Curb detection method and device, equipment and storage medium

Abstract

The invention relates to the technical field of automatic driving, and discloses a curb detection method, device and equipment and a storage medium, which are used for improving the accuracy of curb detection. The curb detection method comprises the steps that an image in the surrounding environment of a vehicle is acquired, and the image is captured by a camera on the vehicle and reflects the position and the shape of an object in the surrounding environment; simulating rays emitted from the center of the camera to the surrounding environment based on natural light reflection rays in the image, and recording intersection points of the rays and the object and attributes of the intersection points; judging whether adjacent target intersection points belong to the same road edge or not according to the intersection points and the attributes of the intersection points; and connecting the target intersection points belonging to the same curb to obtain a curb boundary.

Description

Technical Field

[0001] The present invention relates to the field of autonomous driving technology, and in particular to a curb detection method, device, equipment and storage medium. Background Art

[0002] The current curb detection method usually captures the surrounding environment image through the camera on the vehicle, and then uses edge detection technology to identify the outline of the curb based on the pixel information in the image, and finally obtains the curb boundary.

[0003] In the existing technology, since it relies on the edge features of the image for recognition, it is easily disturbed by factors such as lighting conditions, shadows, and noise, resulting in low accuracy of curb detection. Summary of the invention

[0004] The present invention provides a curb detection method, device, equipment and storage medium to solve the problem in the prior art that the curb detection has low accuracy due to reliance on edge features of an image for recognition.

[0005] The first aspect of the present invention provides a curb detection method, comprising: the curb detection method comprises: acquiring an image of the vehicle's surrounding environment, the image being captured by a camera on the vehicle and reflecting the position and shape of objects in the surrounding environment; simulating rays emitted from the center of the camera to the surrounding environment based on natural light reflected light in the image, and recording the intersections of the rays and the objects, as well as the attributes of each intersection; judging whether adjacent target intersections belong to the same curb based on each intersection and the attributes of each intersection; connecting the target intersections belonging to the same curb to obtain a curb boundary.

[0006] In a feasible implementation manner, based on the natural light reflected in the image, the rays emitted from the center of the camera to the surrounding environment are simulated, and the intersection points of the rays with the objects and the properties of each intersection point are recorded, including: analyzing the source, direction and intensity of the light according to the light distribution in the image; simulating the rays triggered from the center of the camera and emitted along the direction of the light according to the analysis results, and assigning a corresponding intensity value to each ray; performing collision detection between the rays and the objects in the image, and recording the corresponding intersection points and the properties of the intersection points.

[0007] In a feasible implementation, the collision detection between the ray and the object in the image includes: using image processing technology to identify the object from the image and extract the corresponding contour; calculating whether each ray intersects with the identified object contour; and determining the intersection position of the ray and the object contour through geometric calculation.

[0008] In a feasible implementation, the method of judging whether the intersections of adjacent rays belong to the same curb based on the intersections and the attributes of the intersections includes: filtering out target intersections belonging to the curb based on the attributes of each intersection; calculating the spatial distance and direction angle between adjacent target intersections; and judging whether the adjacent target intersections belong to the same curb based on the spatial distance and the direction angle.

[0009] In a feasible implementation, the calculation of the spatial distance and direction angle between adjacent target intersections includes: obtaining the coordinates of adjacent target intersections; calculating the straight-line distance between adjacent target intersections using a distance formula based on the coordinates of the adjacent target intersections to obtain the spatial distance between the adjacent target intersections; and obtaining the direction angle between the adjacent target intersections by calculating the angle between the connection of the adjacent target intersections and the positive direction of the horizontal axis.

[0010] In a feasible implementation, the determining whether adjacent target intersections belong to the same curb based on the spatial distance and the directional angle includes: setting thresholds for the spatial distance and the directional angle according to the continuity characteristics of the curb; comparing the spatial distance and the directional angle between adjacent target intersections with the set thresholds; and attributing adjacent target intersections whose spatial distance and directional angle are within the set thresholds to the same curb.

[0011] In a feasible implementation, before obtaining the reflected light of natural light in the vehicle's surrounding environment, it also includes: real-time detection of the lighting conditions of the vehicle's surrounding environment; and dynamically adjusting relevant parameters of the camera according to the lighting conditions to adapt to curb detection requirements under different lighting conditions.

[0012] The second aspect of the present invention provides a curb detection device, including: an acquisition module, used to acquire an image of the vehicle's surrounding environment, the image is captured by a camera on the vehicle, and reflects the position and shape of objects in the surrounding environment; a processing module, used to simulate rays emitted from the center of the camera to the surrounding environment based on natural light reflected light in the image, and record the intersection of the rays and the objects, as well as the attributes of each intersection; a judgment module, used to judge whether adjacent target intersections belong to the same curb based on each intersection and the attributes of each intersection; a connection module, used to connect target intersections belonging to the same curb to obtain a curb boundary.

[0013] In a feasible implementation, the processing module includes: an analysis unit, used to analyze the source, direction and intensity of light according to the light distribution in the image; a simulation unit, used to simulate, according to the analysis results, rays triggered from the center of the camera and emitted along the direction of the light, and assign a corresponding intensity value to each ray; a processing unit, used to perform collision detection between the rays and the objects in the image, and record the corresponding intersection points and the attributes of the intersection points.

[0014] In a feasible implementation, the processing unit is specifically used to: use image processing technology to identify objects from the image and extract corresponding contours; calculate whether each ray intersects with the identified object contour; and determine the intersection position of the ray and the object contour through geometric calculation.

[0015] In a feasible implementation, the judgment module includes: a screening unit, used to screen out target intersections belonging to the curb based on the attributes of each intersection; a calculation unit, used to calculate the spatial distance and direction angle between adjacent target intersections; and a judgment unit, used to judge whether adjacent target intersections belong to the same curb based on the spatial distance and the direction angle.

[0016] In a feasible implementation, the calculation unit is specifically used to: obtain the coordinates of adjacent target intersections; calculate the straight-line distance between adjacent target intersections using a distance formula based on the coordinates of adjacent target intersections to obtain the spatial distance between adjacent target intersections; and obtain the direction angle between adjacent target intersections by calculating the angle between the connection of adjacent target intersections and the positive direction of the horizontal axis.

[0017] In a feasible implementation, the judgment unit is specifically used to: set thresholds for spatial distance and direction angle according to the continuity characteristics of the curb; compare the spatial distance and direction angle between adjacent target intersections with the set thresholds; and assign adjacent target intersections whose spatial distance and direction angle are within the set thresholds to the same curb.

[0018] In a feasible implementation, the curb detection device further includes: an adjustment module for detecting the lighting conditions of the vehicle's surrounding environment in real time; and dynamically adjusting relevant parameters of the camera according to the lighting conditions to adapt to curb detection requirements under different lighting conditions.

[0019] A third aspect of the present invention provides a curb detection device, comprising: a memory and at least one processor, wherein the memory stores instructions; the at least one processor calls the instructions in the memory so that the curb detection device executes the above-mentioned curb detection method.

[0020] A fourth aspect of the present invention provides a computer-readable storage medium, wherein the computer-readable storage medium stores instructions, which, when executed on a computer, enable the computer to execute the above-mentioned curb detection method.

[0021] In the technical solution provided by the present invention, an image of the vehicle's surroundings is obtained, the image is captured by a camera on the vehicle, and reflects the position and shape of objects in the surroundings; based on the natural light reflected light in the image, the rays emitted from the center of the camera to the surroundings are simulated, and the intersections of the rays and the objects, as well as the attributes of each intersection, are recorded; based on each intersection and the attributes of each intersection, it is determined whether adjacent target intersections belong to the same curb; the target intersections belonging to the same curb are connected to obtain a curb boundary. In an embodiment of the present invention, by utilizing the image captured by the vehicle camera, combining the simulation of natural light reflected light with the intersection record, it is determined whether adjacent intersections belong to the same curb, and the intersections belonging to the same curb are connected to form a curb boundary, thereby improving the accuracy of curb detection and enhancing detection stability. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] Figure 1 A schematic diagram of an embodiment of a curb detection method in an embodiment of the present invention;

[0023] Figure 2 A schematic diagram of another embodiment of the curb detection method in an embodiment of the present invention;

[0024] Figure 3 A schematic diagram of an embodiment of a curb detection device in an embodiment of the present invention;

[0025] Figure 4 A schematic diagram of another embodiment of a curb detection device in an embodiment of the present invention;

[0026] Figure 5 Schematic diagram of an embodiment of a curb detection device in an embodiment of the present invention. DETAILED DESCRIPTION

[0027] The embodiments of the present invention provide a curb detection method, device, equipment and storage medium, which simulate natural light reflection and record intersection properties, determine and connect intersections belonging to a curb to obtain a curb boundary, thereby improving the accuracy of curb detection.

[0028] The terms "first", "second", "third", "fourth", etc. (if any) in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments described herein can be implemented in an order other than that illustrated or described herein. In addition, the terms "including" or "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0029] It is understandable that the execution subject of the present invention may be a curb detection device, or a terminal or a server, which is not limited here. The embodiment of the present invention is described by taking a server as the execution subject as an example.

[0030] For ease of understanding, the specific process of the embodiment of the present invention is described below. Figure 1 , an embodiment of the curb detection method in the embodiment of the present invention includes:

[0031] 101. Acquire an image of the vehicle's surroundings, where the image is captured by a camera on the vehicle and reflects the position and shape of objects in the surroundings;

[0032] The camera installed on the vehicle is used to capture high-definition images of its surroundings. These cameras are usually configured on the front, side or rear of the vehicle to ensure that the vehicle's driving path and its surrounding environment are fully covered. The images captured by the camera not only contain rich color and texture information, but also accurately reflect the position and shape of objects in the surrounding environment. The camera is usually equipped with a high-resolution sensor and image processing algorithm to ensure that clear and accurate image information can be captured even in complex and changing road environments. In addition, the camera may also have the function of automatically adjusting the focus, exposure and color balance to adapt to different lighting conditions and weather conditions.

[0033] 102. Based on the natural light reflection in the image, simulate the rays emitted from the center of the camera to the surrounding environment, and record the intersection points of the rays and the objects, as well as the properties of each intersection point;

[0034] Using image processing technology and optical principles, based on the light information formed by the reflection of natural light on the surface of the object in the image, a series of virtual rays emitted from the center of the camera to the surrounding environment are simulated. These rays are not real rays, but are calculated by the algorithm based on the visual features such as brightness and color in the image. They simulate the process of light starting from the camera position and propagating along a straight line to each object in the image. In order to accurately capture the intersection of these virtual rays and objects in the image, edge detection and image segmentation algorithms can be used. These algorithms can identify the outline and boundary of objects in the image, thereby determining the precise intersection of rays and objects. While recording these intersections, machine learning or deep learning models are also used to classify and identify objects in the image, and each intersection is assigned a unique obstacle ID or curb ID. These IDs not only identify the object to which the intersection belongs, but also contain rich attribute information about the object type, location, shape, etc.

[0035] 103. Determine whether adjacent target intersections belong to the same curb according to each intersection and the attributes of each intersection;

[0036] Using the ID information in the recorded attributes of each intersection, the intersections are preliminarily classified to identify the target intersections belonging to the curb. This can be achieved by comparing the ID attributes of the intersections with the preset curb ID or obstacle ID library. If the ID of the intersection matches the curb ID, the intersection is determined to be a point on the curb. The distances between adjacent target intersections are calculated, and combined with the spatial scale in the image, it is determined whether these distances are within a reasonable continuous range of the curb. At the same time, the directionality and continuity of the intersections are considered, that is, whether the lines connecting adjacent target intersections maintain a certain directional trend and there are no sudden turns or breaks. In addition, other feature information in the image, such as texture, color, etc., is used as an auxiliary judgment basis to enhance the robustness of recognition. If the distance, direction, and continuity of adjacent intersections all meet the characteristics of the curb and their ID attributes are the same, it can be determined that these adjacent intersections belong to the same curb.

[0037] 104. Connect the target intersection points belonging to the same curb to obtain the curb boundary.

[0038] Each target intersection is sorted to ensure that they are connected along the direction of the curb. Then, adjacent target intersections are connected using interpolation algorithms or curve fitting algorithms to form a smooth curb boundary. These algorithms can generate smooth curves that conform to the geometric characteristics of the curb based on the location and attribute information of the intersections. Finally, the connected curb boundary is verified and optimized to ensure that it conforms to the characteristics and requirements of the actual road environment. For example, the continuity, smoothness, and accuracy of the curb boundary can be checked to ensure that it meets the needs of the vehicle's autonomous driving or assisted driving system.

[0039] In an embodiment of the present invention, by utilizing images captured by a vehicle camera, combined with simulation of natural light reflection and intersection records, it is determined whether adjacent intersections belong to the same curb, and the intersections belonging to the same curb are connected to form a curb boundary, thereby improving the accuracy of curb detection and enhancing detection stability.

[0040] See also Figure 2 Another embodiment of the curb detection method in the embodiment of the present invention includes:

[0041] 201. Real-time detection of the lighting conditions of the vehicle's surrounding environment, and dynamic adjustment of the relevant parameters of the camera according to the lighting conditions;

[0042] Through the light sensor integrated in the vehicle or the image analysis technology captured by the camera, the lighting conditions of the environment are monitored in real time, including the intensity, direction and possible shadow changes of the light. According to the detected lighting conditions, the relevant parameters of the camera, such as exposure time, aperture size and ISO value, are dynamically adjusted to ensure that the camera can capture clear images with moderate contrast in different lighting environments. This dynamic adjustment not only helps to reduce image blur and overexposure / underexposure caused by lighting changes, but also improves the recognition rate and accuracy of objects in the image.

[0043] 202. Acquire an image of the environment surrounding the vehicle, where the image is captured by a camera on the vehicle and reflects the position and shape of objects in the surrounding environment;

[0044] The execution process of step 202 is similar to the execution process of step 101, and will not be described again here.

[0045] 203. Based on the natural light reflection in the image, simulate the rays emitted from the center of the camera to the surrounding environment, and record the intersection points of the rays and the objects, as well as the properties of each intersection point;

[0046] According to the light distribution in the image, the source, direction and intensity of the light are analyzed; according to the analysis results, the rays triggered from the center of the camera and emitted along the direction of the light are simulated, and each ray is assigned a corresponding intensity value; the rays are collided with the objects in the image and the corresponding intersection points and their attributes are recorded.

[0047] Gradient calculation is used to identify areas of brightness change in the image. These areas are often closely related to the source and direction of the light. Subsequently, geometric transformation methods such as Hough transform are applied to detect straight line structures in the image. These straight line structures usually represent the main direction of the light. At the same time, the intensity of the light is quantified by analyzing the brightness histogram and calculating the average brightness value of the local area. In addition, color space conversion technology can be combined, such as converting from RGB to HSV color space, to more intuitively analyze the distribution and intensity changes of light at different wavelengths; image processing software and algorithms are used to simulate a series of rays starting from the center of the camera and along the direction of the light. The number, direction and initial position of these rays are set according to the actual situation of the light in the image. According to the intensity information of the light, a corresponding intensity value is assigned to each ray. This intensity value reflects the brightness performance of the ray in the image. It is usually obtained by calculating the average brightness or maximum brightness of the area through which the ray passes. During the simulation process, physical phenomena such as light attenuation and scattering are also considered to ensure the accuracy and authenticity of the simulation results. The object in the image is identified and segmented using image processing algorithms to obtain the contour and boundary information of the object. The image processing algorithms include edge detection and contour extraction. The simulated ray starts from the center of the camera and passes through the image space along the preset light direction. It is tested for intersection with the identified object contour. The intersection test uses efficient geometric algorithms, such as the line segment and polygon intersection algorithm, to determine whether the ray intersects with the object contour and calculate the position of the intersection. At the same time, the attribute information of each intersection is recorded. The attribute information includes the coordinates of the intersection in the image, the ID of the object to which it belongs, and the light intensity or reflectivity at the intersection estimated based on factors such as the ray intensity value and the surface material of the object.

[0048] The collision detection between rays and objects in the image may be performed by: using image processing technology to identify objects from the image and extracting corresponding contours; calculating whether each ray intersects with the identified object contour; and determining the intersection position of the ray and the object contour through geometric calculation.

[0049] Edge detection algorithms are used to identify edge information in images. These edge information usually corresponds to the contour of an object. Edge detection algorithms can be Canny edge detection, etc. Morphological operations are used to optimize the integrity and smoothness of the contour to ensure that the contour can accurately reflect the shape of the object. Morphological operations such as expansion and corrosion, etc. The simulated ray starts from a preset starting point and passes through the image space along the direction of the light. In order to determine whether the ray intersects with the contour of the object, a line segment and polygon intersection algorithm is used. This algorithm determines whether the ray enters or leaves the object area by calculating the intersection points of the ray and each edge of the contour. If it is determined that the ray intersects with the contour, the position of the intersection point in the image is determined by geometric calculations, which include linear interpolation, vector operations, etc.

[0050] 204. Filter out target intersections belonging to curbs based on the attributes of each intersection;

[0051] Based on the coordinate information of the intersection in the image, the possible intersection position is preliminarily located, and the ID of the object to which each intersection belongs is checked. By comparing with the preset curb object ID database, the intersection related to the curb is identified. At the same time, the accuracy of the intersection is further verified by combining the attribute information such as the light intensity or reflectivity at the intersection estimated based on the ray intensity value and the surface material properties of the object. Specifically, the image processing algorithm and optical model are used to calculate the light intensity or reflectivity at the intersection and compare it with the known reflective characteristics of the curb material. If the light characteristics at the intersection are consistent with the reflective characteristics of the curb material, then the intersection is considered to be the target intersection belonging to the curb.

[0052] 205. Calculate the spatial distance and direction angle between adjacent target intersection points;

[0053] Obtain the coordinates of adjacent target intersections; calculate the straight-line distance between adjacent target intersections using the distance formula based on the coordinates of adjacent target intersections to obtain the spatial distance between adjacent target intersections; obtain the direction angle between adjacent target intersections by calculating the angle between the connection of adjacent target intersections and the positive direction of the horizontal axis.

[0054] Extract the coordinate information of the target intersection in the image, traverse all the target intersections, find out the adjacent pairs of target intersections by comparing their coordinate values, and use the distance formula to calculate the straight-line distance between each pair of adjacent target intersections. This distance represents their actual interval in the image space, that is, the spatial distance between adjacent target intersections. In addition, in order to determine the direction angle between adjacent target intersections, calculate the angle formed by the line connecting the adjacent intersections and the positive direction of the horizontal axis. This is usually achieved through vector operations, that is, first construct a vector representing the line connecting the adjacent intersections, and then calculate the angle between the vector and the positive direction vector of the horizontal axis to obtain the direction angle between the adjacent target intersections.

[0055] 206. Determine whether adjacent target intersections belong to the same curb based on spatial distance and direction angle;

[0056] According to the continuity characteristics of the curb, the thresholds of spatial distance and direction angle are set; the spatial distance and direction angle between adjacent target intersections are compared with the set thresholds; and the adjacent target intersections whose spatial distance and direction angle are within the set thresholds are attributed to the same curb.

[0057] Based on historical data and empirical analysis, the ranges of spatial distance and directional angles that the curb usually exhibits in the image are determined. These ranges reflect the stability and consistency of the curb as a continuous boundary in the image space. According to these ranges, the thresholds of spatial distance and directional angle are set. For spatial distance, a maximum threshold is set to ensure that the interval between adjacent target intersections is not too large, thereby maintaining the continuity of the curb. For directional angle, a variation range threshold is set to allow the direction of the line between adjacent target intersections to have a certain variation, but the variation range must be kept within the preset threshold to ensure that the directionality of the curb does not change suddenly. The setting of these thresholds is completed through repeated experiments and adjustments to ensure that they can accurately reflect the continuity characteristics of the curb.

[0058] Traverse all adjacent pairs of target intersections, and for each pair of intersections, calculate their spatial distance and direction angle respectively, and compare these values ​​with the previously set spatial distance threshold and direction angle threshold. If the spatial distance of a pair of intersections is less than or equal to the spatial distance threshold, and the range of their direction angle is within the set direction angle threshold, then the pair of intersections are considered to be continuous in space and consistent in direction, so they are classified as belonging to the same curb.

[0059] 207. Connect the target intersection points belonging to the same curb to obtain the curb boundary.

[0060] All target intersections that have been determined to belong to the same curb are sorted to ensure that they are arranged in a natural order on the curb. Linear interpolation or curve fitting algorithms are used to connect adjacent intersections one by one according to the coordinate information of these intersections. For straight line segments, the equation of the straight line between the two points is directly calculated. For curved segments, polynomial fitting, Bezier curves or other high-order curve models may be used to more accurately describe the curved shape of the curb. During the connection process, the continuity and smoothness of the boundary need to be maintained to avoid abrupt turns, while ensuring that the accuracy of the boundary meets the needs of subsequent applications. By connecting the target intersections that belong to the same curb, a complete and continuous curb boundary is obtained.

[0061] In the embodiment of the present invention, by real-time detection of lighting conditions and dynamic adjustment of camera parameters, the image capture problem in different lighting environments is effectively addressed, the stability and accuracy of the image quality are ensured, and the target intersection points belonging to the curb are screened out by simulating the intersection of rays and objects and recording the attributes. Based on the calculation of spatial distance and direction angle, accurate reconstruction of the curb boundary is achieved, providing the vehicle with reliable curb recognition information, and significantly improving the safety and reliability of the autonomous driving system.

[0062] The above describes the curb detection method in the embodiment of the present invention. The following describes the curb detection device in the embodiment of the present invention. Figure 3In one embodiment of the present invention, a curb detection device includes:

[0063] An acquisition module 301 is used to acquire an image in the surrounding environment of the vehicle, where the image is captured by a camera on the vehicle and reflects the position and shape of objects in the surrounding environment;

[0064] The processing module 302 is used to simulate the rays emitted from the center of the camera to the surrounding environment based on the natural light reflected in the image, and record the intersection points of the rays and the objects, as well as the attributes of each intersection point;

[0065] A judgment module 303 is used to judge whether adjacent target intersections belong to the same curb according to each intersection and the attributes of each intersection;

[0066] The connection module 304 is used to connect the target intersection points belonging to the same curb to obtain the curb boundary.

[0067] In an embodiment of the present invention, by utilizing images captured by a vehicle camera, combined with simulation of natural light reflection and intersection records, it is determined whether adjacent intersections belong to the same curb, and the intersections belonging to the same curb are connected to form a curb boundary, thereby improving the accuracy of curb detection and enhancing detection stability.

[0068] See also Figure 4 Another embodiment of the curb detection device in the embodiment of the present invention includes:

[0069] An acquisition module 301 is used to acquire an image in the surrounding environment of the vehicle, where the image is captured by a camera on the vehicle and reflects the position and shape of objects in the surrounding environment;

[0070] The processing module 302 is used to simulate the rays emitted from the center of the camera to the surrounding environment based on the natural light reflected in the image, and record the intersection points of the rays and the objects, as well as the attributes of each intersection point;

[0071] A judgment module 303 is used to judge whether adjacent target intersections belong to the same curb according to each intersection and the attributes of each intersection;

[0072] The connection module 304 is used to connect the target intersection points belonging to the same curb to obtain the curb boundary. Optionally, the processing module 302 includes:

[0073] An analysis unit 3021, used to analyze the source, direction and intensity of light according to the light distribution in the image;

[0074] The simulation unit 3022 is used to simulate the rays triggered from the camera center and emitted along the light direction according to the analysis result, and assign a corresponding intensity value to each ray;

[0075] The processing unit 3023 is used to perform collision detection between the ray and the object in the image, and record the corresponding intersection point and the attributes of the intersection point.

[0076] Optionally, the processing unit 3023 may be specifically configured to:

[0077] Use image processing technology to identify objects from images and extract corresponding contours; calculate whether each ray intersects with the identified object contour; and determine the intersection position of the ray and the object contour through geometric calculation.

[0078] Optionally, the judging module 303 includes:

[0079] A screening unit 3031 is used to screen out target intersection points belonging to the curb based on the attributes of each intersection point;

[0080] A calculation unit 3032 is used to calculate the spatial distance and direction angle between adjacent target intersection points;

[0081] The judging unit 3033 is used to judge whether adjacent target intersection points belong to the same curb based on the spatial distance and the direction angle.

[0082] Optionally, the calculation unit 3032 may be specifically configured to:

[0083] Obtain the coordinates of adjacent target intersections; calculate the straight-line distance between adjacent target intersections using the distance formula based on the coordinates of adjacent target intersections to obtain the spatial distance between adjacent target intersections; obtain the direction angle between adjacent target intersections by calculating the angle between the connection of adjacent target intersections and the positive direction of the horizontal axis.

[0084] Optionally, the determining unit 3033 may be specifically configured to:

[0085] According to the continuity characteristics of the curb, the thresholds of spatial distance and direction angle are set; the spatial distance and direction angle between adjacent target intersections are compared with the set thresholds; and the adjacent target intersections whose spatial distance and direction angle are within the set thresholds are attributed to the same curb.

[0086] Optionally, the curb detection device further includes:

[0087] The adjustment module 305 is used to detect the lighting conditions of the vehicle's surrounding environment in real time; according to the lighting conditions, the relevant parameters of the camera are dynamically adjusted to meet the requirements of curb detection under different lighting conditions.

[0088] In the embodiment of the present invention, by real-time detection of lighting conditions and dynamic adjustment of camera parameters, the image capture problem in different lighting environments is effectively addressed, the stability and accuracy of the image quality are ensured, and the target intersection points belonging to the curb are screened out by simulating the intersection of rays and objects and recording the attributes. Based on the calculation of spatial distance and direction angle, accurate reconstruction of the curb boundary is achieved, providing the vehicle with reliable curb recognition information, and significantly improving the safety and reliability of the autonomous driving system.

[0089] above Figure 3 and Figure 4 The curb detection device in the embodiment of the present invention is described in detail from the perspective of modular functional entities, and the curb detection device in the embodiment of the present invention is described in detail from the perspective of hardware processing.

[0090] See also Figure 5 As shown, the curb detection device includes a processor 500 and a memory 501 . The memory 501 stores machine executable instructions that can be executed by the processor 500 . The processor 500 executes the machine executable instructions to implement the above curb detection method.

[0091] Further, Figure 5 The curb detection device shown further includes a bus 502 and a communication interface 503 , and the processor 500 , the communication interface 503 and the memory 501 are connected via the bus 502 .

[0092] Among them, the memory 501 may include a high-speed random access memory (RAM), and may also include a non-volatile memory (non-volatile memory), for example, at least one disk storage. The communication connection between the system network element and at least one other network element is realized through at least one communication interface 503 (which can be wired or wireless), and the Internet, wide area network, local area network, metropolitan area network, etc. can be used. The bus 502 can be an ISA bus, a PCI bus or an EISA bus, etc. The bus can be divided into an address bus, a data bus, a control bus, etc. For ease of representation, Figure 5 Only one bidirectional arrow is used in the diagram, but this does not mean that there is only one bus or only one type of bus.

[0093] The processor 500 may be an integrated circuit chip with signal processing capabilities. In the implementation process, each step of the above method can be completed by the hardware integrated logic circuit or software instructions in the processor 500. The above processor 500 may be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), etc.; it can also be a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA) or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components. The disclosed methods, steps and logic block diagrams in the embodiments of the present disclosure can be implemented or executed. The general-purpose processor can be a microprocessor or the processor can also be any conventional processor, etc. The steps of the method disclosed in conjunction with the embodiments of the present disclosure can be directly embodied as a hardware decoding processor to be executed, or a combination of hardware and software modules in the decoding processor can be executed. The software module can be located in a mature storage medium in the field such as a random access memory, a flash memory, a read-only memory, a programmable read-only memory or an electrically erasable programmable memory, a register, etc. The storage medium is located in the memory 501 , and the processor 500 reads the information in the memory 501 and completes the method steps of the above-mentioned embodiment in combination with its hardware.

[0094] The present invention also provides a curb detection device, the computer device comprising a memory and a processor, the memory storing computer-readable instructions, and when the computer-readable instructions are executed by the processor, the processor executes the steps of the curb detection method in the above-mentioned embodiments.

[0095] The present invention also provides a computer-readable storage medium, which may be a non-volatile computer-readable storage medium or a volatile computer-readable storage medium. Instructions are stored in the computer-readable storage medium. When the instructions are executed on a computer, the computer executes the steps of the curb detection method.

[0096] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0097] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention is essentially or the part that contributes to the prior art or all or part of the technical solution can be embodied in the form of a software product. The computer software product is stored in a storage medium, including a number of instructions to enable a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM), random access memory (RAM), disk or optical disk and other media that can store program codes.

[0098] As described above, the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features thereof may be replaced by equivalents. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A curb detection method, characterized in that: The curb detection method comprises: Acquire an image of the environment surrounding the vehicle, wherein the image is captured by a camera on the vehicle and reflects the position and shape of objects in the surrounding environment; Based on the natural light reflected in the image, simulate the rays emitted from the center of the camera to the surrounding environment, and record the intersection points of the rays and the object, as well as the properties of each intersection point; Determine whether adjacent target intersections belong to the same curb based on each intersection and its attributes; Connect the target intersection points belonging to the same curb to obtain the curb boundary.

2. The method for detecting a road edge according to claim 1, characterized in that: Based on the natural light reflected in the image, the rays emitted from the center of the camera to the surrounding environment are simulated, and the intersection points of the rays and the objects and the properties of each intersection point are recorded, including: Analyzing the source, direction and intensity of the light according to the light distribution in the image; According to the analysis results, simulate the rays triggered from the center of the camera and emitted along the direction of the light, and assign corresponding intensity values ​​to each ray; The ray is subjected to collision detection with the object in the image, and the corresponding intersection points and the attributes of the intersection points are recorded.

3. The road edge detection method according to claim 2, characterized in that: The performing collision detection between the ray and the object in the image includes: Using image processing technology to identify objects from the image and extract corresponding contours; Calculate whether each ray intersects with the identified object contour; The intersection point of the ray and the object contour is determined through geometric calculation.

4. The method for detecting a road edge according to claim 1, characterized in that: The step of judging whether the intersection points of adjacent rays belong to the same curb according to the intersection points and the attributes of the intersection points includes: Filter out target intersections belonging to the curb based on the attributes of each intersection; Calculate the spatial distance and direction angle between adjacent target intersection points; It is determined whether adjacent target intersection points belong to the same curb based on the spatial distance and the direction angle.

5. The method for detecting a road edge according to claim 4, characterized in that: The calculating of the spatial distance and direction angle between adjacent target intersection points includes: Get the coordinates of the intersection points of adjacent targets; According to the coordinates of adjacent target intersection points, the straight-line distance between adjacent target intersection points is calculated using the distance formula to obtain the spatial distance between adjacent target intersection points; The direction angle between adjacent target intersection points is obtained by calculating the angle between the connection of adjacent target intersection points and the positive direction of the horizontal axis.

6. The method for detecting a road edge according to claim 4, characterized in that: The determining whether adjacent target intersection points belong to the same curb based on the spatial distance and the direction angle includes: According to the continuity characteristics of the curb, the thresholds of spatial distance and direction angle are set; Compare the spatial distance and direction angle between adjacent target intersection points with the set threshold; Adjacent target intersections whose spatial distance and direction angle are within the set threshold are attributed to the same curb.

7. The method for detecting a curb according to any one of claims 1 to 6, characterized in that: Before obtaining the natural light reflected from the vehicle's surroundings, it also includes: Real-time detection of lighting conditions around the vehicle; According to the lighting conditions, the relevant parameters of the camera are dynamically adjusted to meet the needs of curb detection under different lighting conditions.

8. A curb detection device, characterized in that: The curb detection device comprises: An acquisition module is used to acquire an image in the surrounding environment of the vehicle, wherein the image is captured by a camera on the vehicle and reflects the position and shape of objects in the surrounding environment; A processing module, for simulating rays emitted from the center of the camera to the surrounding environment based on natural light reflected light in the image, and recording intersections of the rays with the object and properties of each intersection; A judgment module, used for judging whether adjacent target intersections belong to the same curb according to the intersections and the attributes of the intersections; The connection module is used to connect the target intersection points belonging to the same curb to obtain the curb boundary.

9. A road edge detection device, characterized in that: The curb detection device comprises: a memory and at least one processor, wherein the memory stores instructions; The at least one processor calls the instructions in the memory to enable the curb detection device to perform the curb detection method according to any one of claims 1 to 7.

10. A computer-readable storage medium having instructions stored thereon, characterized in that: When the instructions are executed by a processor, the curb detection method according to any one of claims 1 to 7 is implemented.

Images