Path planning method, target object motion control method, device and electronic equipment

Abstract

The present disclosure provides a path planning method, a target object motion control method, a device and an electronic device, and relates to the technical field of intelligent driving and the technical field of artificial intelligence. The path planning method comprises: performing obstacle detection on an initial image representing a target object moving space to obtain an intermediate image, wherein the intermediate image comprises an obstacle image area representing obstacles in the target object moving space, and a starting position and a stopping position of the target object; performing image feature extraction on the intermediate image to obtain intermediate image features; and determining a motion path of the target object from the starting position to the stopping position according to the intermediate image features.

Description

Technical Field

[0001] This disclosure relates to the fields of intelligent driving technology and artificial intelligence technology, and more specifically, to a path planning method, a target object motion control method, a device, an electronic device, a storage medium, and a computer program product. Background Technology

[0002] With the rapid development of technology, autonomous driving functions are widely used in passenger cars, logistics vehicles, and automated inspection vehicles, enabling vehicles to perform autonomous driving functions such as automatic parking and automatic inspection based on the generated movement path, thereby improving the driving efficiency and operational efficiency of vehicles.

[0003] In realizing the concept disclosed herein, the inventors discovered that the related technologies have at least the following problems: the motion paths generated in the related technologies have low accuracy in avoiding obstacles in the moving space, and there are situations where they collide with obstacles during the movement, resulting in low motion efficiency of target objects such as vehicles, which makes it difficult to meet actual needs. Summary of the Invention

[0004] In view of this, the present disclosure provides a path planning method, a target object motion control method, an apparatus, an electronic device, a storage medium, and a computer program product.

[0005] One aspect of this disclosure provides a path planning method, comprising:

[0006] Obstacle detection is performed on the initial image representing the movement space of the target object to obtain an intermediate image, wherein the intermediate image includes an obstacle image region representing obstacles in the movement space of the target object, as well as the starting position and the stopping position of the target object;

[0007] Image feature extraction is performed on the aforementioned intermediate image to obtain intermediate image features; and

[0008] Based on the aforementioned intermediate image features, the motion path of the target object from the aforementioned starting position to the aforementioned stopping position is determined.

[0009] According to embodiments of this disclosure, the initial image includes an initial obstacle image region that characterizes the shape of the obstacle.

[0010] Obstacle detection on the initial image representing the movement space of the target object includes:

[0011] The initial image is then processed into a grid to obtain an initial grid image, wherein the initial grid image comprises multiple initial grid image units;

[0012] Overlap detection is performed on the initial grid image units and the initial obstacle image regions to obtain overlap detection results; and

[0013] Based on the overlap detection results of each of the initial grid image units, the initial pixel values ​​of each of the initial grid image units are updated to obtain the intermediate image.

[0014] According to embodiments of this disclosure, updating the initial pixel values ​​of each of the initial mesh image units based on the overlap detection results of each of the initial mesh image units includes:

[0015] If the overlap detection result corresponding to the initial mesh image unit is the first detection result, the initial pixel value of the initial mesh image unit is updated to the first pixel value, wherein the first detection result indicates that the initial mesh image unit overlaps with the initial obstacle image region.

[0016] If the overlap detection result corresponding to the initial mesh image unit is the second detection result, the initial pixel value of the initial mesh image unit is updated to the second pixel value, wherein the second detection result indicates that the initial mesh image unit is spaced apart from the initial obstacle image region.

[0017] According to embodiments of this disclosure, overlap detection is performed on the initial mesh image unit and the initial obstacle image region to obtain the overlap detection result, including:

[0018] When the initial grid image unit overlaps at least partially with the initial obstacle image region, the overlap detection result is determined as the first detection result.

[0019] According to embodiments of this disclosure, overlap detection is performed on the initial mesh image unit and the initial obstacle image region to obtain the overlap detection result, including:

[0020] If the overlap between the initial grid image unit and the initial obstacle image region is greater than a preset overlap threshold, the overlap detection result is determined as the first detection result.

[0021] According to embodiments of this disclosure, obstacle detection on an initial image characterizing the movement space of a target object includes:

[0022] Based on the object detection algorithm, the initial image is processed to obtain object detection boxes representing the obstacles; and

[0023] Based on the target detection bounding box described above, the intermediate image is determined.

[0024] According to embodiments of this disclosure, determining the motion path of the target object from the starting position to the stopping position based on the aforementioned intermediate image features includes:

[0025] The aforementioned intermediate image features are input into a path prediction model constructed based on a neural network algorithm, which outputs the path positions between the starting position and the stopping position; and

[0026] Based on the locations mentioned above, the aforementioned movement path is determined.

[0027] According to embodiments of this disclosure, the path prediction model includes at least one of the following:

[0028] Recurrent neural network model, long short-term memory network model, bidirectional long short-term memory network model, attention network model.

[0029] According to embodiments of this disclosure, the target object includes at least one of the following:

[0030] Unmanned vehicles and intelligent robots.

[0031] Another aspect of this disclosure provides a method for motion control of a target object, comprising:

[0032] The motion path is determined according to the method described above; and

[0033] Based on the motion path, control the target object to perform motion operations.

[0034] Another aspect of this disclosure provides a path planning apparatus, comprising:

[0035] The detection module is used to perform obstacle detection on the initial image representing the movement space of the target object to obtain an intermediate image, wherein the intermediate image includes an obstacle image region representing obstacles in the movement space of the target object, as well as the starting position and the stopping position of the target object.

[0036] The image feature extraction module is used to extract image features from the intermediate image to obtain intermediate image features; and

[0037] The first motion path determination module is used to determine the motion path of the target object from the starting position to the stopping position based on the intermediate image features.

[0038] Another aspect of this disclosure provides a target object motion control device, comprising:

[0039] The second motion path determination module is used to determine the motion path according to the method described above; and

[0040] The motion operation control module is used to control the target object to perform motion operations according to the motion path.

[0041] Another aspect of this disclosure provides an electronic device comprising: one or more processors; and a memory for storing one or more programs, wherein, when the one or more programs are executed by the one or more processors, the one or more processors perform the methods described above.

[0042] Another aspect of this disclosure provides a computer-readable storage medium storing computer-executable instructions that, when executed, are used to implement the method described above.

[0043] Another aspect of this disclosure provides a computer program product including computer-executable instructions that, when executed, are used to implement the method described above.

[0044] According to the embodiments of this disclosure, by performing obstacle detection on the initial image and extracting image features from the intermediate image obtained after detection, the technical problem of low obstacle detection accuracy in related technologies is at least partially overcome. By determining the motion path between the starting position and the stopping position based on the obtained intermediate image features, the obtained motion path can accurately avoid obstacles. Thus, the target object can be controlled to perform motion operations based on the motion path, enabling the target object to accurately avoid obstacles in the target object's movement space during motion. This at least partially overcomes the technical problem of low obstacle avoidance accuracy during the target object's movement, achieving the technical effect of improving the target object's motion efficiency and obstacle avoidance success rate. Attached Figure Description

[0045] The above and other objects, features and advantages of this disclosure will become clearer from the following description of embodiments with reference to the accompanying drawings, in which:

[0046] Figure 1 This illustration schematically shows an exemplary system architecture to which path planning methods and apparatus can be applied according to embodiments of the present disclosure;

[0047] Figure 2 A flowchart illustrating a path planning method according to an embodiment of the present disclosure is shown schematically.

[0048] Figure 3 The flowchart illustrating obstacle detection of an initial image characterizing the movement space of a target object according to an embodiment of the present disclosure is shown schematically.

[0049] Figure 4 The flowchart illustrating obstacle detection of an initial image characterizing the movement space of a target object is shown schematically according to another embodiment of the present disclosure;

[0050] Figure 5 This diagram illustrates an application scenario of the path planning method according to an embodiment of the present disclosure.

[0051] Figure 6a A flowchart illustrating a target object motion control method according to an embodiment of the present disclosure is shown schematically.

[0052] Figure 6b This diagram schematically illustrates an application scenario of the target object motion control method according to an embodiment of the present disclosure.

[0053] Figure 7 A block diagram of a path planning apparatus according to an embodiment of the present disclosure is shown schematically;

[0054] Figure 8 A block diagram schematically illustrates a target object motion control device according to an embodiment of the present disclosure; and

[0055] Figure 9 A block diagram of an electronic device suitable for implementing a path planning method and a target object motion control method according to embodiments of the present disclosure is shown schematically. Detailed Implementation

[0056] The embodiments of the present disclosure will now be described with reference to the accompanying drawings. However, it should be understood that these descriptions are exemplary only and are not intended to limit the scope of the disclosure. In the following detailed description, numerous specific details are set forth to provide a thorough understanding of the embodiments of the present disclosure for ease of explanation. However, it will be apparent that one or more embodiments may be practiced without these specific details. Furthermore, descriptions of well-known structures and techniques are omitted in the following description to avoid unnecessarily obscuring the concepts of the present disclosure.

[0057] The terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit this disclosure. The terms “comprising,” “including,” etc., as used herein indicate the presence of the stated features, steps, operations, and / or components, but do not exclude the presence or addition of one or more other features, steps, operations, or components.

[0058] All terms used herein (including technical and scientific terms) have the meanings commonly understood by those skilled in the art, unless otherwise defined. It should be noted that the terms used herein are to be interpreted in a manner consistent with the context of this specification, and not in an idealized or overly rigid way.

[0059] When using expressions such as "at least one of A, B, and C", they should generally be interpreted in accordance with the meaning that is commonly understood by a person skilled in the art (e.g., "a system having at least one of A, B, and C" should include, but is not limited to, a system having A alone, a system having B alone, a system having C alone, a system having A and B, a system having A and C, a system having B and C, and / or a system having A, B, and C, etc.).

[0060] In the embodiments disclosed herein, the collection, updating, analysis, processing, use, transmission, provision, disclosure, and storage of data (e.g., including but not limited to user personal information) comply with relevant laws and regulations, are used for legitimate purposes, and do not violate public order and good morals. In particular, necessary measures have been taken to prevent unauthorized access to user personal information data and to safeguard user personal information security, network security, and national security.

[0061] In the embodiments disclosed herein, user authorization or consent is obtained before acquiring or collecting user personal information.

[0062] In realizing the inventive concept of this disclosure, the inventors discovered that related technologies typically rely on fully discrete spatial point search methods, artificial potential field methods, or sampling methods to plan the motion paths of target objects such as unmanned vehicles. However, the computational overhead and time consumption of generating motion paths based on fully discrete spatial point search methods or sampling methods are too high, placing excessive demands on the processing power of computing hardware and making it difficult to meet the needs of real-time path planning. Numerical optimization methods for motion path planning have a low success rate and suffer from technical problems such as motion path generation failures. Furthermore, it is difficult to find a universal potential function applicable to a wide variety of path planning scenarios using artificial potential field methods, and they heavily rely on expert experience. Simultaneously, since artificial potential fields use repulsive forces to guide the path away from obstacles, these repulsive forces can be canceled out by attractive forces. Therefore, planning motion paths based on artificial potential fields is difficult to achieve precise control of the target object to avoid obstacles, posing certain safety hazards.

[0063] Embodiments of this disclosure provide a path planning method, a target object motion control method, an apparatus, an electronic device, a storage medium, and a computer program product. The path planning method includes: performing obstacle detection on an initial image representing the movement space of a target object to obtain an intermediate image, wherein the intermediate image includes obstacle image regions representing obstacles in the movement space of the target object, and the starting position and stopping position of the target object; extracting image features from the intermediate image to obtain intermediate image features; and determining a motion path for the target object to move from the starting position to the stopping position based on the intermediate image features.

[0064] According to embodiments of this disclosure, by performing obstacle detection on the initial image, the resulting intermediate image can accurately represent the obstacles in the target object's movement space. By extracting image features from the intermediate image and determining the movement path from the starting position to the stopping position based on the obtained intermediate image features, the obtained movement path can accurately avoid obstacles. Thus, the target object can be controlled to perform movement operations based on the movement path, enabling the target object to accurately avoid obstacles in its movement space during movement, thereby improving the obstacle avoidance accuracy.

[0065] Figure 1 The illustration schematically depicts an exemplary system architecture to which path planning methods and apparatus can be applied according to embodiments of the present disclosure. It should be noted that... Figure 1 The examples shown are merely examples of system architectures that can be applied to the embodiments of this disclosure, in order to help those skilled in the art understand the technical content of this disclosure, but do not mean that the embodiments of this disclosure cannot be used in other devices, systems, environments or scenarios.

[0066] like Figure 1 As shown, the system architecture 100 according to this embodiment may include terminal devices 101 and 102, vehicle 103, network 104, and server 105. Network 104 is used as a medium to provide a communication link between terminal devices 101 and 102, vehicle 103, and server 105. Network 104 may include various connection types, such as wired and / or wireless communication links, etc.

[0067] Users can use terminal devices 101, 102, and vehicle 103 to interact with server 105 via network 104 to receive or send messages, etc. Various communication client applications can be installed on terminal devices 101 and 102, such as shopping applications, web browser applications, search applications, instant messaging tools, email clients, and / or social media platforms (for example only). Vehicle 103 can be equipped with corresponding client applications or information collection devices.

[0068] Terminal devices 101, 102, and vehicle 103 can be various electronic devices with displays and web browsing capabilities, including but not limited to smartphones, tablets, laptops, and desktop computers.

[0069] Server 105 can be a server that provides various services, such as a backend management server that supports websites browsed by users using terminal devices 101 and 102 (for example only). The backend management server can analyze and process data such as received user requests, and feed back the processing results (such as web pages, information, or data obtained or generated according to user requests) to the terminal devices. Server 105 can also be a server that processes information collected or detected by vehicle 103. Server 105 can process information such as image information collected by vehicle 103 to generate motion paths, or server 105 can generate motion paths based on images collected by other image acquisition devices, or server 105 can control vehicle 103 to perform motion operations based on the generated motion paths.

[0070] It should be noted that the path planning method provided in this embodiment can generally be executed by vehicle 103. Accordingly, the path planning device provided in this embodiment can generally be installed in vehicle 103. Alternatively, the path planning method provided in this embodiment can also be executed by terminal devices 101 and 102. Accordingly, the path planning device provided in this embodiment can also be installed in terminal devices 101 and 102. Alternatively, the path planning method provided in this embodiment can also be executed by server 105. Accordingly, the path planning device provided in this embodiment can also be installed in server 105.

[0071] The path planning method provided in this disclosure can also be executed by a server or server cluster that is different from server 105 and is capable of communicating with terminal devices 101, 102, vehicle 103 and / or server 105. Correspondingly, the path planning device provided in this disclosure can also be set in a server or server cluster that is different from server 105 and is capable of communicating with terminal devices 101, 102, vehicle 103 and / or server 105.

[0072] It should be understood that Figure 1 The number of terminal devices, networks, and servers shown is merely illustrative. Depending on implementation needs, any number of terminal devices, networks, and servers can be included.

[0073] Figure 2 A flowchart illustrating a path planning method according to an embodiment of the present disclosure is shown schematically.

[0074] like Figure 2 As shown, the path planning method includes operations S210 to S230.

[0075] In operation S210, obstacle detection is performed on the initial image representing the movement space of the target object to obtain an intermediate image. The intermediate image includes obstacle image regions representing obstacles in the movement space of the target object, as well as the starting position and stopping position of the target object.

[0076] In operation S220, image features are extracted from the intermediate image to obtain the intermediate image features.

[0077] In operation S230, the motion path of the target object from the starting position to the stopping position is determined based on the features of the intermediate image.

[0078] According to embodiments of this disclosure, the initial image may include a bird's-eye view image representing the movement space of the target object. The bird's-eye view image may include obstacles and their locations within the movement space, as well as the starting and ending positions of the target object. However, it is not limited to this; the initial image may also include a map image representing the movement space of the target object. For example, if the movement space of the target object is a space constituted by a logistics warehouse, the initial image may be represented by a top-down layout diagram of the logistics warehouse.

[0079] It should be noted that the embodiments of this disclosure do not limit the specific image type of the initial image. Those skilled in the art can choose according to actual needs, as long as it can represent the starting and stopping positions of the target object and the obstacles in the target object's movement space.

[0080] According to embodiments of this disclosure, the initial image can also be generated after image acquisition or target detection of the target object's movement space from a top-down angle or other angles. For example, it can be generated after image acquisition of the space inside the logistics warehouse based on an image acquisition device set in the corner of the logistics warehouse. Embodiments of this disclosure do not limit the image acquisition method or angle of the initial image; those skilled in the art can select according to actual needs.

[0081] According to embodiments of this disclosure, the obstacle image region can be an image region capable of representing at least one obstacle attribute information, such as the position, shape, and size of the obstacle in the initial image. By extracting image features from the intermediate image, the obtained intermediate image features can more fully retain the obstacle attribute information in the intermediate image, thereby enabling the motion path determined based on the intermediate image features to more accurately avoid obstacles in the target object's movement space.

[0082] It should be understood that the positional attribute information contained in the intermediate image, such as the starting position, ending position, and obstacle position, can be represented by the pixel position of the intermediate image. Alternatively, it can be represented by the image coordinate position corresponding to the intermediate image. The embodiments of this disclosure do not limit the specific method of representing the position of the intermediate image, and those skilled in the art can choose according to actual needs.

[0083] According to embodiments of this disclosure, obstacle detection in an initial image can be achieved by detecting pixels representing obstacles in that initial image. However, this is not limited to this method; obstacle detection can also be performed on the initial image based on relevant object detection models, such as the R-CNN (Region with Convolutional Neural Networks features) model. The embodiments of this disclosure do not limit the specific method for obstacle detection. Those skilled in the art can choose the appropriate method according to actual needs, as long as obstacles in the initial image can be detected and the obstacle image region obtained.

[0084] According to embodiments of this disclosure, intermediate image features can be processed based on neural network algorithms, such as recurrent neural network algorithms, but not limited to this. Other types of neural network algorithms can also be used to process intermediate image features. Embodiments of this disclosure do not limit the specific type of algorithm for processing intermediate image features, and those skilled in the art can choose according to actual needs.

[0085] According to embodiments of this disclosure, by performing obstacle detection on the initial image, the resulting intermediate image can accurately represent the obstacles in the target object's movement space. By extracting image features from the intermediate image and determining the motion path from the starting position to the stopping position based on the obtained intermediate image features, the motion path can accurately avoid obstacles. Thus, by controlling the target object to perform motion operations based on the motion path, the target object can accurately avoid obstacles in its movement space during motion, thereby improving the obstacle avoidance accuracy during the motion control of the target object.

[0086] The following describes specific embodiments and references. Figures 3-5 right Figure 2 The method shown will be further explained.

[0087] According to embodiments of this disclosure, the initial image includes an initial obstacle image region that characterizes the shape of the obstacle.

[0088] Figure 3 The flowchart illustrating obstacle detection of an initial image characterizing the movement space of a target object according to an embodiment of the present disclosure is shown schematically.

[0089] like Figure 3 As shown, in operation S210, obstacle detection is performed on the initial image representing the movement space of the target object, including operations S310 to S330.

[0090] In operation S310, the initial image is meshed to obtain an initial mesh image, wherein the initial mesh image includes multiple initial mesh image units.

[0091] In operation S320, overlap detection is performed on the initial grid image unit and the initial obstacle image region to obtain the overlap detection result.

[0092] In operation S330, the initial pixel values ​​of each of the initial grid image units are updated according to the overlap detection results of each of the initial grid image units to obtain the intermediate image.

[0093] According to embodiments of this disclosure, the initial obstacle image region can be constructed by the pixel values ​​of pixels representing obstacles, but it is not limited to this. The initial obstacle image region can also be generated by annotating the initial image based on a preset obstacle region or obstacle shape. The embodiments of this disclosure do not limit the specific method of determining the initial obstacle image region, and those skilled in the art can choose according to actual needs.

[0094] According to embodiments of this disclosure, an initial image can be meshed based on a mesh diagram containing multiple mesh cells to obtain the regions of the initial image corresponding to the mesh cells, i.e., the initial mesh image units surrounded by the mesh cells.

[0095] According to embodiments of this disclosure, the overlap detection result can characterize the degree of overlap between the initial mesh image unit and the initial obstacle image region. For example, if the initial mesh image unit and the initial obstacle image region completely overlap, the overlap detection result can be 1; if the initial mesh image unit and the initial obstacle image region do not overlap, the overlap detection result can be 0. Accordingly, the overlap detection result can also include a value between 0 and 1 to accurately characterize the degree of overlap between the initial mesh image unit and the initial obstacle image region.

[0096] According to embodiments of this disclosure, the overlap detection result can also characterize whether the initial grid image unit overlaps with the initial obstacle image region by representing a positive detection result that represents partial overlap and a negative detection result that represents no overlap.

[0097] It should be noted that the embodiments of this disclosure do not limit the specific setting method of the overlap detection result. Those skilled in the art can choose according to actual needs, as long as it can satisfy the requirement of characterizing the degree of overlap between the initial grid image unit and the initial obstacle image region.

[0098] According to embodiments of this disclosure, an intermediate image value corresponding to the overlap detection result can be determined based on the overlap detection result to obtain an updated intermediate image.

[0099] In one embodiment of this disclosure, a preset overlap range can be set. When the overlap detection result is 0, the initial grid image unit is updated to the green pixel value corresponding to the green pixel. When the overlap detection result is (0, 0.4), the initial grid image unit is updated to the yellow pixel value corresponding to the yellow pixel. Correspondingly, when the overlap detection result is (0.4, 1), the initial grid image unit is updated to the red pixel value corresponding to the red pixel. The regions in the intermediate image with red and green pixel values ​​can be image regions representing obstacles, thereby enabling the clear representation of obstacle attribute information such as position, shape, and size in the target object's movement space based on the intermediate image.

[0100] According to embodiments of this disclosure, the motion path may include path information generated in an intermediate image, or it may also include path information generated in an initial image, or it may also include path information generated in a preset coordinate system. Embodiments of this disclosure do not limit the specific representation method of the motion path, and those skilled in the art can choose according to actual needs.

[0101] According to embodiments of this disclosure, by updating the initial pixel values ​​based on the overlap detection results, the pixel values ​​in each updated grid image can clearly characterize the obstacle attribute information of obstacles in the target object's movement space. This allows the subsequently extracted intermediate image features to more accurately retain obstacle attribute information, avoid the loss of obstacle attribute information, improve the accuracy of the generated motion path in avoiding obstacles in the target object's movement space, and ultimately improve the accuracy and success rate of the target object avoiding obstacles during motion operations.

[0102] It should be noted that the above embodiments are merely illustrative of the method for updating the initial pixel value of the initial grid image unit in the embodiments of this disclosure, and are not intended to limit the method provided in the embodiments of this disclosure. Those skilled in the art can update the initial pixel value of the initial grid image unit according to actual needs.

[0103] According to an embodiment of this disclosure, in operation S330, updating the initial pixel value of each of the initial grid image units based on the overlap detection result corresponding to each of the initial grid image units may include the following operations.

[0104] If the overlap detection result corresponding to the initial mesh image unit is a first detection result, the initial pixel value of the initial mesh image unit is updated to a first pixel value, wherein the first detection result indicates that the initial mesh image unit overlaps with the initial obstacle image region; and if the overlap detection result corresponding to the initial mesh image unit is a second detection result, the initial pixel value of the initial mesh image unit is updated to a second pixel value, wherein the second detection result indicates that the initial mesh image unit is spaced apart from the initial obstacle image region.

[0105] According to embodiments of this disclosure, updating the initial grid image with a first pixel value and a second pixel value based on a first detection result (positive detection result) and a second detection result (negative detection result) respectively can reduce the number of pixel value types in the obtained intermediate image, enabling the intermediate image to clearly distinguish between obstacles and non-obstacles in the target object's movement space, and reducing the error in the intermediate image's representation of obstacles.

[0106] It should be noted that the first pixel value and the second pixel value can be determined based on the choice of those skilled in the art. The embodiments of this disclosure do not limit the specific setting method of the first pixel value and the second pixel value, as long as it can distinguish between obstacles and non-obstacles.

[0107] According to an embodiment of this disclosure, in operation S320, the overlap detection of the initial grid image unit and the initial obstacle image region, and the resulting overlap detection result, may include the following operations.

[0108] If the initial grid image cell and the initial obstacle image region at least partially overlap, the overlap detection result is determined as the first detection result.

[0109] According to embodiments of this disclosure, the first detection result can characterize the overlap between the initial grid image and the initial obstacle image region. By determining the overlap detection result as the first detection result when the initial grid image unit and the initial obstacle image region at least partially overlap, the determination speed of the first detection result can be accelerated, thereby enabling rapid generation of intermediate images and improving the efficiency of subsequent motion path generation. Simultaneously, the image region characterizing the obstacle in the intermediate image can be appropriately expanded, avoiding the probability that the subsequent motion path cannot bypass the obstacle due to low obstacle detection accuracy, thus increasing the success rate of the target object avoiding obstacles during subsequent motion operations.

[0110] According to an embodiment of this disclosure, in operation S320, the overlap detection of the initial grid image unit and the initial obstacle image region, and the resulting overlap detection result, may include the following operations.

[0111] If the overlap between the initial grid image unit and the initial obstacle image region is greater than a preset overlap threshold, the overlap detection result is determined as the first detection result.

[0112] Determining the first detection result using the method provided in the embodiments of this disclosure can improve the detection accuracy of the intermediate image for obstacle attribute information, thereby improving the representation accuracy of the intermediate image for obstacles in the target object's movement space, enabling the subsequently generated motion path to reduce the distance between it and the obstacle, and improving the motion efficiency of the target object in performing motion operations.

[0113] According to embodiments of this disclosure, the starting and stopping positions in the initial image can also be updated to third pixel values, thereby effectively distinguishing the starting and stopping positions from the first and second pixel values, improving the planning accuracy of the motion path, and enabling the target object to move accurately from the starting position to the stopping position according to the path planning method provided in the embodiments of this disclosure.

[0114] Figure 4 The flowchart illustrating obstacle detection of an initial image characterizing the movement space of a target object is shown schematically according to another embodiment of the present disclosure.

[0115] like Figure 4 As shown, in operation S210, obstacle detection on the initial image representing the movement space of the target object may include operations S410 to S420.

[0116] In operation S410, the initial image is processed based on the object detection algorithm to obtain object detection boxes representing obstacles.

[0117] In operation S420, the intermediate image is determined based on the target detection box.

[0118] According to embodiments of this disclosure, the target detection algorithm may include algorithms used in related technologies for detecting targets such as vehicles and signs in image information, such as the SMOKE (Single-Stage Monocular 3D Object Detection via Keypoint Estimation) algorithm. The embodiments of this disclosure do not limit the specific type of target detection algorithm, and those skilled in the art can select one according to actual needs.

[0119] According to embodiments of this disclosure, the target detection box can at least partially characterize obstacle attribute information such as the type, location, and size of the obstacle. Furthermore, the target detection box can completely enclose the obstacle, preventing situations where the detected obstacle size is too small, causing the target object to collide with the obstacle during subsequent movement operations, thus improving the success rate of obstacle avoidance. Moreover, the target detection box has high generation efficiency and strong generalization ability, making it suitable for detecting various types of obstacles. Therefore, determining the intermediate image containing the obstacle image region based on the target detection box is applicable to scenarios where the location or type of obstacles in the target object's movement space changes frequently, improving the adaptability of the target object's movement operations.

[0120] According to an embodiment of this disclosure, in operation S230, determining the motion path of the target object from the starting position to the stopping position based on the intermediate image features may include the following operations.

[0121] The intermediate image features are input into the path prediction model built based on the neural network algorithm, and the output is the path position between the starting position and the stopping position; and the motion path is determined based on the path position.

[0122] According to embodiments of this disclosure, by processing intermediate image features through a path prediction model, the trajectory of the target object between the starting position and the stopping position can be characterized by rapidly generated path positions. Then, the motion path can be determined based on the path positions. This enables the relevant motion control module to efficiently and accurately control the target object to perform motion operations according to the motion path based on the path positions in the motion path, thereby improving the accuracy of the target object's motion and the precision of obstacle avoidance.

[0123] According to embodiments of this disclosure, the path prediction model includes at least one of the following:

[0124] Recurrent neural network model, long short-term memory network model, bidirectional long short-term memory network model, attention network model.

[0125] According to embodiments of this disclosure, the attention network model may include neural network models built based on attention mechanisms in related technologies, such as the Transformer model, BERT (Bidirectional Encoder Representation from Transformers) model, etc.

[0126] According to embodiments of this disclosure, by using a bidirectional long short-term memory network model to process intermediate image features, obstacle attribute information contained in the intermediate image features can be learned bidirectionally from the starting position to the stopping position and from the stopping position to the starting position, thereby generating a motion path that avoids obstacles with higher accuracy.

[0127] According to embodiments of this disclosure, the target object includes at least one of the following:

[0128] Unmanned vehicles and intelligent robots.

[0129] According to embodiments of this disclosure, unmanned vehicles may include unmanned logistics vehicles, unmanned passenger vehicles, unmanned boats, and other transportation vehicles in related technologies. The embodiments of this disclosure do not limit the specific type of unmanned vehicle, as long as it can perform motion operations according to the motion path.

[0130] According to the embodiments of this disclosure, intelligent robots may include any type of robot with motion functions, such as inspection robots and logistics robots. The embodiments of this disclosure do not limit the specific type of intelligent robot, as long as it can perform motion operations according to the motion path.

[0131] Figure 5 The diagram illustrates an application scenario of the path planning method according to an embodiment of the present disclosure.

[0132] like Figure 5 As shown, in the application scenario of this embodiment, an initial image 510 representing the movement space of the target object may be included. The initial image 510 may be a planar map image viewed from above the movement space of the target object. The initial image 510 may include obstacle objects 511a, 511b, and 511c, and correspondingly may also include a starting position object 512 and a stopping position object 513. An image coordinate system including an X-axis and a Y-axis may be set for the initial image 510. The obstacle position information, obstacle shape information, and obstacle size information of the obstacle objects 511a, 511b, and 511c, as well as the starting position object 512 and the stopping position object 513, can be represented by coordinate elements corresponding to the image coordinate system.

[0133] The initial image 510 is input into the meshing processing model 521 of the path planning module 520, which can perform meshing processing on the initial image 510 based on the grid-like mesh, and output multiple initial mesh image units to achieve the discretization of the initial image 510.

[0134] Multiple initial mesh image units can be input into an intermediate image generation model 522. Obstacle detection is performed on all initial mesh image units based on the intermediate image generation model 522. For example, overlap detection can be performed on the initial mesh image units and initial obstacle image regions (i.e., obstacle objects 511a, 511b, and 511c) to obtain an overlap detection result. If the overlap detection result indicates that the initial mesh image and the initial obstacle image region at least partially overlap, the overlap detection result is determined as the first detection result. Then, the pixel value of the initial mesh image unit corresponding to the first detection result is updated to 0.

[0135] Accordingly, the initial mesh image units corresponding to the starting position object 512 and the stopping position object 513 can be updated to a pixel value of 127, and the initial mesh image units that do not overlap with the starting position object 512 and / or the stopping position object 513 can be updated to a pixel value of 255, thereby obtaining an intermediate image. The regions with a pixel value of 255 (i.e., white regions) in the intermediate image can be areas where the target object can move, while the regions with a pixel value of 0 (i.e., black regions) can represent the obstacle attributes such as the position, size, and shape of the obstacle objects 511a, 511b, and 511c. Simultaneously, the regions with a pixel value of 127 in the intermediate image can represent the starting position object 512 or the stopping position object 513, thus enabling the intermediate image to accurately represent the movement space of the target object.

[0136] The generated intermediate image is input into the image feature extraction model 523, which can be constructed based on a convolutional neural network model to extract the image features of the intermediate image and output the intermediate image features.

[0137] The intermediate image features can be input into the path prediction model 524, which can be constructed based on a Long Short-Term Memory (LSTM) model. This model can predict the path positions of the target object between the starting position object 512 and the stopping position object 513 based on the intermediate image features. For example, it can output multiple path position coordinates between the starting position object 512 and the stopping position object 513, obtaining a path position coordinate sequence. The intermediate image can be updated based on this path position coordinate sequence to obtain the target image 530.

[0138] The target image 530 can be generated based on an intermediate image. The target image 530 includes target obstacles 531a, 531b, and 531c with updated pixel values, and may also include a target starting position object 532 and a target stopping position object 533. Furthermore, the target image 530 may include a motion path object 540 representing the motion path. The motion path object 540 can be generated based on the path prediction model 524's output path position coordinate sequence, allowing relevant personnel to clearly view the motion trajectory of the target moving object in the target's movement space and successfully avoid obstacles in the target's movement space. This achieves motion path generation through an end-to-end module (i.e., path planning module 520), improving motion path planning efficiency.

[0139] It should be noted that the image feature extraction model 523 and the path prediction model 524 in the path planning module 520 can be trained based on supervised training methods in related technologies.

[0140] For example, in actual autonomous vehicle operation, a map of the autonomous vehicle's movement path can be accumulated. Initial sample images (map images without movement paths) and sample labels (sample movement paths) are generated. The initial image feature extraction model and the initial path prediction model are trained using the open-source deep learning library PyTorch. The predicted path location coordinates output by the initial path prediction model are... The coordinates of the sample path location representing the sample label are: n represents the number of path location coordinates. The following formula (2) is used as the cost function to form the supervision signal. This yields the trained image feature extraction model 523 and path prediction model 524.

[0141]

[0142] In formula (2), n represents the number of path location coordinates and N represents the number of sample movement paths.

[0143] The trained image feature extraction model 523 and path prediction model 524 can transform the path search problem in related technologies into a regression problem. Based on the trained image feature extraction model and path prediction model, the end-to-end output of motion paths can be achieved, improving the efficiency of motion path generation and the accuracy of obstacle avoidance.

[0144] It should be noted that the data (e.g., image information of the target object's mobile space) involved in the technical solution disclosed herein, including its collection, use, storage, sharing, and transfer, all comply with the provisions of relevant laws and regulations, and require notification to relevant users or institutions and obtaining their consent or authorization. Where applicable, the personal information of users involved has been de-identified and / or anonymized and / or encrypted.

[0145] Figure 6a A flowchart illustrating a target object motion control method according to an embodiment of the present disclosure is shown schematically.

[0146] like Figure 6a As shown, the motion control method for the target object includes operations S610 to S620.

[0147] In operation S610, the motion path is determined according to the path planning method provided in the above embodiment.

[0148] When operating S620, the target object is controlled to perform motion operations according to the motion path. Figure 6b The diagram illustrates an application scenario of the target object motion control method according to an embodiment of the present disclosure.

[0149] like Figure 6b As shown, in the application scenario of this embodiment, there may be a target object movement space 610, which includes obstacles 611a, 611b and 611c, and sets the starting position 612 and the stopping position 613 of the target object 620.

[0150] By acquiring images of the target object's movement space 610 from a top-down perspective, an initial image can be obtained. According to the target object motion control method provided in this embodiment, a motion path between the starting position 612 and the stopping position 613 can be generated.

[0151] By controlling the target object 620 to perform motion operations through the generated motion path 630, the target object 620 can be controlled to perform motion operations according to the motion path trajectory 630, and finally controlled to move to the stop position 613. Obviously, the target object 620 can move from the starting position 612 to the stop position 613 efficiently while accurately avoiding obstacles 611a, 611b and 611c, thus achieving efficient control of the target object 620.

[0152] Figure 7 A block diagram of a path planning apparatus according to an embodiment of the present disclosure is shown schematically.

[0153] like Figure 7As shown, the path planning 700 includes a detection module 710, an image feature extraction module 720, and a first motion path determination module 730.

[0154] The detection module 710 is used to perform obstacle detection on the initial image representing the movement space of the target object to obtain an intermediate image. The intermediate image includes an obstacle image region representing obstacles in the movement space of the target object, as well as the starting position and stopping position of the target object.

[0155] The image feature extraction module 720 is used to extract image features from the intermediate image to obtain the intermediate image features.

[0156] The first motion path determination module 730 is used to determine the motion path of the target object from the starting position to the stopping position based on the features of the intermediate image.

[0157] According to embodiments of this disclosure, the initial image includes an initial obstacle image region that characterizes the shape of the obstacle.

[0158] The detection module includes: a meshing processing unit, an overlap detection unit, and a pixel value update unit.

[0159] The meshing processing unit is used to mesh the initial image to obtain an initial mesh image, wherein the initial mesh image includes multiple initial mesh image units.

[0160] The overlap detection unit is used to detect the overlap between the initial grid image unit and the initial obstacle image region, and obtain the overlap detection result.

[0161] The pixel value update unit is used to update the initial pixel value of each of the initial grid image units according to the overlap detection result of each of the initial grid image units, so as to obtain the intermediate image.

[0162] According to embodiments of this disclosure, the pixel value update unit includes: a first pixel value update subunit and a second pixel value update subunit.

[0163] The first pixel value update subunit is used to update the initial pixel value of the initial grid image unit to the first pixel value when the overlap detection result corresponding to the initial grid image unit is the first detection result, wherein the first detection result indicates that the initial grid image unit overlaps with the initial obstacle image region.

[0164] The second pixel value update subunit is used to update the initial pixel value of the initial grid image unit to the second pixel value when the overlap detection result corresponding to the initial grid image unit is the second detection result, wherein the second detection result represents the initial grid image unit, which is spaced apart from the initial obstacle image region.

[0165] According to embodiments of this disclosure, the overlap detection unit includes a first overlap detection subunit.

[0166] The first overlap detection subunit is used to determine the overlap detection result as the first detection result when the initial grid image unit and the initial obstacle image region at least partially overlap.

[0167] According to embodiments of this disclosure, the overlap detection unit includes a second overlap detection subunit.

[0168] The second overlap detection subunit is used to determine the overlap detection result as the first detection result when the overlap between the initial grid image unit and the initial obstacle image region is greater than a preset overlap threshold.

[0169] According to embodiments of this disclosure, the detection module includes: a target detection box determination unit and an intermediate image generation unit.

[0170] The target detection box determination unit is used to process the initial image based on the target detection algorithm to obtain target detection boxes representing obstacles.

[0171] The intermediate image generation unit is used to determine the intermediate image based on the target detection box.

[0172] According to embodiments of this disclosure, the motion path determination module includes: a path location determination unit and a motion path determination unit.

[0173] The path location determination unit is used to input intermediate image features into the path prediction model built based on a neural network algorithm, and output the path location between the starting position and the stopping position.

[0174] The motion path determination unit is used to determine the motion path based on the location along the way.

[0175] According to embodiments of this disclosure, the path prediction model includes at least one of the following:

[0176] Recurrent neural network model, long short-term memory network model, bidirectional long short-term memory network model, attention network model.

[0177] According to embodiments of this disclosure, the target object includes at least one of the following:

[0178] Unmanned vehicles and intelligent robots.

[0179] Figure 8 A block diagram of a target object motion control device according to an embodiment of the present disclosure is shown schematically.

[0180] like Figure 8As shown, the target object motion control device 800 includes a second motion path determination module 810 and a motion operation control module 820. Any plurality of modules, units, and sub-units according to embodiments of the present disclosure, or at least part of the functions of any plurality of such modules, units, and sub-units, can be implemented in one module. Any one or more modules, units, and sub-units according to embodiments of the present disclosure can be implemented by being divided into multiple modules. Any one or more modules, units, and sub-units according to embodiments of the present disclosure can be at least partially implemented as hardware circuits, such as field-programmable gate arrays (FPGAs), programmable logic arrays (PLAs), systems-on-a-chip, systems-on-a-substrate, systems-on-package, application-specific integrated circuits (ASICs), or implemented by hardware or firmware in any other reasonable manner by integrating or packaging circuits, or implemented in any one of software, hardware, and firmware, or in a suitable combination of any of these. Alternatively, one or more modules, units, and sub-units according to embodiments of the present disclosure can be at least partially implemented as computer program modules, which, when run, can perform corresponding functions.

[0181] For example, any multiple of the detection module 710, image feature extraction module 720, and first motion path determination module 730, or second motion path determination module 810 and motion operation control module 820, can be combined into one module / unit / subunit, or any one of these modules / units / subunits can be split into multiple modules / units / subunits. Alternatively, at least some of the functions of one or more of these modules / units / subunits can be combined with at least some of the functions of other modules / units / subunits and implemented in one module / unit / subunit. According to embodiments of this disclosure, at least one of the detection module 710, image feature extraction module 720, and first motion path determination module 730, or second motion path determination module 810 and motion operation control module 820, can be at least partially implemented as hardware circuitry, such as a field-programmable gate array (FPGA), programmable logic array (PLA), system-on-a-chip, system-on-a-substrate, system-on-package, application-specific integrated circuit (ASIC), or any other reasonable method of integrating or packaging circuitry, or implemented in software, hardware, or firmware, or in any suitable combination of any of these three methods. Alternatively, at least one of the detection module 710, image feature extraction module 720, and first motion path determination module 730, or second motion path determination module 810 and motion operation control module 820, can be at least partially implemented as a computer program module, which, when run, can perform corresponding functions.

[0182] It should be noted that the target object motion control device part in the embodiments of this disclosure corresponds to the target object motion control method part in the embodiments of this disclosure. For a detailed description of the target object motion control device part, please refer to the target object motion control method part, which will not be repeated here.

[0183] It should be noted that the path planning device part in the embodiments of this disclosure corresponds to the path planning method part in the embodiments of this disclosure. The specific description of the path planning device part is referred to in the path planning method part, and will not be repeated here.

[0184] Figure 9 A block diagram of an electronic device suitable for implementing a path planning method and a target object motion control method according to embodiments of the present disclosure is shown schematically. Figure 9 The electronic device shown is merely an example and should not be construed as limiting the functionality and scope of the embodiments disclosed herein.

[0185] like Figure 9 As shown, an electronic device 900 according to an embodiment of the present disclosure includes a processor 901, which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 902 or a program loaded from a storage portion 908 into a random access memory (RAM) 903. The processor 901 may include, for example, a general-purpose microprocessor (e.g., a CPU), an instruction set processor and / or an associated chipset and / or a special-purpose microprocessor (e.g., an application-specific integrated circuit (ASIC)), etc. The processor 901 may also include onboard memory for caching purposes. The processor 901 may include a single processing unit or multiple processing units for performing different actions of the method flow according to an embodiment of the present disclosure.

[0186] RAM 903 stores various programs and data required for the operation of electronic device 900. Processor 901, ROM 902, and RAM 903 are interconnected via bus 904. Processor 901 performs various operations of the method flow according to embodiments of the present disclosure by executing programs in ROM 902 and / or RAM 903. It should be noted that the programs may also be stored in one or more memories other than ROM 902 and RAM 903. Processor 901 may also perform various operations of the method flow according to embodiments of the present disclosure by executing programs stored in said one or more memories.

[0187] According to embodiments of this disclosure, the electronic device 900 may further include an input / output (I / O) interface 905, which is also connected to a bus 904. The system 900 may also include one or more of the following components connected to the I / O interface 905: an input section 906 including a keyboard, mouse, etc.; an output section 907 including a cathode ray tube (CRT), liquid crystal display (LCD), etc., and a speaker, etc.; a storage section 908 including a hard disk, etc.; and a communication section 909 including a network interface card such as a LAN card, modem, etc. The communication section 909 performs communication processing via a network such as the Internet. A drive 910 is also connected to the I / O interface 905 as needed. A removable medium 911, such as a disk, optical disk, magneto-optical disk, semiconductor memory, etc., is installed on the drive 910 as needed so that computer programs read from it can be installed into the storage section 908 as needed.

[0188] According to embodiments of this disclosure, the method flow according to embodiments of this disclosure can be implemented as a computer software program. For example, embodiments of this disclosure include a computer program product comprising a computer program carried on a computer-readable storage medium, the computer program containing program code for performing the methods shown in the flowchart. In such embodiments, the computer program can be downloaded and installed from a network via communication section 909, and / or installed from removable medium 911. When the computer program is executed by processor 901, it performs the functions defined in the system of embodiments of this disclosure. According to embodiments of this disclosure, the systems, devices, apparatuses, modules, units, etc., described above can be implemented by computer program modules.

[0189] This disclosure also provides a computer-readable storage medium, which may be included in the device / apparatus / system described in the above embodiments; or it may exist independently and not assembled into the device / apparatus / system. The computer-readable storage medium carries one or more programs that, when executed, implement the method according to the embodiments of this disclosure.

[0190] According to embodiments of this disclosure, the computer-readable storage medium can be a non-volatile computer-readable storage medium. Examples include, but are not limited to: portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof. In this disclosure, the computer-readable storage medium can be any tangible medium that contains or stores a program that can be used by or in conjunction with an instruction execution system, apparatus, or device.

[0191] For example, according to embodiments of this disclosure, a computer-readable storage medium may include the ROM 902 and / or RAM 903 described above and / or one or more memories other than ROM 902 and RAM 903.

[0192] Embodiments of this disclosure also include a computer program product comprising a computer program containing program code for performing the methods provided in the embodiments of this disclosure. When the computer program product is run on an electronic device, the program code is used to enable the electronic device to implement the path planning method and the target object motion control method provided in the embodiments of this disclosure.

[0193] When the computer program is executed by the processor 901, it performs the functions defined in the system / apparatus of this disclosure embodiments. According to embodiments of this disclosure, the systems, apparatuses, modules, units, etc., described above can be implemented by computer program modules.

[0194] In one embodiment, the computer program may rely on a tangible storage medium such as an optical storage device or a magnetic storage device. In another embodiment, the computer program may also be transmitted and distributed in the form of signals over a network medium, and downloaded and installed via the communication section 909, and / or installed from a removable medium 911. The program code contained in the computer program can be transmitted using any suitable network medium, including but not limited to: wireless, wired, etc., or any suitable combination thereof.

[0195] According to embodiments of this disclosure, program code for executing the computer programs provided in embodiments of this disclosure can be written in any combination of one or more programming languages. Specifically, these computational programs can be implemented using high-level procedural and / or object-oriented programming languages, and / or assembly / machine languages. Programming languages ​​include, but are not limited to, languages ​​such as Java, C++, Python, "C", or similar programming languages. The program code can execute entirely on the user's computing device, partially on the user's device, partially on a remote computing device, or entirely on a remote computing device or server. In cases involving remote computing devices, the remote computing device can be connected to the user's computing device via any type of network, including a local area network (LAN) or a wide area network (WAN), or it can be connected to an external computing device (e.g., via the Internet using an Internet service provider).

[0196] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of the present disclosure. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two consecutively indicated blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in a block diagram or flowchart, and combinations of blocks in a block diagram or flowchart, may be implemented using a dedicated hardware-based system that performs the specified function or operation, or using a combination of dedicated hardware and computer instructions. Those skilled in the art will understand that the features recited in the various embodiments and / or claims of this disclosure can be combined and / or combined in various ways, even if such combinations or combinations are not expressly described in this disclosure. In particular, the features described in the various embodiments and / or claims of this disclosure may be combined and / or combined in various ways without departing from the spirit and teachings of this disclosure. All such combinations and / or combinations fall within the scope of this disclosure.

[0197] The embodiments of this disclosure have been described above. However, these embodiments are for illustrative purposes only and are not intended to limit the scope of this disclosure. Although various embodiments have been described above, this does not mean that the measures in the various embodiments cannot be used advantageously in combination. The scope of this disclosure is defined by the appended claims and their equivalents. Various substitutions and modifications can be made by those skilled in the art without departing from the scope of this disclosure, and all such substitutions and modifications should fall within the scope of this disclosure.

Claims

1. A path planning method, comprising: Obstacle detection is performed on an initial image representing the movement space of a target object to obtain an intermediate image, wherein the intermediate image includes an obstacle image region representing obstacles in the movement space of the target object, as well as the starting position and the stopping position of the target object; Image feature extraction is performed on the intermediate image to obtain intermediate image features; and Based on the intermediate image features, determine the motion path of the target object from the starting position to the stopping position; Obstacle detection on the initial image representing the target object's movement space includes: The initial image is processed based on an object detection algorithm to obtain object detection boxes surrounding the obstacle. These boxes characterize the obstacle type and size. The intermediate image is determined based on the target detection bounding box.

2. The method according to claim 1, wherein, The initial image includes an initial obstacle image region that characterizes the shape of the obstacle; Obstacle detection on the initial image representing the movement space of the target object includes: The initial image is subjected to meshing processing to obtain an initial mesh image, wherein the initial mesh image includes multiple initial mesh image units; Overlap detection is performed on the initial grid image unit and the initial obstacle image region to obtain the overlap detection result; and Based on the overlap detection results corresponding to each of the initial grid image units, the initial pixel values ​​of each of the initial grid image units are updated to obtain the intermediate image.

3. The method according to claim 2, wherein, Updating the initial pixel values ​​of each of the initial grid image units based on the overlap detection results includes: If the overlap detection result corresponding to the initial mesh image unit is a first detection result, the initial pixel value of the initial mesh image unit is updated to the first pixel value, wherein the first detection result indicates that the initial mesh image unit overlaps with the initial obstacle image region; If the overlap detection result corresponding to the initial mesh image unit is the second detection result, the initial pixel value of the initial mesh image unit is updated to the second pixel value, wherein the second detection result indicates that the initial mesh image unit is spaced apart from the initial obstacle image region.

4. The method according to claim 3, wherein, Overlap detection is performed on the initial grid image unit and the initial obstacle image region to obtain the overlap detection results, including: If the initial grid image unit at least partially overlaps with the initial obstacle image region, the overlap detection result is determined as the first detection result.

5. The method according to claim 3, wherein, Overlap detection is performed on the initial grid image unit and the initial obstacle image region to obtain the overlap detection results, including: If the overlap between the initial grid image unit and the initial obstacle image region is greater than a preset overlap threshold, the overlap detection result is determined as the first detection result.

6. The method according to claim 1, wherein, Determining the motion path of the target object from the starting position to the stopping position based on the intermediate image features includes: The intermediate image features are input into a path prediction model constructed based on a neural network algorithm, which outputs the path location between the starting position and the stopping position; and The movement path is determined based on the locations along the route.

7. The method according to claim 6, wherein, The path prediction model includes at least one of the following: Recurrent neural network model, long short-term memory network model, bidirectional long short-term memory network model, attention network model.

8. The method according to any one of claims 1 to 7, wherein, The target object includes at least one of the following: Unmanned vehicles and intelligent robots.

9. A method for motion control of a target object, comprising: The motion path is determined according to the method described in any one of claims 1 to 8; as well as Based on the motion path, control the target object to perform motion operations.

10. A path planning device, comprising: The detection module is used to perform obstacle detection on an initial image representing the movement space of a target object to obtain an intermediate image, wherein the intermediate image includes an obstacle image region representing obstacles in the movement space of the target object, as well as the starting position and the stopping position of the target object; An image feature extraction module is used to extract image features from the intermediate image to obtain intermediate image features; and The first motion path determination module is used to determine the motion path of the target object from the starting position to the stopping position based on the intermediate image features. The detection module is configured as follows: The initial image is processed based on an object detection algorithm to obtain object detection boxes surrounding the obstacle. These boxes characterize the obstacle type and size. The intermediate image is determined based on the target detection bounding box.

11. A target object motion control device, comprising: The second motion path determination module is used to determine the motion path according to the method described in any one of claims 1 to 8; as well as The motion operation control module is used to control the target object to perform motion operations according to the motion path.

12. An electronic device, comprising: One or more processors; Memory, used to store one or more programs. When the one or more programs are executed by the one or more processors, the one or more processors implement the method of any one of claims 1 to 9.

13. A computer-readable storage medium having stored thereon executable instructions that, when executed by a processor, cause the processor to perform the method of any one of claims 1 to 9.

14. A computer program product comprising a computer program that, when executed by a processor, implements the method according to any one of claims 1 to 9.

Images