map by combining the obstacle information in the target region.

Description

Technical Field

[0001] This application belongs to the field of automatic control technology, specifically relating to a map generation method for a self-moving device, a self-moving device, and a storage medium. Background Technology

[0002] Self-moving devices are those that can move automatically within a target area without manual intervention. To avoid collisions with obstacles during operation, self-moving devices acquire a map of the target area before commencing work, including areas where they cannot pass. This allows them to avoid impassable areas while moving according to the map.

[0003] In a typical method for generating a regional map of a target area, the mobile device moves within the target area and gradually builds the regional map during the movement. If a collision occurs during the movement, an impassable area is marked on the already built regional map based on the collision location. After traversing the target area, a complete regional map of the target area is obtained.

[0004] However, regional maps can only show whether each area is passable, and the resulting regional map does not match the actual environment of the target area very well. Summary of the Invention

[0005] This application provides a map generation method for self-moving devices, a self-moving device, and a storage medium, which solves the problem that traditional regional maps only show whether each area is passable, resulting in a low degree of matching between the obtained regional map and the actual environment of the target area. This application provides the following technical solution:

[0006] Firstly, a map generation method for self-moving devices is provided, the method comprising:

[0007] During the movement of the self-moving device within the target area, a regional map of the target area is constructed. The regional map includes impassable areas where obstacles are located within the target area, and the impassable areas are marked using a preset representation method.

[0008] Obtain obstacle information of the obstacle obtained during the movement;

[0009] Based on the obstacle information, the regional structure of the impassable area and / or the preset representation method are corrected to obtain a corrected regional map.

[0010] Optionally, the preset representation method is corrected based on the obstacle information to obtain a corrected area map, including:

[0011] Determine the desired representation of the obstacle information;

[0012] The desired representation method is used to identify the impassable areas corresponding to the obstacles, thus obtaining the corrected area map.

[0013] Optionally, the obstacle information includes obstacle type; correspondingly, the desired representation is used to indicate the obstacle type;

[0014] And / or,

[0015] The obstacle information includes obstacle color; correspondingly, the desired representation includes representation using the obstacle color.

[0016] Optionally, the regional structure of the impassable area is corrected based on the obstacle information, including:

[0017] Based on the obstacle information, determine whether the impassable area is a false area;

[0018] If the impassable area is not a false area, the area map is completed to obtain the corrected area map.

[0019] Optionally, the self-moving device is equipped with a laser sensor and a collision sensor; correspondingly, the obstacle information includes a first obstacle detection result determined based on a first sensing result collected by the laser sensor, and a second obstacle detection result determined based on a second sensing result collected by the collision sensor.

[0020] The step of determining whether the impassable area is a false area based on the obstacle information includes:

[0021] If the first obstacle detection result indicates that there is no obstacle and the second obstacle detection result indicates that there is an obstacle, then the impassable area is determined not to be the false area.

[0022] Optionally, the step of completing the regional map includes:

[0023] The first sensing result is used to complete the area map.

[0024] Optionally, determining whether the impassable area is a false area based on the obstacle information includes:

[0025] The obstacle information is input into a pre-trained neural network model to obtain obstacle recognition results; the obstacle recognition results are used to indicate whether the area where the obstacle is located is a false area.

[0026] Optionally, after correcting the regional structure and / or the preset representation of the impassable area based on the obstacle information to obtain the corrected regional map, the method further includes:

[0027] The edges of the regions in the corrected regional map are straightened.

[0028] Optionally, the process of straightening the regional edges in the corrected regional map includes:

[0029] The location of the corner within the target area is determined based on the obstacle information;

[0030] The edge of the area is straightened based on the corner position.

[0031] In a second aspect, a self-moving device is provided, the self-moving device including a processor and a memory; the memory stores a program, the program being loaded and executed by the processor to implement the map generation method of the self-moving device provided in the first aspect.

[0032] Thirdly, a computer-readable storage medium is provided, wherein a program is stored therein, which, when executed by a processor, is used to implement the map generation method for a self-moving device provided in the first aspect.

[0033] The beneficial effects of this application include at least the following: constructing a regional map of the target area during the movement of the self-moving device within the target area; the regional map includes impassable areas where obstacles are located within the target area, and these impassable areas are marked using a preset representation; acquiring obstacle information obtained during the movement; correcting the regional structure and / or preset representation of the impassable areas based on the obstacle information to obtain a corrected regional map; solving the problem of low matching degree between the regional map and the actual environment of the target area; and improving the matching degree between the regional map and the actual environment of the target area by combining obstacle information within the target area, since obstacle information is not used when constructing the regional map.

[0034] In addition, since the expected representation method corresponding to obstacle information is used to identify the impassable areas corresponding to obstacles, different obstacles can be represented differently in the corrected area map, which can more intuitively distinguish different obstacles. Therefore, the corrected area map can more intuitively reflect the actual environment of the target area and improve the matching degree between the area map and the actual environment of the target area.

[0035] In addition, since obstacle information includes obstacle type, different types of obstacles can be represented differently in the corrected area map. This allows for a more intuitive distinction between different types of obstacles and avoids the problem of not being able to distinguish different types of obstacles from the area map when all obstacles are marked using a preset representation. Therefore, the corrected area map can more intuitively reflect the actual environment of the target area and improve the matching degree between the area map and the actual environment of the target area.

[0036] In addition, since obstacle information includes obstacle color, the corrected area map can represent obstacles of different colors differently, making it easier to distinguish between obstacles of different colors. This avoids the problem of not being able to distinguish between obstacles of different colors on the area map when all obstacles are marked using a preset representation. Therefore, it can improve the matching degree between the area map and the actual environment of the target area.

[0037] Furthermore, since the area map is completed and corrected after determining that the impassable area is not a false area based on obstacle information, the problem of identifying an actual area in the target area as an impassable area can be avoided. Therefore, the matching degree between the area map and the actual environment of the target area can be improved.

[0038] Furthermore, since obstacle information includes the first obstacle detection result determined by the first sensing result collected by the laser sensor and the second obstacle detection result determined by the second sensing result collected by the collision sensor, obstacle information can accurately determine whether an impassable area is a false area, thereby determining whether the area map should be corrected. Therefore, the matching degree between the area map and the actual environment of the target area can be improved.

[0039] Furthermore, since the area map is completed based on the first sensing result after determining that the impassable area is not a false area, and the laser sensing signal emitted by the laser sensor can pass through objects with high light transmittance, the environmental information of the area can be determined based on the first sensing result, thereby completing the area map and improving the matching degree between the area map and the actual environment of the target area.

[0040] Furthermore, since obstacle recognition results based on pre-trained neural network models determine whether the area where the obstacle is located is a false area, and thus determine whether the area map needs to be corrected, the matching degree between the area map and the actual environment of the target area can be improved.

[0041] In addition, straightening the edges of regions in the corrected regional map can avoid errors in the construction and correction process of the regional map, thus improving the matching degree between the regional map and the actual environment of the target region.

[0042] In addition, since the location of the wall can be determined based on the corner position, the edge of the area can be straightened, thus improving the accuracy of the straightening process and the degree of matching between the area map and the actual environment of the target area. Attached Figure Description

[0043] 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.

[0044] Figure 1 This is a schematic diagram of the structure of a self-moving device provided in one embodiment of this application;

[0045] Figure 2 This is a schematic diagram of a map generation method for a self-moving device provided in one embodiment of this application;

[0046] Figure 3 This is a block diagram of a map generation apparatus for a self-moving device provided in one embodiment of this application;

[0047] Figure 4 This is a block diagram of an electronic device provided in one embodiment of this application. Detailed Implementation

[0048] The technical solutions of this application will now be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this application. The application will be described in detail below with reference to the accompanying drawings and embodiments. It should be noted that, unless otherwise specified, the embodiments and features in the embodiments of this application can be combined with each other.

[0049] It should be noted that the terms "first," "second," etc., in the specification, claims, and drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence.

[0050] In this application, unless otherwise stated, directional terms such as "upper," "lower," "top," and "bottom" are generally used in relation to the direction shown in the accompanying drawings, or in relation to the vertical, perpendicular, or gravitational direction of the component itself; similarly, for ease of understanding and description, "inner" and "outer" refer to the inner and outer contours of each component itself, but the above directional terms are not used to limit this utility model.

[0051] First, several terms used in the embodiments of this application will be introduced.

[0052] Laser sensor: It is a sensor that includes a laser generating component and a photosensitive component. When in use, the laser sensor emits a laser towards the target object. The laser produces diffuse reflection on the target object, and the reflected light forms an image on the photosensitive component. The position and shape of the target object can be measured by checking the changes in the position and shape of the image.

[0053] Collision sensors: These generate sensing signals when the device comes into contact with other objects, indicating a collision. Collision sensors are typically mounted on the surface of self-moving equipment.

[0054] Figure 1 This is a schematic diagram of the structure of a self-moving device provided in one embodiment of this application. The self-moving device includes, but is not limited to, devices with automatic movement functions such as sweeping robots, floor scrubbers, and combined sweeping and mopping machines. This embodiment does not limit the type of self-moving device. Figure 1 It is known that the self-moving device includes at least a housing 110, a moving mechanism 120, and a controller 130.

[0055] The housing 110 is the outer shell of the self-moving device. The shape of the housing 110 can be a regular geometric shape, such as a circle or a square, or it can be set to other shapes according to the actual application scenario. This embodiment does not limit the shape of the housing 110.

[0056] The housing 110 primarily serves a protective and supportive function. The housing 110 can be integrally formed or have a detachable structure; this embodiment does not limit the implementation of the housing 110.

[0057] The structure of the housing 110 is generally flat, such as a disc. This embodiment does not limit the shape of the housing 110.

[0058] The moving mechanism 120 is located at the bottom of the housing 110 and is used to drive the self-moving device to move. The moving mechanism 120 can be wheeled or tracked, and this embodiment does not limit the implementation of the moving mechanism 120.

[0059] The moving mechanism 120 is connected to the controller 130 to drive the self-moving device to move under the control of the controller 130.

[0060] The controller 130 can be a microcontroller unit installed inside the self-moving device, or any component with control functions. This embodiment does not limit the type of controller.

[0061] In this embodiment, the controller 130 is used to: construct a regional map of the target area during the movement of the self-moving device within the target area, the regional map including impassable areas where obstacles are located within the target area, the impassable areas being marked using a preset representation method; acquire obstacle information of obstacles obtained during the movement; and correct the regional structure and / or preset representation method of the impassable areas based on the obstacle information to obtain a corrected regional map.

[0062] Optionally, an environmental sensor 140 is provided on the self-moving device to collect obstacle information. The environmental sensor 140 is used to collect obstacle information of obstacles in the target area as the self-moving device moves in the target area.

[0063] The environmental sensor 140 can be a vision sensor, a laser sensor, or a collision sensor. The vision sensor includes, but is not limited to, a charge-coupled device (CCD) or a metal-oxide-semiconductor (CMOS) device. This embodiment does not limit the type of the environmental sensor 140.

[0064] In one example, the environmental sensor 140 includes a vision sensor. Accordingly, the obstacle information includes obstacle color.

[0065] Optionally, there may be one environmental sensor 140, or at least two. This embodiment does not limit the number of environmental sensors 140.

[0066] Optionally, the environmental sensor 140 may be located on the top and / or side of the housing 110. The acquisition range of the environmental sensor 140 includes, but is not limited to: in front of the mobile device in the direction of travel, to the left of the mobile device in the direction of travel, and / or to the left of the mobile device in the direction of travel. This embodiment does not limit the specific installation position and acquisition range of the environmental sensor 140.

[0067] In actual implementation, the self-moving device may also include other components, such as a battery pack and a side brush. This embodiment will not list all the components included in the self-moving device.

[0068] In this embodiment, a regional map of the target area is constructed as the self-moving device moves within the target area. The regional map includes impassable areas where obstacles are located within the target area, and these impassable areas are marked using a preset representation method. Obstacle information obtained during the movement is acquired. Based on the obstacle information, the regional structure and / or preset representation method of the impassable areas are corrected to obtain a corrected regional map. This can solve the problem of low matching degree between the regional map and the actual environment of the target area. Since obstacle information is not used when constructing the regional map, the matching degree between the regional map and the actual environment of the target area can be improved by combining obstacle information within the target area to correct the regional map.

[0069] The following section provides a detailed description of the map generation method for self-moving devices provided in this application.

[0070] This embodiment provides a map generation method for self-mobile devices, such as... Figure 2 As shown. This embodiment uses this method for... Figure 1 The self-moving device shown is used as an example for illustration. In other embodiments, it can also be performed by other devices that are communicatively connected to the self-moving device, such as remotely controlling the cleaning equipment through a mobile phone, computer, tablet, or other device. This embodiment does not limit the implementation method of other devices or the execution subject of each embodiment. The map generation method includes at least the following steps:

[0071] Step 201: Construct a regional map of the target area as the self-moving device moves within the target area.

[0072] The area map includes impassable areas where obstacles are located within the target area, and these impassable areas are marked using a preset representation method.

[0073] Optionally, the target area can be a residence, an office, or a factory; this embodiment does not limit the type of the target area. This embodiment uses a residence as an example for illustration.

[0074] Optionally, the obstacle can be a wall within the target area, or it can be furniture within the target area, such as a table, coffee table, wardrobe, bed, etc. This embodiment does not limit the type of obstacle.

[0075] The preset representation method refers to the traditional method of marking impassable areas during the construction of regional maps. Optionally, the preset representation method can be text, such as marking impassable areas with the text "Impassable Area"; or it can be numbers, such as marking impassable areas with the number "1"; or it can be a fixed color, such as using gray to mark impassable areas. This embodiment does not limit the type of preset representation method.

[0076] Optionally, the method for constructing a regional map of the target area includes, but is not limited to, one of the following methods:

[0077] The first method involves constructing a regional map of the target area based on the movement trajectory of the mobile device within the target area. This construction includes: determining traversable areas within the target area based on the movement trajectory of the mobile device; and determining impassable areas within the target area based on the boundary information and traversable areas, thus constructing the regional map of the target area.

[0078] The second approach involves constructing a regional map of the target area based on the collision locations of the self-moving device during its movement within the target area. In this case, constructing the target area map includes: obtaining the collision locations of the self-moving device during its movement within the target area; and identifying the collision locations as obstacle locations to construct the regional map of the target area.

[0079] Optionally, the collision of the self-moving device can be determined based on the sensing information of the collision sensor installed on the side of the self-moving device's housing, or it can be determined based on the sensing information of the acceleration sensor installed on the self-moving device. This embodiment does not limit the collision detection method of the self-moving device.

[0080] In one example, a collision is determined when the self-moving device detects an acceleration in a direction other than travel that exceeds a preset acceleration threshold.

[0081] In actual implementation, self-mobile devices can also construct regional maps of the target area in other ways. This embodiment does not limit the way self-mobile devices construct regional maps of the target area.

[0082] Step 202: Obtain obstacle information of obstacles obtained during the movement.

[0083] Optionally, obstacle information includes, but is not limited to, the following: obstacle type, obstacle color, and / or obstacle edge position. This embodiment does not limit the type of obstacle information.

[0084] In one example, obstacles are categorized into furniture and other obstacles, with furniture further divided into types such as tables, chairs, and cabinets.

[0085] In another example, obstacles are categorized into movable obstacles and immovable obstacles. Immovable obstacles include types such as walls and wardrobes; movable obstacles include types such as tables and chairs.

[0086] In actual implementation, obstacles can also be classified according to other classification methods. This embodiment does not limit the classification method of obstacles.

[0087] Optionally, obstacle information may be generated by the self-moving device based on environmental information collected by environmental sensors, or it may be sent to the self-moving device by other devices. This embodiment does not limit the method of obtaining obstacle information.

[0088] In one example, obstacle information is generated by the mobile device based on environmental information collected by environmental sensors. In this case, acquiring obstacle information obtained during the movement includes at least one of the following:

[0089] In the first scenario, obstacle information includes obstacle type. In this case, the environmental sensors include visual sensors or laser sensors, and the environmental information is image information of the area where the obstacle is located. Therefore, acquiring obstacle information obtained during movement includes determining the obstacle type based on the image information of the area where the obstacle is located.

[0090] Optionally, determining the obstacle type based on image information of the area where the obstacle is located includes: inputting the image of the area where the obstacle is located into a pre-trained obstacle recognition model to obtain the obstacle type.

[0091] The obstacle recognition model is obtained by training the neural network using the first training data. Each set of the first training data includes a first sample image and obstacle type label data in the first sample image.

[0092] For example, the training process of an obstacle recognition model includes: creating an initial network model; inputting sample images and obstacle type label data from the sample images into the initial network model to obtain the model results; iteratively updating the parameters of the initial network model based on the model results and the corresponding obstacle type label data; obtaining the obstacle recognition model when the number of iterations reaches a preset number, or when the updated model converges.

[0093] The initial network model can be a BP neural network (Back Propagation Neural Network), an ART neural network (Adaptive Resonance Theory), or a radial basis function (RBF) neural network. This embodiment does not limit the type of the initial network model.

[0094] In the second scenario, obstacle information includes obstacle color. In this case, the environmental sensor includes a vision sensor, and the environmental information is the image information of the area where the obstacle is located. Therefore, acquiring obstacle information obtained during movement includes determining the obstacle's color based on the image information of the area where the obstacle is located.

[0095] The image information acquired by the vision sensor is color image information, which is information composed of data from the red, green, and blue channels in the same row and column.

[0096] In the third scenario, obstacle information includes the location of obstacle edges. In this case, the environmental sensors include laser sensors or vision sensors, and the environmental information is the laser sensing signal of the area where the obstacle is located. Therefore, acquiring obstacle information obtained during movement includes determining the obstacle edge location based on the laser sensing signal of the area where the obstacle is located.

[0097] Among them, the laser sensing signal is used to indicate the height change of the area where the obstacle is located, that is, the laser sensing signal is different for different heights.

[0098] Optionally, determining the edge position of an obstacle based on the laser sensing signal of the area where the obstacle is located includes: determining the location where the height changes in the area where the obstacle is located based on the laser sensing signal; and determining the location where the height changes in the area where the obstacle is located as the edge position of the obstacle.

[0099] In another example, obstacle information is sent from other devices to the self-moving device. In this case, obtaining obstacle information obtained during the movement includes: the self-moving device sending its current location information to other devices during the movement so that other devices can determine the obstacle information of the sub-region where the current location is located based on the current location information and send it to the self-moving device; and receiving obstacle information sent by other devices.

[0100] Step 203: Correct the regional structure and / or preset representation of the impassable area based on the obstacle information to obtain the corrected regional map.

[0101] Optionally, the modified area map is obtained by correcting the preset representation method based on the obstacle information, including: determining the desired representation method corresponding to the obstacle information; using the desired representation method to identify the impassable areas at the corresponding locations of the obstacles, thereby obtaining the modified area map.

[0102] Optionally, different types of obstacle information correspond to different expected representations. Therefore, determining the expected representation of obstacle information includes at least one of the following:

[0103] In the first case, the obstacle information includes the obstacle type. Accordingly, the desired representation is used to indicate the obstacle type. In this case, determining the desired representation corresponding to the obstacle information includes: determining the type data corresponding to the obstacle type as the desired representation.

[0104] Optionally, the type data can be text, such as the obstacle type name, or it can be numbers, such as the number corresponding to the obstacle type, or it can be colors, such as the color corresponding to the obstacle type. This embodiment does not limit the implementation method of the type data.

[0105] In the second scenario, the obstacle information includes obstacle color. Correspondingly, the desired representation includes using obstacle color. In this case, determining the desired representation corresponding to the obstacle information includes: determining the obstacle color as the desired representation.

[0106] For example, if the obstacle is red, then red is determined as the desired representation.

[0107] Optionally, the regional structure of impassable areas is corrected based on obstacle information, including: determining whether the impassable area is a false area based on obstacle information; and, if the impassable area is not a false area, completing the regional map to obtain the corrected regional map.

[0108] False regions refer to areas identified by detecting fake images formed on highly reflective objects. Highly reflective objects are defined as those with a light reflectivity exceeding a preset threshold, through which most light cannot pass. Highly reflective objects include, but are not limited to, mirrors and electroplated products. Because false regions are not real and cannot be entered by mobile devices, they are marked as impassable areas during map building.

[0109] Optionally, the reflectivity threshold is greater than or equal to the light reflectivity corresponding to the minimum amount of light signal required for the self-moving device to identify the false area; the amount of light signal reflected by the object is positively correlated with the object's light reflectivity.

[0110] In traditional area map construction, mobile devices cannot pass through highly transparent objects, so they often designate the area containing such objects as impassable. However, since a highly transparent object may lie between two areas, one of which is the area where the mobile device is located, designating the other area connected to the highly transparent object as impassable would lead to a discrepancy between the area map and the actual environment of the target area.

[0111] High light transmittance objects, also known as transparent objects, refer to objects whose light transmittance is greater than a preset light transmittance threshold, allowing most light to pass through them. High light transmittance objects include, but are not limited to, plexiglass and transparent fiberglass.

[0112] Optionally, the transmittance threshold is greater than or equal to the transmittance corresponding to the minimum amount of light signal required for the self-moving device to identify the object; the amount of light signal transmitted through the object is positively correlated with the transmittance of the object.

[0113] Based on the aforementioned technical issues, in this embodiment, the self-moving device needs to determine whether the impassable area is a false area.

[0114] Optionally, the methods for determining whether an impassable area is a false area based on obstacle information include, but are not limited to, at least one of the following:

[0115] In the first method, a laser sensor and a collision sensor are installed on the mobile device. Accordingly, the obstacle information includes a first obstacle detection result determined based on the first sensing result collected by the laser sensor, and a second obstacle detection result determined based on the second sensing result collected by the collision sensor. In this case, determining whether an impassable area is a false area based on the obstacle information includes: if the first obstacle detection result indicates that there is no obstacle and the second obstacle detection result indicates that there is an obstacle, then determining that the impassable area is not a false area.

[0116] Because the sensing signal of the laser sensor can pass through objects with high light transmittance, the first obstacle recognition result cannot identify such objects. However, since the high-transmittance objects actually exist, the second obstacle recognition result can identify them, and the area containing the high-transmittance objects is determined to be an impassable area. Therefore, by combining the first and second obstacle recognition results, it can be determined whether the impassable area is a false area.

[0117] Since the first sensing result can reflect the environmental information within the area when the impassable area is not a false area, the area map is completed, including: using the first sensing result to complete the area map.

[0118] The second approach determines whether an impassable area is a false area based on obstacle information. This includes: inputting obstacle information into a pre-trained neural network model to obtain obstacle recognition results; and using the obstacle recognition results to indicate whether the area where the obstacle is located is a false area.

[0119] The neural network model is obtained by training the neural network using the second training data. Each set of second training data includes the obstacle type label data in the second sample image and the second sample image.

[0120] The training method for the neural network model is the same as that for the obstacle recognition model described above, and will not be repeated here.

[0121] Optionally, the obstacle information can be an obstacle image, which can be collected by a visual sensor. Alternatively, it can be the outline information of the obstacle, which can be collected by a laser sensor. This embodiment does not limit the type of obstacle information or the method of collecting obstacle information.

[0122] Optionally, the obstacle identification result is an obstacle type, which includes: high-transmittance obstacles and other obstacles, where other obstacles refer to obstacles other than high-transmittance obstacles. If the obstacle type indicates that the obstacle is a high-transmittance obstacle, the impassable area is determined not to be a false area; if the obstacle type indicates that the obstacle is another obstacle, the impassable area is determined to be a false area.

[0123] Since obstacle information can reflect environmental information within an area when it is not a false area, the area map is completed by using obstacle information.

[0124] Optionally, after correcting the regional structure and / or preset representation of the impassable area based on obstacle information to obtain the corrected regional map, the method further includes: performing straightening processing on the regional edges in the corrected regional map.

[0125] Among them, straightening treatment refers to straightening the edges of the curved area.

[0126] A corner refers to the angle between adjacent walls. Since it's difficult for a mobile device to reach the corner of the target area during its movement, the edge information at corner locations in the corrected area map may be inaccurate. This can cause the edges of the area at corner locations to appear curved. Therefore, edge straightening processing is performed on the corrected area map, including: determining the corner locations within the target area based on obstacle information; and straightening the area edges based on the corner locations.

[0127] Optionally, determining the location of the center corner of the target area based on obstacle information includes, but is not limited to, at least one of the following:

[0128] The first scenario: Obstacle information includes the location of obstacle edges. In this case, determining the corner location in the target area based on the obstacle information includes: determining the obstacle edge contour information based on the obstacle edge location; and determining the corner location in the target area based on the obstacle edge contour information.

[0129] Optionally, determining obstacle edge contour information based on obstacle edge location includes connecting adjacent obstacle edge locations to obtain obstacle contour information.

[0130] Optionally, determining the corner position in the target area based on obstacle edge contour information includes: determining the corner position where the turning angle in the obstacle edge contour indicated by the obstacle edge contour information is greater than a preset angle threshold as the corner position.

[0131] The preset angle threshold is pre-stored in the self-moving device.

[0132] In one example, the preset angle threshold is 80 degrees.

[0133] Optionally, the process of determining the corner angle in the obstacle edge contour includes: determining the corner position in the obstacle edge contour; taking the corner position as the starting point and cutting off the edge contour line segments of the region of a preset length in different directions; calculating the line segment angle between different edge contour line segments respectively; and determining the largest line segment angle as the corner angle at the corner position.

[0134] The second scenario: Obstacle information includes obstacle images. In this case, determining the corner position in the target area based on the obstacle image includes: determining the corner position in the target area based on the obstacle image.

[0135] Optionally, determining the corner position in the target area based on the obstacle image includes: inputting the obstacle image into a pre-trained corner recognition model to obtain a corner recognition result; the corner recognition result is used to indicate whether a corner exists in the obstacle image and the position of the corner.

[0136] The corner recognition model is obtained by training the neural network using third training data. Each set of third training data includes a third sample image and corner location label data in the third sample image.

[0137] The training process for the corner recognition model is the same as that for the obstacle recognition model described above, and will not be repeated here.

[0138] Optionally, the methods for straightening the edge of the area based on the corner location include, but are not limited to, at least one of the following methods:

[0139] The first method involves segmenting the edge of the target area based on the corner position; and then performing straightening processing on each segment of the edge.

[0140] Optionally, straightening processing is performed on the edge of each region segment, including: fitting a straight line to the edge of each region segment separately.

[0141] Optionally, the method for line fitting can be the least squares method, or it can be the gradient descent method. This embodiment does not limit the method for line fitting.

[0142] The second method: Determine the wall position based on the corner position; straighten the edge of the area corresponding to the wall position.

[0143] Optionally, the edge of the area corresponding to the wall position is straightened, including: fitting the edge line segment of the area corresponding to the wall position with a straight line.

[0144] The third method: straighten the edges of the corner area.

[0145] Optionally, the edge of the area at the corner is straightened, including: taking the corner as the starting point and cutting off the edge line segments of the area to a preset length in different directions; and fitting the cut-off edge line segments with straight lines.

[0146] To eliminate the influence of noise and abrupt changes in the region edges on the straightening process, the region edges in the corrected region map are straightened. This includes: filtering the region edges in the corrected region map to remove abrupt changes at the region edge locations; and straightening the filtered region edge locations.

[0147] Optionally, the edges of the corrected area map are filtered, including removing locations where the difference between the edge location and its adjacent location is greater than a preset difference.

[0148] In other embodiments, the self-moving device may also perform straightening processing on the area edges in the corrected area map based on the feature information in the corrected area map, such as: identifying the wall positions in the corrected area map; and performing straightening processing on the area edges based on the wall positions. This embodiment does not limit the method of straightening processing on the area edges in the corrected area map.

[0149] In summary, the map generation method for self-moving devices provided in this embodiment constructs a regional map of the target area during the movement of the self-moving device within the target area. The regional map includes impassable areas where obstacles are located within the target area, and these impassable areas are marked using a preset representation. Obstacle information obtained during the movement is acquired. Based on the obstacle information, the regional structure and / or preset representation of the impassable areas are corrected to obtain a corrected regional map. This method can solve the problem of low matching degree between the regional map and the actual environment of the target area. Since obstacle information is not used when constructing the regional map, the matching degree between the regional map and the actual environment of the target area can be improved by combining obstacle information within the target area to correct the regional map.

[0150] In addition, since the expected representation method corresponding to obstacle information is used to identify the impassable areas corresponding to obstacles, different obstacles can be represented differently in the corrected area map, which can more intuitively distinguish different obstacles. Therefore, the corrected area map can more intuitively reflect the actual environment of the target area and improve the matching degree between the area map and the actual environment of the target area.

[0151] In addition, since obstacle information includes obstacle type, different types of obstacles can be represented differently in the corrected area map. This allows for a more intuitive distinction between different types of obstacles and avoids the problem of not being able to distinguish different types of obstacles from the area map when all obstacles are marked using a preset representation. Therefore, the corrected area map can more intuitively reflect the actual environment of the target area and improve the matching degree between the area map and the actual environment of the target area.

[0152] In addition, since obstacle information includes obstacle color, the corrected area map can represent obstacles of different colors differently, making it easier to distinguish between obstacles of different colors. This avoids the problem of not being able to distinguish between obstacles of different colors on the area map when all obstacles are marked using a preset representation. Therefore, it can improve the matching degree between the area map and the actual environment of the target area.

[0153] Furthermore, since the area map is completed and corrected after determining that the impassable area is not a false area based on obstacle information, the problem of identifying an actual area in the target area as an impassable area can be avoided. Therefore, the matching degree between the area map and the actual environment of the target area can be improved.

[0154] Furthermore, since obstacle information includes the first obstacle detection result determined by the first sensing result collected by the laser sensor and the second obstacle detection result determined by the second sensing result collected by the collision sensor, obstacle information can accurately determine whether an impassable area is a false area, thereby determining whether the area map should be corrected. Therefore, the matching degree between the area map and the actual environment of the target area can be improved.

[0155] Furthermore, since the area map is completed based on the first sensing result after determining that the impassable area is not a false area, and the laser sensing signal emitted by the laser sensor can pass through objects with high light transmittance, the environmental information of the area can be determined based on the first sensing result, thereby completing the area map and improving the matching degree between the area map and the actual environment of the target area.

[0156] Furthermore, since obstacle recognition results based on pre-trained neural network models determine whether the area where the obstacle is located is a false area, and thus determine whether the area map needs to be corrected, the matching degree between the area map and the actual environment of the target area can be improved.

[0157] In addition, straightening the edges of regions in the corrected regional map can avoid errors in the construction and correction process of the regional map, thus improving the matching degree between the regional map and the actual environment of the target region.

[0158] In addition, since the location of the wall can be determined based on the corner position, the edge of the area can be straightened, thus improving the accuracy of the straightening process and the degree of matching between the area map and the actual environment of the target area.

[0159] This embodiment provides a map generation device for self-mobile devices, such as... Figure 3 As shown. This embodiment applies the device to... Figure 1 The controller of the self-moving device shown includes at least the following modules: map building module 310, information acquisition module 320, and map correction module 330.

[0160] The map building module 310 is used to build a regional map of the target area during the movement of the self-moving device in the target area. The regional map includes impassable areas where obstacles are located in the target area. The impassable areas are marked using a preset representation method.

[0161] The information acquisition module 320 is used to acquire obstacle information of obstacles obtained during the movement;

[0162] The map correction module 330 is used to correct the regional structure and / or preset representation of impassable areas based on obstacle information, thereby obtaining a corrected regional map. For related details, please refer to the above-described method and device embodiments.

[0163] It should be noted that the map generation device for self-mobile devices provided in the above embodiments is only illustrated by the division of the above functional modules. In practical applications, the above functions can be assigned to different functional modules as needed, that is, the internal structure of the map generation device for self-mobile devices can be divided into different functional modules to complete all or part of the functions described above. In addition, the map generation device for self-mobile devices provided in the above embodiments and the map generation method embodiments for self-mobile devices belong to the same concept, and the specific implementation process is detailed in the method embodiments, which will not be repeated here.

[0164] This embodiment provides an electronic device, such as... Figure 4 As shown. Electronic devices can be Figure 1 A self-moving device. The electronic device includes at least a processor 401 and a memory 402.

[0165] Processor 401 may include one or more processing cores, such as a quad-core processor or an octa-core processor. Processor 401 may be implemented using at least one hardware form selected from DSP (Digital Signal Processing), FPGA (Field-Programmable Gate Array), and PLA (Programmable Logic Array). Processor 401 may also include a main processor and a coprocessor. The main processor, also known as a CPU (Central Processing Unit), is used to process data in the wake-up state; the coprocessor is a low-power processor used to process data in the standby state. In some embodiments, processor 401 may integrate a GPU (Graphics Processing Unit), which is responsible for rendering and drawing the content to be displayed on the screen. In some embodiments, processor 401 may also include an AI (Artificial Intelligence) processor, which is used to handle computational operations related to machine learning.

[0166] The memory 402 may include one or more computer-readable storage media, which may be non-transitory. The memory 402 may also include high-speed random access memory and non-volatile memory, such as one or more disk storage devices or flash memory devices. In some embodiments, the non-transitory computer-readable storage media in the memory 402 are used to store at least one instruction, which is executed by the processor 401 to implement the map generation method for a self-moving device provided in the method embodiments of this application.

[0167] In some embodiments, the electronic device may also optionally include: a peripheral device interface and at least one peripheral device. The processor 401, memory 402, and peripheral device interface can be connected via a bus or signal line. Each peripheral device can be connected to the peripheral device interface via a bus, signal line, or circuit board. Indicatively, peripheral devices include, but are not limited to: radio frequency circuitry, a touch display screen, audio circuitry, and a power supply.

[0168] Of course, electronic devices may also include fewer or more components, and this embodiment does not limit this.

[0169] Optionally, this application also provides a self-moving device, which includes a processor and a memory; the memory stores a program, which is loaded and executed by the processor to implement the map generation method of the self-moving device described in the above method embodiments.

[0170] Optionally, this application also provides a computer-readable storage medium storing a program that is loaded and executed by a processor to implement the map generation method of the self-moving device described in the above method embodiments.

[0171] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0172] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the invention patent. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this patent application should be determined by the appended claims.

Claims

1. A map generation method for self-moving devices, characterized in that, The method includes: During the movement of the self-moving device within the target area, a regional map of the target area is constructed. The regional map includes impassable areas where obstacles are located within the target area, and the impassable areas are marked using a preset representation method. Obtain obstacle information of the obstacle obtained during the movement; Based on the obstacle information, the regional structure of the impassable area is corrected to obtain a corrected regional map; The process of correcting the regional structure of the impassable area based on the obstacle information includes: Based on the obstacle information, determine whether the impassable area is a false area; If the impassable area is not a false area, the area map is completed to obtain the corrected area map; The false area refers to a false area obtained by identifying false images formed on a high reflectivity object. The high reflectivity object is defined as an object with a light reflectivity greater than a preset reflectivity threshold. The self-moving device marks the false area as an impassable area during the map construction process.

2. The method according to claim 1, characterized in that, Also includes: Based on the obstacle information, the regional structure of the impassable area and the preset representation method are corrected to obtain the corrected regional map.

3. The method according to claim 2, characterized in that, Based on the obstacle information, the regional structure of the impassable area and the preset representation method are corrected to obtain a corrected regional map, including: Determine the desired representation of the obstacle information; The desired representation method is used to identify the impassable areas corresponding to the obstacles, thus obtaining the corrected area map.

4. The method according to claim 3, characterized in that, The obstacle information includes obstacle types; correspondingly, the expected representation is used to indicate the obstacle type; And / or, The obstacle information includes obstacle color; correspondingly, the desired representation includes representation using the obstacle color.

5. The method according to claim 1, characterized in that, The self-moving device is equipped with a laser sensor and a collision sensor; accordingly, the obstacle information includes a first obstacle detection result determined based on a first sensing result collected by the laser sensor, and a second obstacle detection result determined based on a second sensing result collected by the collision sensor. The step of determining whether the impassable area is a false area based on the obstacle information includes: If the first obstacle detection result indicates that there is no obstacle and the second obstacle detection result indicates that there is an obstacle, then the impassable area is determined not to be the false area.

6. The method according to claim 5, characterized in that, The process of completing the regional map includes: The first sensing result is used to complete the area map.

7. The method according to claim 1, characterized in that, The step of determining whether the impassable area is a false area based on the obstacle information includes: The obstacle information is input into a pre-trained neural network model to obtain obstacle recognition results; the obstacle recognition results are used to indicate whether the area where the obstacle is located is a false area.

8. The method according to claim 1, characterized in that, After correcting the regional structure of the impassable area based on the obstacle information to obtain the corrected regional map, the method further includes: The edges of the regions in the corrected regional map are straightened.

9. The method according to claim 8, characterized in that, The process of straightening the edges of regions in the corrected region map includes: The location of the corner within the target area is determined based on the obstacle information; The edge of the area is straightened based on the corner position.

10. A self-moving device, characterized in that, The self-moving device includes a processor and a memory; the memory stores a program that is loaded and executed by the processor to implement the map generation method of the self-moving device as described in any one of claims 1 to 9.

11. A computer-readable storage medium, characterized in that, The storage medium stores a program that, when executed by a processor, is used to implement the map generation method for a self-moving device as described in any one of claims 1 to 9.

Images