Collision detection method and device, electronic equipment and computer readable storage medium

Abstract

The embodiment of the invention discloses a collision detection method and device, electronic equipment and a computer readable storage medium, and the method comprises the steps: carrying out the shape representation of a target object and an environment object according to the position information, and determining a first shape corresponding to the target object and a second shape corresponding to the environment object; moving the second shape along the boundary of the first shape, and determining a collision area of the environment object relative to the target object; determining a position relationship between the environment object and the collision area according to the position information of the environment object; and determining a collision detection result of the target object and the environmental object according to the position relationship. Therefore, in the embodiment, the collision detection result is determined by determining the collision area of the environment object relative to the target object and according to the position relation between the environment object and the collision area, and the determination efficiency and accuracy of the fault detection result can be improved.

Description

Technical Field

[0001] The present invention relates to the field of computer technologies, and particularly to a collision detection method, apparatus, electronic device, and computer-readable storage medium. Background Art

[0002] During the prediction of the moving trajectory of an object, collision detection can be used to predict the risk of collision between objects, which can help the object correct its moving trajectory in time and improve the safety of movement. Taking the autonomous driving scenario as an example, by performing collision detection on an autonomous driving vehicle and surrounding traffic participants or other objects, it can provide a reliable basis for the autonomous vehicle to more accurately and safely plan future trajectories.

[0003] Existing collision detection methods often determine whether two objects collide based on whether there is an overlap between the geometric shapes of different objects. However, when performing overlap calculations, it is necessary to consider geometric features in various situations such as whether a point is included in an edge and whether edges coincide. A large number of loop statements are required in the calculation process, resulting in high computational complexity and long time consumption, and the determination efficiency and accuracy of the collision detection results both need to be improved. Summary of the Invention

[0004] In view of this, the purpose of the embodiments of the present invention is to provide a collision detection method, apparatus, electronic device, and computer-readable storage medium to improve the determination efficiency and accuracy of the fault detection results.

[0005] In a first aspect, an embodiment of the present invention aims to provide a collision detection method, and the method includes:

[0006] Obtain the position information of a target object and an environmental object;

[0007] Perform shape representation on the target object and the environmental object according to the position information, and determine a first shape corresponding to the target object and a second shape corresponding to the environmental object;

[0008] Move the second shape along the boundary of the first shape to determine the collision area of the environmental object relative to the target object;

[0009] Determine the positional relationship between the environmental object and the collision area according to the position information of the environmental object, and the positional relationship is used to characterize the overlap situation between the environmental object and the collision area;

[0010] Determine the collision detection result of the target object and the environmental object according to the positional relationship.

[0011] Further, the first shape and the second shape are rectangles. Moving the second shape along the boundary of the first shape to determine the collision area of the environmental object relative to the target object includes:

[0012] Moving the second shape to be circumscribed with the first shape according to the position and orientation of the second shape;

[0013] Starting from the vertices of the first shape, alternately translating the second shape along the boundaries of the first shape and the second shape to determine at least one circumscribed shape;

[0014] Determining the collision area of the environmental object relative to the target object according to each of the circumscribed shapes.

[0015] Further, determining the collision area of the environmental object relative to the target object according to each of the circumscribed shapes includes:

[0016] Determining the center points of each of the circumscribed shapes as the vertices of the collision area;

[0017] Connecting the vertices to determine the closed area formed after connection as the collision area of the environmental object relative to the target object.

[0018] Further, the position information includes position coordinates and position orientation, and the position coordinates are used to represent the location of the target object or the environmental object.

[0019] Further, the contour of the first shape is circumscribed with the contour of the target object, the position orientation of the first shape corresponds to the position orientation of the target object, and the position coordinates of the first shape correspond to the position coordinates of the target object.

[0020] Further, determining the positional relationship between the environmental object and the collision area according to the position information of the environmental object includes:

[0021] Constructing a detection coordinate system according to the position coordinates and position orientation of the target object, with the origin of the detection coordinate system being the location of the target object. The detection coordinate system includes a first direction and a second direction, the first direction is consistent with the position orientation of the target object, and the second direction is perpendicular to the position orientation of the target object;

[0022] Determining the vertex coordinates of each vertex corresponding to the collision area in the detection coordinate system;

[0023] Determining the positional relationship between the environmental object and the collision area according to the position coordinates of the environmental object and the vertex coordinates of each of the area vertices.

[0024] Further, determining the positional relationship between the environmental object and the collision region according to the position coordinates of the environmental object and the vertex coordinates of each vertex includes:

[0025] Determining at least one detection vector group according to the position coordinates of the environmental object and the vertex coordinates of each region vertex, where the detection vector group includes a first detection vector formed from the environmental object to the region vertex and a second detection vector formed between two adjacent region vertices;

[0026] Determining the positional relationship between the target object and the collision region according to the directions of the first detection vector and the second detection vector in each detection vector group.

[0027] Further, determining the positional relationship between the target object and the collision region according to the directions of the first detection vector and the second detection vector in each detection vector group includes:

[0028] In response to the directions of the first detection vector and the second detection vector in each detection vector group being consistent, determining that there is an overlap between the environmental object and the collision region;

[0029] In response to the directions of the first detection vector and the second detection vector in at least one detection vector group being inconsistent, determining that there is no overlap between the environmental object and the collision region.

[0030] Further, determining the collision detection result between the target object and the environmental object according to the positional relationship includes:

[0031] In response to the positional relationship indicating that there is an overlap between the environmental object and the collision region, determining that a collision occurs between the target object and the environmental object;

[0032] In response to the positional relationship indicating that there is no overlap between the environmental object and the collision region, determining that no collision occurs between the target object and the environmental object.

[0033] Further, the method further includes:

[0034] Obtaining at least one predicted trajectory of the target object, where the predicted trajectory includes at least one environmental object.

[0035] Further, the method further includes:

[0036] Evaluating the predicted trajectory according to the collision detection results between the target object and each environmental object.

[0037] In a second aspect, an embodiment of the present invention aims to provide a collision detection device, and the device includes:

[0038] An information acquisition unit, configured to acquire the position information of the target object and the environmental object;

[0039] A shape representation unit, configured to perform shape representation on the target object and the environmental object according to the position information, and determine a first shape corresponding to the target object and a second shape corresponding to the environmental object;

[0040] A region determination unit, configured to move the second shape along the boundary of the first shape to determine the collision region of the environmental object relative to the target object;

[0041] A position detection unit, configured to determine the position relationship between the environmental object and the collision region according to the position information of the environmental object, where the position relationship is used to characterize the overlapping situation between the environmental object and the collision region; and determine the collision detection result between the target object and the environmental object according to the position relationship.

[0042] In a third aspect, an embodiment of the present invention aims at an electronic device, including a memory and a processor, where the memory is configured to store one or more computer program instructions, and wherein the one or more computer program instructions are executed by the processor to implement the method described in any one of the above.

[0043] In a fourth aspect, an embodiment of the present invention aims to provide a computer-readable storage medium, where a computer program is stored in the computer-readable storage medium, and when the computer program is executed by a processor, the method steps described in any one of the above are implemented.

[0044] The technical solution of the embodiment of the present invention acquires the position information of the target object and the environmental object, performs shape representation on the target object and the environmental object according to the position information, and determines a first shape corresponding to the target object and a second shape corresponding to the environmental object; moves the second shape along the boundary of the first shape to determine the collision region of the environmental object relative to the target object; determines the position relationship between the environmental object and the collision region according to the position information of the environmental object; and determines the collision detection result between the target object and the environmental object according to the position relationship. Thus, in this embodiment, the position area with a collision risk around the target object can be objectively expressed through the collision region, providing a reliable basis for determining the collision detection result according to the position relationship between the environmental object and the collision region; and by determining the collision detection result according to the position relationship between the environmental object and the collision region, on the basis of ensuring the accuracy of the collision result, the complex calculation amount can be reduced, thereby improving the determination efficiency and accuracy of the fault detection result. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] Through the following description of the embodiments of the present invention with reference to the accompanying drawings, the above and other objects, features and advantages of the present invention will become more clear. In the drawings:

[0046] Figure 1 is a flowchart of the collision detection method according to an embodiment of the present invention;

[0047] Figure 2 is a flowchart of another collision detection method according to an embodiment of the present invention;

[0048] Figure 3 is a schematic diagram of the shape representation according to an embodiment of the present invention;

[0049] Figure 4 is a schematic diagram of the first shape and the second shape according to an embodiment of the present invention;

[0050] Figure 5 is a flowchart of determining the collision area according to an embodiment of the present invention;

[0051] Figure 6 is a schematic diagram of circumscribing the second shape with the first shape according to an embodiment of the present invention;

[0052] Figure 7 is a schematic diagram of moving the second shape along the boundary of the first shape according to an embodiment of the present invention;

[0053] Figure 8 is a schematic diagram of moving the second shape along the boundary of the second shape according to an embodiment of the present invention;

[0054] Figure 9 is a schematic diagram of the collision area according to an embodiment of the present invention;

[0055] Figure 10 is a schematic diagram of the detection coordinate system according to an embodiment of the present invention;

[0056] Figure 11 is a schematic diagram of determining the vertex coordinates of the area vertices according to an embodiment of the present invention;

[0057] Figure 12 is a schematic diagram of the vector from the area vertex to the circumscribed vertex according to an embodiment of the present invention;

[0058] Figure 13 is a schematic diagram of the collision detection device according to an embodiment of the present invention;

[0059] Figure 14 is a schematic diagram of the electronic device according to an embodiment of the present invention. Detailed implementation manners

[0060] The present application will be described based on embodiments, but the present application is not limited to these embodiments. In the following detailed description of the present application, some specific details are described in detail. Those skilled in the art can fully understand the present application without the description of these details. In order to avoid obscuring the essence of the present application, well-known methods, processes, procedures, components, and circuits are not described in detail.

[0061] In addition, those of ordinary skill in the art should understand that the accompanying drawings provided herein are for illustrative purposes only and are not necessarily drawn to scale.

[0062] Unless the context clearly requires otherwise, the words such as "including" and "comprising" in the entire application document should be construed in an inclusive sense rather than an exclusive or exhaustive sense; that is, it is the meaning of "including but not limited to".

[0063] In the description of the present application, it should be understood that the terms "first", "second", etc. are only used for descriptive purposes and cannot be construed as indicating or implying relative importance. In addition, in the description of the present application, unless otherwise specified, the meaning of "a plurality" is two or more.

[0064] For the solutions described in this specification and the embodiments, if they involve personal information processing, they will all be processed on the premise of having a legal basis (such as obtaining the consent of the personal information subject, or being necessary for performing a contract, etc.), and will only be processed within the specified or agreed scope. If the user refuses to process personal information other than the necessary information required for the basic functions, it will not affect the user's use of the basic functions.

[0065] Through collision detection, the risk of collision between objects can be predicted, which helps the objects correct their movement trajectories in time and improves the safety of movement. In view of this, the embodiments of the present invention aim to provide a fault detection method to improve the determination efficiency and accuracy of fault detection results.

[0066] Figure 1 is a flowchart of the collision detection method according to the embodiments of the present invention. As Figure 1 shown, the collision detection method in this embodiment includes the following steps.

[0067] In step S110, obtain the position information of the target object and the environmental object.

[0068] In this embodiment, the target object is the object to be detected, and the environmental object is the risk object around the target object that has a collision risk with the target object. For example, the target object can be an autonomous vehicle, and the environmental objects are other traffic participants on the road other than the target object. The position information includes position coordinates and position direction. The position coordinates are used to represent the location of the target object or the environmental object, and the position direction is used to represent the moving direction of the target object or the environmental object.

[0069] Optionally, for the convenience of subsequent processing, the position coordinates in this embodiment can be the geographical coordinates of the central position (i.e., the center point) on the physical form of the object. That is, the position coordinates of the target object and the environmental object are the geographical coordinates of the center point of the target object and the physical coordinates of the center point of the environmental object, respectively. The position directions of the target object and the environmental object can be represented by an angle value. Thus, by obtaining the position information of the target object and the environmental object, it is convenient to timely and intuitively understand the locations and moving directions of the target object and the environmental object, and then it is convenient to perform fault detection based on the locations and moving directions of the target object and the environmental object.

[0070] In step S120, according to the position information, the target object and the environmental object are represented in terms of shape, and the first shape corresponding to the target object and the second shape corresponding to the environmental object are determined.

[0071] In this embodiment, shape representation is a process of geometrically representing an object. To make the result of shape representation closer to the real object, in this embodiment, the method of using a preset shape to circumscribe the outer contour of the target object or the environmental object is adopted to represent the target object and the environmental object in terms of shape, and the first shape corresponding to the target object and the second shape corresponding to the environmental object are determined.

[0072] Optionally, the preset shape in this embodiment can be a rectangle, a circle, an ellipse, a polygon, etc. The target object and the environmental object can be represented in terms of shape using the same or different shapes. Preferably, in this embodiment, the target object and the environmental object are represented in terms of shape by a rectangle. Compared with other shapes, the rectangle shape representation has better simplicity, stability, robustness, and generality, making it more convenient for a computer to describe, identify, and calculate during subsequent collision detection based on shape representation, maintaining good stability under noise and interference conditions, and adapting to different scales and rotational changes, ensuring the processing efficiency and the accuracy of the processing result.

[0073] Further, to flexibly determine the collision detection range and facilitate early warnings of collisions at different levels, in this embodiment, when representing the shape of the external contour of the target object or the environmental object by the circumscribed shape of the preset shape, taking the shape representing the first shape of the target object as an example, in an alternative implementation, the first shape is a preset shape circumscribed to the target object. The contour of the first shape circumscribes the contour of the target object, the position and orientation of the first shape correspond to those of the target object, and the position coordinates of the first shape correspond to those of the target object. In another alternative implementation, the first shape is a shape obtained by overall enlarging the preset shape outward on the basis of the preset shape circumscribed to the target object. There is a certain distance between the contour of the first shape and the preset shape circumscribing the contour of the target object, the position and orientation of the first shape correspond to those of the target object, and the position coordinates of the first shape correspond to those of the target object. At the same time, the second shape in this embodiment can be set as a preset shape circumscribing the contour of the environmental object or a shape of the same type as the preset shape formed by expanding outward on the basis of the preset shape circumscribing the contour of the environmental object in the same manner as the first shape.

[0074] In step S130, move the second shape along the boundary of the first shape to determine the collision area of the environmental object relative to the target object.

[0075] In this embodiment, after determining the first shape corresponding to the target object and the second shape corresponding to the environmental object, a collision area of the environmental object relative to the target object is constructed by the first shape and the second shape, and the collision area is the position area where there is a collision risk. Specifically, in this embodiment, the collision area of the environmental object relative to the target object is determined by moving the second shape along the boundary of the first shape, so that the collision area can comprehensively cover the collision risk area between the target object and the environmental object in the current position and orientation, and accurately judge the collision risk between the target object and the environmental object based on the collision area, thereby improving the accuracy of collision detection.

[0076] It should be noted that in this embodiment, the position coordinates and the position and orientation of the first shape when determining the collision area are the same as those of the target object; while the position and orientation of the second shape when determining the collision area are the same as those of the environmental object, but the position coordinates are only the same as those of the environmental object at the beginning of the first movement, and the position coordinates after each movement will change with the movement process. Thus, by moving the second shape along the boundary of the first shape, the collision area of the environmental object relative to the target object can be determined according to the position where at least one second shape is located after the movement.

[0077] Further, in this embodiment, each second shape during the movement along the boundary of the first shape is defined as a circumscribed shape. For a second shape that is not circumscribed about the first shape before the movement starts, the second shape is first moved along its own position direction to be circumscribed about the first shape, and then the second shape is continuously moved along the boundary of the first shape to determine multiple circumscribed shapes, and the collision area is determined according to the area enclosed by the multiple circumscribed shapes. Additionally, to further improve the collision detection efficiency, each position coordinate in this embodiment is the center point of the corresponding object or geometric shape, so the determined collision area is the area enclosed by the center points of the multiple circumscribed shapes, that is, the vertices of the geometric shape corresponding to the collision area are the center points of each circumscribed shape.

[0078] In step S140, the positional relationship between the environmental object and the collision area is determined according to the position information of the environmental object, and the positional relationship is used to characterize the overlapping situation between the environmental object and the collision area.

[0079] In this embodiment, after determining the collision area of the environmental object relative to the target object, the positional relationship between the environmental object and the collision area is determined according to the position information of the environmental object.

[0080] Optionally, to facilitate quickly determining the positional relationship between the environmental object and the collision area, in this embodiment, the positional relationship between the environmental object and the collision area is determined according to the relative positional relationship between the position where the environmental object is located and the positions of the vertices of the collision area, which can convert the collision detection between the environmental object and the target object from the judgment of area overlap between shapes to the judgment of relative positional relationship between points, reduce the computational complexity and time consumption of collision detection, and improve the collision detection efficiency.

[0081] Further, the positional relationship in this embodiment includes that the environmental object overlaps with the collision area and that the environmental object does not overlap with the collision area. When the position coordinates of the environmental object fall within the collision area or on the boundary of the collision area, it is determined that the environmental object overlaps with the collision area. When the position coordinates of the environmental object fall outside the collision area, it is determined that the environmental object does not overlap with the collision area.

[0082] In step S150, the collision detection result between the target object and the environmental object is determined according to the positional relationship.

[0083] In this embodiment, the collision detection is determined according to whether the environmental object overlaps with the collision area characterized by the positional relationship between the environmental object and the collision area. Specifically, in this embodiment, in response to the positional relationship indicating that the environmental object overlaps with the collision area, it is determined that the target object collides with the environmental object; in response to the positional relationship indicating that the environmental object does not overlap with the collision area, it is determined that the target object does not collide with the environmental object.

[0084] The technical solution of the embodiment of the present invention represents the shapes of the target object and the environmental object by obtaining the position information of the target object and the environmental object, and determines the collision area of the environmental object relative to the target object by moving the second shape corresponding to the environmental object along the first shape corresponding to the target object, so that the collision area can comprehensively cover and objectively and intuitively express the collision risk area between the target object and the environmental object, providing a reliable basis for accurately judging the collision risk between the target object and the environmental object based on the collision area subsequently, and further improving the accuracy of collision detection; moreover, by determining the positional relationship between the environmental object and the collision area according to the position information of the environmental object, and determining the collision detection result between the target object and the environmental object according to the positional relationship, the collision detection between the environmental object and the target object can be converted from the area overlap judgment between shapes to the relative positional relationship judgment between points, reducing the computational complexity and computational amount of collision detection while ensuring the accuracy of the collision result, thereby reducing the time consumption and improving the collision detection efficiency.

[0085] For ease of understanding, in this embodiment, the collision detection method will be described in detail by taking the collision detection in the process of trajectory prediction of an autonomous driving vehicle as an example. It should be understood, however, that the collision detection method in this embodiment is also applicable to other objects (such as intelligent mobile shelves, express sorting robots, food delivery robots, etc.) or scenarios (such as goods transportation, express sorting, food delivery, etc.) that require collision risk detection, and no limitation is imposed here.

[0086] In the autonomous driving scenario, trajectory prediction and collision detection are two closely related links. Trajectory prediction analyzes the states and behaviors of the autonomous driving vehicle and other traffic participants to predict the future positions and movement trajectories of the autonomous driving vehicle and other traffic participants, helping the autonomous driving system better plan the driving path of the vehicle and avoid potential collision risks. Collision detection, on the other hand, determines whether a collision will occur by analyzing the positions and speeds of the vehicle and obstacles, so as to facilitate the autonomous driving system to take measures such as adjusting the trajectory to avoid risks when a collision risk is detected. Therefore, by performing collision detection in trajectory prediction, the predicted movement trajectory can be evaluated to guide the safe driving of the autonomous driving vehicle.

[0087] Figure 2 is a flowchart of another collision detection method according to an embodiment of the present invention. As Figure 2 shown, the collision detection method in this embodiment includes the following steps.

[0088] In step S210, at least one predicted trajectory of the target object is obtained.

[0089] In this embodiment, the target object is an autonomous vehicle, and the environmental object is an object that appears in the predicted trajectory of the autonomous vehicle, such as other traffic participants other than the target object. Among them, the predicted trajectory is a moving trajectory in a future period of time predicted based on the historical information and current information of the target object or the environmental object, and the moving trajectory can be represented by a set of trajectory points in a time series. For example, assuming that the prediction duration is 9 seconds and each 0.1 second is a sampling moment, a moving trajectory includes 91 trajectory points and can be expressed as traj = {s0, s1,..., si,..., s90}, where the position point si includes five-dimensional data, namely t, (X, Y), θ, l, w. t represents the moment; X and Y are the position coordinates of the trajectory point, indicating the position of the object in the global coordinate system at this moment; θ represents the position direction of the object at this moment; l and w respectively represent the length and width of the object when its geometric shape is represented by a rectangle.

[0090] In this embodiment, when evaluating the predicted trajectory of an autonomous vehicle through collision detection, considering that there will be multiple predicted trajectories in the trajectory prediction of an autonomous vehicle and there will be multiple relevant environmental objects on one predicted trajectory, at least one predicted trajectory of the target object will be obtained in this embodiment, and the predicted trajectory includes at least one environmental object. Further, when evaluating each predicted trajectory, for each predicted trajectory, first, the collision detection results between each environmental object and the target object at a moment are respectively determined, the fault detection result corresponding to this moment is determined according to the collision detection results between each environmental object and the target object at this moment, then the fault detection results at other moments in this predicted trajectory are respectively determined by the same method, and the evaluation result of this predicted trajectory is determined according to the fault detection results at each moment in this predicted trajectory.

[0091] In step S220, the position information of the target object and the environmental object is obtained.

[0092] In this embodiment, taking the collision detection between one environmental object and the target object at a moment in one predicted trajectory of the target object as an example, the collision detection of each environmental object at each trajectory and each moment can be performed by the same method, and the corresponding collision detection results can be determined.

[0093] Optionally, since the trajectory points in the predicted trajectory have definite time attributes and position attributes, in this embodiment, the environmental objects existing in the scene can be screened by the appearance time and appearance position of the trajectory points to determine the environmental objects corresponding to each predicted trajectory of the target object. Then, collision detection is respectively performed on each environmental object and the target object in each predicted trajectory, and the predicted trajectories where each environmental object is located are evaluated based on the collision detection results of each environmental object and the target object. It should be understood that the purpose of screening in this embodiment is to preferentially perform collision detection on environmental objects with earlier appearance times and closer distances to the target object, rather than ignoring some environmental objects and not performing collision detection on them.

[0094] In step S230, according to the position information, the shapes of the target object and the environmental object are represented to determine the first shape corresponding to the target object and the second shape corresponding to the environmental object.

[0095] Figure 3 is a schematic diagram of the shape representation of an embodiment of the present invention. As Figure 3 shown, in this embodiment, taking the shape representation of the target object V as an example, when representing the shape of the target object V, first determine the outer contour of the target object V, and then determine a rectangle A circumscribing the target object V along this outer contour, and define this circumscribing rectangle A as the shape representation of the target object V, that is, the first shape A. The environmental object can adopt the same shape representation method as the target object.

[0096] Figure 4 is a schematic diagram of the first shape and the second shape of an embodiment of the present invention. Combining Figure 3 and Figure 4 shown, the first shape A in this embodiment is the shape representation of the target object V at the moment t of the mth modality (i.e., the mth predicted trajectory), and the second shape B is the shape representation of the environmental object N at the moment t i . The contour of the first shape A circumscribes the contour of the target object V, and the position direction is the position direction θ of the target object V i , and the position coordinates of the center point A0 are the position coordinates (X A , Y A ) of the center point of the target object V. The length is the maximum length l in the length direction of the target object V A , and the width is the maximum width w in the width direction of the target object V A . The contour of the second shape B circumscribes the contour of the environmental object N, and the position direction is the position direction θ of the environmental object N A , and the position coordinates of the center point B0 are the position coordinates (X B , Y B ) of the center point of the environmental object N. The length is the maximum length l in the length direction of the environmental object N B , and the width is the maximum width w in the width direction of the environmental object N B, with a width of the maximum width w in the width direction of the environmental object N B .

[0097] In step S240, move the second shape along the boundary of the first shape to determine the collision area of the environmental object relative to the target object.

[0098] Figure 5 is the flowchart of determining the collision area in an embodiment of the present invention. As Figure 5 shown, in this embodiment, the collision area of the environmental object relative to the target object is determined through the following steps.

[0099] In step S510, move the second shape to be circumscribed with the first shape according to the position direction of the second shape.

[0100] In this embodiment, considering that there are various relative position relationships between the second shape and the first shape, the second shape will be moved according to the position direction of the second shape and / or the direction perpendicular to the position direction of the second shape and / or the direction opposite to the position direction of the second shape, so that the moved second shape is circumscribed with the first shape.

[0101] Optionally, assume that the relative position relationship between the second shape B and the first shape A is as Figure 6 shown. Then, when moving the second shape B to be circumscribed with the first shape A, the second shape B can be first moved along the position direction of the second shape B according to the path ① in the figure, and then moved along the direction perpendicular to the position direction of the second shape B according to the path ② in the figure to be tangent to the vertex of the first shape A.

[0102] In step S520, starting from the vertex of the first shape, translate the second shape alternately along the boundaries of the first shape and the second shape to determine at least one circumscribed shape.

[0103] In this embodiment, after moving the second shape to be circumscribed with the first shape, starting from the vertex where the second shape is circumscribed with the first shape, continue to translate the second shape alternately along the boundaries of the first shape and the second shape, and determine at least one circumscribed shape. Optionally, to ensure the complete coverage of the generated collision area, the operation of alternately translating along the boundaries of the first shape and the second shape in this embodiment includes moving along the long side of the first shape, moving along the short side of the first shape, moving along the long side of the second shape, and moving along the short side of the second shape.

[0104] It should be understood that when the position directions of the first shape and the second shape are neither parallel nor perpendicular, the first shape and the second shape in the translation operation are circumscribed only through one point. At this time, each circumscribed shape and the first shape have only one circumscribed point; while when the position directions of the first shape and the second shape are parallel or perpendicular, the first shape and the second shape are circumscribed through one side, and correspondingly, each circumscribed shape and the first shape have only one circumscribed side.

[0105] Further, in this embodiment, as Figure 7 and Figure 8 shown, the processes and effects of translation along the first shape boundary and along the second shape boundary are respectively shown. At the same time, Figure 9 shows the formation process of the circumscribed shape corresponding to each translation. In this embodiment, the first second shape circumscribing the first shape is used as the circumscribed shape B1, and then the circumscribed shape B1 is translated along the short side of the second shape in a direction perpendicular to the position direction of the second shape to obtain the circumscribed shape B2, the circumscribed shape B2 is translated along the short side of the first shape in a direction perpendicular to the position direction of the first shape to obtain the circumscribed shape B3, the circumscribed shape B3 is translated along the long side of the second shape in a direction opposite to the position direction of the second shape to obtain the circumscribed shape B4, the circumscribed shape B4 is translated along the long side of the first shape in a direction opposite to the position direction of the first shape to obtain the circumscribed shape B5, the circumscribed shape B5 is translated along the short side of the second shape in a direction perpendicular to the position direction of the second shape to obtain the circumscribed shape B6, the circumscribed shape B6 is translated along the short side of the first shape in a direction perpendicular to the position direction of the first shape to obtain the circumscribed shape B7, the circumscribed shape B7 is translated along the long side of the second shape in the position direction of the second shape to obtain the circumscribed shape B8, and the direction from the circumscribed shape B8 to the circumscribed shape B1 is consistent with the position direction of the first shape.

[0106] In step S530, the collision region of the environmental object relative to the target object is determined according to each circumscribed shape.

[0107] In this embodiment, as Figure 9 shown, after determining each circumscribed shape, first, the center points of each circumscribed shape are determined as the vertices of the collision region, and then the vertices are connected to determine the enclosed region formed after the connection as the collision region of the environmental object relative to the target object. At the same time, it should be understood that when the position directions of the first shape and the second shape are neither parallel nor perpendicular, the shape of the collision region of the environmental object relative to the target object is a convex octagon. And when the position directions of the first shape and the second shape are parallel or perpendicular, the shape of the collision region of the environmental object relative to the target object will degenerate from an octagon to a rectangle.

[0108] In step S250, the positional relationship between the environmental object and the collision region is determined according to the position information of the environmental object, and the positional relationship is used to characterize the overlapping situation between the environmental object and the collision region.

[0109] In this embodiment, when determining the positional relationship between the environmental object and the collision area according to the position information of the environmental object, first construct a detection coordinate system based on the position coordinates and position direction of the target object; then determine the vertex coordinates of each area vertex corresponding to the collision area in the detection coordinate system; finally, determine the positional relationship between the environmental object and the collision area according to the position coordinates of the environmental object and the vertex coordinates of each area vertex.

[0110] Optionally, to facilitate determining the positional relationship between the environmental object and the collision area and improve the collision detection efficiency, as Figure 10 shown, the origin of the detection coordinate system in this embodiment is the position where the center point A0 of the target object is located. The detection coordinate system includes a first direction x and a second direction y. The first direction x is consistent with the position direction θ A of the target object, and the second direction y is perpendicular to the position direction θ A of the target object.

[0111] After establishing the detection coordinate system, combined with Figure 9 and Figure 10 , since each area vertex B n (n = 1, 2,..., 8) of the collision area is the center point of the circumscribed shape Bn of each vertex of the first shape A, the relative positional relationship between each area vertex B n and the vertex A n (n = 1, 2, 3, 4) (i.e., the circumscribed vertex) of the corresponding circumscribed shape Bn and the first shape A is determined, and this relative positional relationship can be represented by the vector from the circumscribed vertex A n to the area vertex B n . Therefore, when determining the vertex coordinates of each area vertex B n corresponding to the collision area in the detection coordinate system, in this embodiment, the vector coordinates of the vector A n from the corresponding circumscribed vertex A n to the area vertex B n will be determined first, and then the vertex coordinates of each area vertex B n will be determined according to the vector A n B n and the vector coordinates of the vector A0A n from the coordinate system origin A0 (i.e., the center point of the first shape) to the circumscribed vertex A n . n n

[0112] For example, as Figure 10As shown, taking the determination of the vertex coordinates of the regional vertex B1 as an example, the regional vertex B1 is the center point of the circumscribed shape B1, and the circumscribed shape B1 is circumscribed with the vertex A1 of the first shape A. When determining the vertex coordinates of the regional vertex B1 in the detection coordinate system, the vector coordinates of the vector A1B1 from the circumscribed vertex A1 to the regional vertex B1 corresponding to the regional vertex B1 can be determined first, and then the vector coordinates of the vector A0B1 from the origin A0 of the coordinate system to the regional vertex B1 can be determined according to the vector coordinates of the vector A1B1 and the vector coordinates of the vector A0A1 from the origin A0 of the coordinate system to the circumscribed vertex A1. Since the vector A0B1 is the sum of the vector A0A1 and the vector A1B1, the vector coordinates of the vector A0B1 are the sum of the vector coordinates of the vector A0A1 and the vector coordinates of the vector A1B1.

[0113] Optionally, in this embodiment, the coordinate system origin A0 to the circumscribed vertex A n The vector A0A n The vector coordinates of the corresponding circumscribed vertex A n For example, the coordinates of vector A0A1 are consistent with the coordinates of circumscribed vertex A1 in the detection coordinate system, that is, (l A / 2,w A / 2).

[0114] In determining the vertices B of each region n The corresponding circumscribed vertex A n To region vertex B n The vector A n B n When Figure 11 As shown, in this embodiment, a first offset vector A is introduced n B n1 and the second offset vector A n B n2 , by the first offset vector A n B n1 and the second offset vector A n B n2 Perform vector calculations to determine vector A n B n Among them, the first offset vector A n B n1 Used to represent the vertex B of the region n Relative to the circumscribed vertex A n The offset along the position direction of the circumscribed shape Bn, the second offset vector A n B n2 Used to represent the vertex B of the region n Relative to the circumscribed vertex A n The offset in the direction perpendicular to the position direction of the circumscribed shape Bn. And, the first offset vector A n B n1and the second offset vector A n B n2 The corresponding offsets both include offset values in the x and y directions. The offset corresponding to the first offset vector can be expressed as The offset corresponding to the second offset vector can be expressed as θ is the position direction θ of the circumscribed shape B and the position direction θ of the first shape A The angular difference, that is, θ = θ B - θ A . Thus, according to the value and direction of the offset, the first offset vector A n B n1 and the second offset vector A n B n2 can be determined. Then, by summing the vector coordinates of the first offset vector A n B n1 and the second offset vector A n B n2 the vector coordinates of the circumscribed vertex A n corresponding to each region vertex B n to the region vertex B n vector A n B n can be determined. Finally, by summing the vector coordinates of the vector A0A n from the coordinate system origin A0 to the circumscribed vertex A n and the vector coordinates of the vector A n from the circumscribed vertex A n corresponding to each region vertex B n to the region vertex B n B n the vertex coordinates of each region vertex B n are determined.

[0115] Furthermore, with reference to Figure 12 and in combination with the above determination method, the vertex coordinates of each region vertex B n in the collision region of this embodiment are respectively expressed by the following formulas:

[0116]

[0117]

[0118]

[0119]

[0120]

[0121]

[0122]

[0123]

[0124] Optionally, in this embodiment, after determining the vertex coordinates of the vertices of each region, the positional relationship between the environmental object and the collision region is determined according to the position coordinates of the environmental object and the vertex coordinates of each vertex. When determining the positional relationship between the environmental object and the collision region, at least one detection vector group is determined according to the position coordinates of the environmental object and the vertex coordinates of each region vertex. The detection vector group includes a first detection vector formed from the environmental object to the region vertex and a second detection vector formed between two adjacent region vertices; then, the positional relationship between the target object and the collision region is determined according to the directions of the first detection vector and the second detection vector in each detection vector group. Further, the fault detection method in this embodiment determines that there is an overlap between the environmental object and the collision region in response to the directions of the first detection vector and the second detection vector in each detection vector group being the same; or determines that there is no overlap between the environmental object and the collision region in response to the directions of the first detection vector and the second detection vector in at least one detection vector group being different.

[0125] Specifically, in this embodiment, the detection vector group is denoted as B0B n B n-1 , and each detection vector group includes a first vector B0B n and a second vector B n B n-1 , where n = 1, 2,..., 8. It should be noted that when n is 1, the corresponding detection vector group B0B1B8 includes the first vector B0B1 and the second vector B1B8. Thus, after determining each detection vector group and the first vector and the second vector in each detection vector group, the cross product calculation is respectively performed on the first vector and the second vector in each detection vector group. When the cross product calculation result is positive, it indicates that the directions of the first vector and the second vector in the current detection vector group are the same; when the cross product calculation result is negative, it indicates that the directions of the first vector and the second vector in the current detection vector group are different. Further, when the directions of the first vector and the second vector in each detection vector group are all the same, it indicates that the center point B0 of the environmental object falls into the collision region, and at this time, there is an overlap between the environmental object and the collision region; or, when the directions of the first vector and the second vector in at least one detection vector group are different, it indicates that the center point B0 of the environmental object does not fall into the collision region, and at this time, there is no overlap between the environmental object and the collision region.

[0126] In step S260, the collision detection result between the target object and the environmental object is determined according to the positional relationship.

[0127] In this embodiment, when the positional relationship indicates that the environmental object overlaps with the collision area, it is determined that a collision will occur between the target object and the environmental object; when the positional relationship indicates that the environmental object does not overlap with the collision area, it is determined that a collision will not occur between the target object and the environmental object.

[0128] In step S270, the predicted trajectory is evaluated according to the collision detection results of the target object and each environmental object.

[0129] In this embodiment, after determining the collision detection result of one environmental object and the target object based on the above steps, the same method can be used to determine the collision detection results of other environmental objects and the target object in the same predicted trajectory. Then, the predicted trajectory is evaluated according to the collision detection results of the target object and each environmental object. For example, the more collision detection results indicating a collision occur in a predicted trajectory, the more times the predicted trajectory needs to be adjusted or the less available it is, and the worse its quality; on the contrary, the fewer collision detection results indicating a collision in the predicted trajectory, the better the quality of the predicted trajectory, so as to ensure the safe and reliable operation of the autonomous driving vehicle.

[0130] The technical solution of this embodiment can obtain at least one predicted trajectory of the target object and the environmental objects related to the predicted trajectory, and can determine the collision area of the environmental object relative to the target object based on the positional relationship between the environmental object and the target object at different times, and determine the collision detection result of the environmental object and the target object at the corresponding time based on the positional relationship between the environmental object and the collision area, improving the accuracy and efficiency of collision detection. At the same time, since the collision detection process in this embodiment involves data in five dimensions: the time, position coordinates, position direction of the target object and the environmental object, and the lengths and widths of the corresponding shape representations, the above processes of determining the collision area and the collision detection result can be implemented based on the matrix parallel computing framework of the graphics card, avoiding the use of multiple loops by the graphics card for collision detection, improving the processing efficiency of the graphics card, making the collision detection more convenient, taking less time for collision detection, and further improving the collision detection efficiency and the trajectory evaluation efficiency, thereby providing support for the safe and reliable operation of the autonomous driving vehicle.

[0131] Figure 13 is a schematic diagram of the collision detection device according to an embodiment of the present invention. As Figure 13As shown in the figure, the collision detection device in this embodiment includes an information acquisition unit 1, a shape representation unit 2, a region determination unit 3, and a position detection unit 4. Among them, the information acquisition unit 1 is used to acquire the position information of the target object and the environmental object. The shape representation unit 2 is used to represent the shapes of the target object and the environmental object according to the position information, and determine the first shape corresponding to the target object and the second shape corresponding to the environmental object. The region determination unit 3 is used to move the second shape along the boundary of the first shape to determine the collision region of the environmental object relative to the target object. The position detection unit 4 is used to determine the position relationship between the environmental object and the collision region according to the position information of the environmental object, and determine the collision detection result between the target object and the environmental object according to the position relationship. The position relationship is used to characterize the overlapping situation between the environmental object and the collision region.

[0132] Optionally, the position relationship in this embodiment includes position coordinates and position direction. The position coordinates are used to characterize the location of the target object or the environmental object. The first shape is a rectangle, the contour of the first shape circumscribes the contour of the target object, the position direction of the first shape corresponds to the position direction of the target object, and the position coordinates of the first shape correspond to the position coordinates of the target object.

[0133] Optionally, when determining the collision region, the region determination unit 3 in this embodiment is further used to move the second shape to be circumscribed with the first shape according to the position direction of the second shape; starting from the vertex of the first shape, translate the second shape alternately along the boundaries of the first shape and the second shape to determine at least one circumscribed shape; determine the collision region of the environmental object relative to the target object according to each circumscribed shape. Further, when determining the collision region according to each circumscribed shape, the region determination unit 3 is specifically used to determine the center point of each circumscribed shape as the vertex of the collision region; connect the vertices to determine the closed region formed after the connection as the collision region of the environmental object relative to the target object.

[0134] Optionally, the position detection unit 4 in this embodiment is further used to construct a detection coordinate system according to the position coordinates and position direction of the target object, and determine the vertex coordinates of each region vertex corresponding to the collision region in the detection coordinate system; determine the position relationship between the environmental object and the collision region according to the position coordinates of the environmental object and the vertex coordinates of each region vertex. Among them, the origin of the detection coordinate system is the location of the target object, the detection coordinate system includes a first direction and a second direction, the first direction is consistent with the position direction of the target object, and the second direction is perpendicular to the position direction of the target object.

[0135] Further, when determining the positional relationship in this embodiment, the position detection unit 4 is specifically configured to determine at least one detection vector group according to the position coordinates of the environmental object and the vertex coordinates of each region vertex. The detection vector group includes a first detection vector formed from the environmental object to the region vertex and a second detection vector formed between two adjacent region vertices; and determine the positional relationship between the target object and the collision region according to the directions of the first detection vector and the second detection vector in each detection vector group. In response to the directions of the first detection vector and the second detection vector in each detection vector group being consistent, it is determined that the environmental object and the collision region overlap; in response to the directions of the first detection vector and the second detection vector in at least one detection vector group being inconsistent, it is determined that the environmental object and the collision region do not overlap.

[0136] Optionally, the position detection unit 4 in this embodiment is further configured to determine that the target object collides with the environmental object in response to the positional relationship indicating that the environmental object and the collision region overlap; and determine that the target object does not collide with the environmental object in response to the positional relationship indicating that the environmental object and the collision region do not overlap.

[0137] Optionally, the fault detection device in this embodiment further includes a trajectory processing unit 5. The trajectory processing unit 5 is configured to obtain at least one predicted trajectory of the target object, where the predicted trajectory includes at least one environmental object; and evaluate the predicted trajectory according to the collision detection results between the target object and each environmental object.

[0138] Figure 14 is a schematic diagram of the electronic device according to an embodiment of the present invention. As Figure 14 shown, the electronic device is a general address query device, which includes a general computer hardware structure, and at least includes a processor 61 and a memory 62. The processor 61 and the memory 62 are connected through a bus 63. The memory 62 is adapted to store instructions or programs executable by the processor 61. The processor 61 may be an independent microprocessor or a set of one or more microprocessors. Thus, the processor 61 executes the instructions stored in the memory 62 to execute the method flow of the embodiment of the present invention as described above to implement data processing and control of other devices. The bus 63 connects the above-mentioned multiple components together and at the same time connects the above-mentioned components to a display controller 64, a display device, and an input / output (I / O) device 65. The input / output (I / O) device 65 may be a mouse, a keyboard, a modem, a network interface, a touch input device, a body sensing input device, a printer, and other devices well known in the art. Typically, the input / output (I / O) device 65 is connected to the system through an input / output (I / O) controller 66.

[0139] Those skilled in the art should understand that the embodiments of the present application can be provided as a method, a device (equipment), or a computer program product. Therefore, the present application can take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present application can be implemented as a computer program product on one or more computer-readable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that contain computer-usable program code.

[0140] The present application is described with reference to the flowcharts of methods, devices (equipment), and computer program products according to the embodiments of the present application. It should be understood that each process in the flowchart can be implemented by computer program instructions.

[0141] These computer program instructions can be stored in a computer-readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer-readable memory produce a manufactured article including an instruction device that implements the process Figure 1 specified functions in one or more of these processes.

[0142] These computer program instructions can also be provided to the processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device produce a device for implementing the Figure 1 specified functions in one or more of these processes.

[0143] Another embodiment of the present invention relates to a non-volatile storage medium for storing a computer-readable program, and the computer-readable program is used for a computer to execute some or all of the above method embodiments.

[0144] That is, those skilled in the art can understand that all or part of the steps in implementing the above method embodiments can be completed by specifying relevant hardware through a program. The program is stored in a storage medium and includes several instructions to enable a device (which can be a single-chip microcomputer, a chip, etc.) or a processor to execute all or part of the steps of the methods described in the embodiments of the present application. The aforementioned storage medium includes: USB flash drives, mobile hard disks, read-only memories (ROMs), random access memories (RAMs), magnetic disks, or optical discs, etc., which can store program codes.

[0145] The above are only the preferred embodiments of the present application and are not intended to limit the present application. For those skilled in the art, various modifications and changes can be made to the present application. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application shall be included within the protection scope of the present application.

Claims

1. A collision detection method, characterized in that, The method includes: Obtaining the position information of the target object and the environmental object; Performing shape representation on the target object and the environmental object according to the position information, and determining a first shape corresponding to the target object and a second shape corresponding to the environmental object; Moving the second shape along the boundary of the first shape to determine the collision area of the environmental object relative to the target object; Determining the positional relationship between the environmental object and the collision area according to the position information of the environmental object, where the positional relationship is used to characterize the overlapping situation between the environmental object and the collision area; Determining the collision detection result between the target object and the environmental object according to the positional relationship.

2. The method according to claim 1, characterized in that, The first shape and the second shape are rectangles. The moving the second shape along the boundary of the first shape to determine the collision area of the environmental object relative to the target object includes: Moving the second shape to be circumscribed with the first shape according to the position direction of the second shape; Starting from the vertex of the first shape, alternately translating the second shape along the boundaries of the first shape and the second shape to determine at least one circumscribed shape; Determining the collision area of the environmental object relative to the target object according to each circumscribed shape.

3. The method according to claim 2, characterized in that, The determining the collision area of the environmental object relative to the target object according to each circumscribed shape includes: Determining the center point of each circumscribed shape as the vertex of the collision area; Connecting the vertices to determine the closed area formed after the connection as the collision area of the environmental object relative to the target object.

4. The method according to claim 3, wherein The position information includes position coordinates and a position direction, and the position coordinates are used to characterize the location of the target object or the environmental object.

5. The method according to claim 4, characterized in that, The contour of the first shape is circumscribed with the contour of the target object, the position direction of the first shape corresponds to the position direction of the target object, and the position coordinates of the first shape correspond to the position coordinates of the target object.

6. The method according to claim 4, characterized in that The determining the positional relationship between the environmental object and the collision area according to the position information of the environmental object includes: Constructing a detection coordinate system according to the position coordinates and the position direction of the target object, where the origin of the detection coordinate system is the location of the target object, the detection coordinate system includes a first direction and a second direction, the first direction is consistent with the position direction of the target object, and the second direction is perpendicular to the position direction of the target object; Determining the vertex coordinates of each area vertex corresponding to the collision area in the detection coordinate system; Determining the positional relationship between the environmental object and the collision area according to the position coordinates of the environmental object and the vertex coordinates of each area vertex.

7. The method according to claim 6, wherein The determining the positional relationship between the environmental object and the collision area according to the position coordinates of the environmental object and the vertex coordinates of each vertex includes: Determining at least one detection vector group according to the position coordinates of the environmental object and the vertex coordinates of each area vertex, where the detection vector group includes a first detection vector formed from the environmental object to the area vertex and a second detection vector formed between two adjacent area vertices; Determine the positional relationship between the target object and the collision area according to the directions of the first detection vector and the second detection vector in each of the detection vector groups.

8. The method according to claim 7, wherein The determining the positional relationship between the target object and the collision area according to the directions of the first detection vector and the second detection vector in each of the detection vector groups includes: In response to the directions of the first detection vector and the second detection vector in each of the detection vector groups being the same, determine that there is an overlap between the environmental object and the collision area; In response to the directions of the first detection vector and the second detection vector in at least one of the detection vector groups being different, determine that there is no overlap between the environmental object and the collision area.

9. The method according to claim 1, characterized in that, The determining the collision detection result between the target object and the environmental object according to the positional relationship includes: In response to the positional relationship indicating that there is an overlap between the environmental object and the collision area, determine that a collision occurs between the target object and the environmental object; In response to the positional relationship indicating that there is no overlap between the environmental object and the collision area, determine that no collision occurs between the target object and the environmental object.

10. The method according to claim 1, characterized in that, The method further includes: Obtain at least one predicted trajectory of the target object, where the predicted trajectory includes at least one environmental object.

11. The method according to claim 10, wherein The method further includes: Evaluate the predicted trajectory according to the collision detection results between the target object and each of the environmental objects.

12. A collision detection device, characterized in that, The device includes: An information acquisition unit, configured to acquire the position information of the target object and the environmental object; A shape representation unit, configured to represent the shapes of the target object and the environmental object according to the position information, and determine the first shape corresponding to the target object and the second shape corresponding to the environmental object; A region determination unit, configured to move the second shape along the boundary of the first shape to determine the collision area of the environmental object relative to the target object; A position detection unit, configured to determine the positional relationship between the environmental object and the collision area according to the position information of the environmental object, where the positional relationship is used to characterize the overlap situation between the environmental object and the collision area; and determine the collision detection result between the target object and the environmental object according to the positional relationship.

13. An electronic device, comprising a memory and a processor, characterized in that, The memory is used to store one or more computer program instructions, where the one or more computer program instructions are executed by the processor to implement the method according to any one of claims 1-11.

14. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program, and when the computer program is executed by the processor, the method steps according to any one of claims 1-11 are implemented.

Images