Vehicle outline dimension detection method, system and electronic equipment

Abstract

This application applies to the field of vehicle measurement technology and provides a vehicle outline dimension detection method, system, and electronic device, including: acquiring a first reference image frame, and a first point cloud frame and a second point cloud frame that match the first reference image frame; determining a first vehicle displacement of the vehicle based on each of the first reference image frame and its matching first point cloud frame and second point cloud frame; constructing a three-dimensional point cloud of the vehicle based on the first vehicle displacement, the first point cloud frame, and the second point cloud frame; and determining the vehicle's dimension information based on the three-dimensional point cloud. This application has low detection site requirements and can effectively improve the efficiency of vehicle outline detection.

Description

Technical Field

[0001] The present application relates to the field of vehicle measurement technology, and in particular to a vehicle profile dimension detection method, system, and electronic equipment. Background Art

[0002] Currently, vehicle dimensions are a mandatory inspection item for overload detection at highway entrances. However, toll booth entrances often have limited space, irregular vehicle movement, and the presence of pedestrians and other distractions. Existing vehicle dimension inspection technologies are complex, require high site requirements, and impose strict inspection conditions, making them unsuitable for the free-flowing conditions at highway toll booth entrances.

[0003] In summary, the existing vehicle profile dimension detection technology has high requirements on the detection site and low detection efficiency. Summary of the Invention

[0004] The embodiments of the present application provide a vehicle outline dimension detection method, system and electronic device, which can solve the problems in the prior art of existing vehicle outline dimension detection technology that the existing technology has high requirements for the detection site and low detection efficiency.

[0005] In a first aspect, an embodiment of the present application provides a vehicle outline dimension detection method, comprising:

[0006] Acquire a first reference image frame, and a first point cloud frame and a second point cloud frame matching the first reference image frame, wherein the first reference image frame is any first vehicle image frame in a sequence of first vehicle image frames captured by photographing the vehicle with a first camera, the first point cloud frame is a point cloud frame scanned by a first scanning laser rangefinder, and the second point cloud frame is a point cloud frame scanned by a second scanning laser rangefinder;

[0007] determining a first vehicle displacement of the vehicle based on each of the first reference image frames and the matched first point cloud frames and the second point cloud frames;

[0008] constructing a three-dimensional point cloud of the vehicle based on the first vehicle displacement, the first point cloud frame, and the second point cloud frame;

[0009] Determine the size information of the vehicle based on the three-dimensional point cloud.

[0010] In a second aspect, an embodiment of the present application provides a vehicle outline dimension detection system, comprising: a first scanning laser rangefinder, a first camera, a second scanning laser rangefinder, and a data processing device;

[0011] The first camera, the first scanning laser rangefinder, and the second scanning laser rangefinder are communicatively connected to the data processing device;

[0012] The first scanning laser rangefinder and the second scanning laser rangefinder are used to scan the vehicle to obtain a point cloud frame;

[0013] The first camera is used to photograph the vehicle and collect a first vehicle image frame sequence;

[0014] The data processing device is configured to acquire a first reference image frame, and a first point cloud frame and a second point cloud frame matching the first reference image frame, wherein the first reference image frame is any first vehicle image frame in the first vehicle image frame sequence, the first point cloud frame is a point cloud frame scanned by a first scanning laser rangefinder, and the second point cloud frame is a point cloud frame scanned by a second scanning laser rangefinder;

[0015] The data processing device is further used to determine the first vehicle displacement of the vehicle based on each of the first reference image frames and their matching first point cloud frames and second point cloud frames; construct a three-dimensional point cloud of the vehicle based on the first vehicle displacement, the first point cloud frame and the second point cloud frame; and determine the size information of the vehicle based on the three-dimensional point cloud.

[0016] In a third aspect, an embodiment of the present application provides an electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the vehicle contour dimension detection method as described in the first aspect above is implemented.

[0017] In a fourth aspect, an embodiment of the present application provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the vehicle contour dimension detection method as described in the first aspect above is implemented.

[0018] In a fifth aspect, an embodiment of the present application provides a computer program product. When the computer program product is run on an electronic device, the electronic device executes the vehicle outline dimension detection method as described in the first aspect above.

[0019] In an embodiment of the present application, by acquiring a first reference image frame, and a first point cloud frame and a second point cloud frame that match the first reference image frame, and then determining the first vehicle displacement of the vehicle based on each of the first reference image frame and its matching first point cloud frame and second point cloud frame, the three-dimensional point cloud of the vehicle can be accurately constructed based on the first vehicle displacement, the first point cloud frame, and the second point cloud frame, and the requirements for the vehicle's driving state are not high. Finally, based on the three-dimensional point cloud, the size information of the vehicle is determined, thereby enabling efficient, accurate, and non-contact detection of the vehicle's contour dimensions. The present application solution has low requirements for the detection site and strong anti-pedestrian interference capabilities, which can effectively improve the efficiency of vehicle contour detection. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the following briefly introduces the drawings required for use in the embodiments or descriptions of the prior art. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0021] Figure 1 This is a system architecture diagram of a vehicle outline dimension detection system provided by an embodiment of the present application;

[0022] Figure 1.1 This is a schematic diagram of an application scenario of the vehicle outline dimension detection system provided in an embodiment of the present application;

[0023] Figure 1.2 This is a schematic diagram of another application scenario of the vehicle outline dimension detection system provided in an embodiment of the present application;

[0024] Figure 2 This is a flow chart of an implementation method for detecting vehicle outline dimensions provided by an embodiment of the present application;

[0025] Figure 3 This is a flowchart of a specific implementation of step S102 of the vehicle outline dimension detection method provided in an embodiment of the present application;

[0026] Figure 3a This is a specific implementation flowchart of extracting features from the first reference image frame to obtain first feature points in an embodiment of the present application;

[0027] Figure 3b is a flowchart of a specific implementation process for calculating the relative pixel displacement between the first reference image frame and each of the first adjacent image frames in an embodiment of the present application;

[0028] Figure 4 A specific implementation flow chart of determining the displacement of the first vehicle in an embodiment of the present application;

[0029] Figure 5 This is a schematic diagram of a scenario for calculating relative displacement information provided by an embodiment of the present application;

[0030] Figure 6 is a flow chart of an implementation method of a vehicle outline dimension detection method provided by another embodiment of the present application;

[0031] Figure 7 This is a flowchart of a specific implementation of step S205 of the vehicle outline dimension detection method provided in an embodiment of the present application;

[0032] Figure 8Schematic diagram of an electronic device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0033] In the following description, specific details such as specific system structures and techniques are provided for purposes of illustration rather than limitation to facilitate a thorough understanding of the embodiments of the present application. However, it will be apparent to those skilled in the art that the present application may be implemented in other embodiments without these specific details. In other cases, detailed descriptions of well-known systems, devices, circuits, and methods are omitted to avoid obscuring the description of the present application with unnecessary detail.

[0034] It should be understood that when used in the present specification and the appended claims, the term "comprising" indicates the presence of described features, integers, steps, operations, elements and / or components, but does not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components and / or collections thereof.

[0035] It will also be understood that the term "and / or" used in this specification and the appended claims refers to and includes any and all possible combinations of one or more of the associated listed items.

[0036] In addition, in the description of the present application specification and the appended claims, the terms "first", "second", "third", etc. are only used to distinguish the descriptions and cannot be understood as indicating or implying relative importance.

[0037] References to "one embodiment" or "some embodiments" in this specification mean that a particular feature, structure, or characteristic described in conjunction with that embodiment is included in one or more embodiments of the present application. Thus, phrases such as "in one embodiment," "in some embodiments," "in other embodiments," and "in other embodiments" appearing in various places in this specification do not necessarily refer to the same embodiment, but rather mean "one or more but not all embodiments," unless otherwise specifically emphasized. The terms "including," "comprising," "having," and variations thereof all mean "including but not limited to," unless otherwise specifically emphasized.

[0038] At present, existing vehicle profile detection technologies have high site requirements, complex equipment components, and strict detection conditions. For example, multi-gantry laser measurement systems not only have high site requirements, but also when measuring length, the ranging sensor used for length detection is far away from the vehicle width and height detection area, which may easily cause tall vehicles to block short vehicles, resulting in inaccurate length measurement of short vehicles. The detection equipment is also easily interfered with by pedestrians. The detection conditions are harsh and cannot adapt to the free flow situation at highway entrance toll stations. The vehicle profile measurement system based on image stitching technology has high site requirements.

[0039] To address the above issues, the present invention provides a method, system, and electronic device for detecting vehicle outline dimensions, which can efficiently, accurately, and contactlessly complete the detection of vehicle outline dimensions. This method not only has low requirements for the detection site, but also has strong resistance to pedestrian interference.

[0040] It should be understood that the vehicle outline dimension detection method provided in the embodiment of the present application can be applied to electronic devices such as servers and ultra-mobile personal computers (UMPCs), and the embodiment of the present application does not impose any restrictions on the specific type of electronic devices.

[0041] Figure 1 A system architecture diagram of a vehicle outline dimension detection system provided in an embodiment of the present application is shown, which is detailed as follows: For ease of explanation, only the parts related to the embodiment of the present invention are shown.

[0042] Reference Figure 1 The vehicle contour dimension detection system includes a first scanning laser rangefinder 1, a second scanning laser rangefinder 2, a first camera 3 and a data processing device 4. The first scanning laser rangefinder 1, the second scanning laser rangefinder 2, the first camera 3 are communicatively connected to the data processing device 4.

[0043] As an implementation method of this application, Figure 1.1 As shown, the first camera is mounted on the crossbar of the gantry 5, which includes a crossbar and two vertical poles. The gantry 5 is mounted on the safety island 6 of the toll station. The first scanning laser rangefinder 1 and the second scanning laser rangefinder 2 are mounted above the vertical poles on both sides of the gantry 5, facing each other. In this embodiment of the present application, the crossbar of the gantry 5 is at least 5 meters above the ground and is perpendicular to the direction of vehicle travel.

[0044] The first scanning laser rangefinder 1 and the second scanning laser rangefinder 2 are used to scan the vehicle to obtain a point cloud frame. In the embodiment of the present application, the vertical height of the gantry 5 is no less than 5 meters above the ground. The scanning cross-section 1 of the first scanning laser rangefinder is parallel to the scanning cross-section of the second scanning laser rangefinder 2, and the angle between them and the direction of travel is within 45 degrees to 90 degrees. The distance between them in the direction of travel is no less than 10 centimeters.

[0045] For example, both the first scanning laser rangefinder 1 and the second scanning laser rangefinder 2 perform two-dimensional line scanning, with a scanning angle of 90 degrees, a scanning frequency of 100 Hz, and an angular resolution of 0.1 degrees. The scanning directions of the first scanning laser rangefinder 1 and the second scanning laser rangefinder 2 are perpendicular to the vehicle's travel direction, and the scanning sections are spaced 30 cm apart in the driving direction.

[0046] The first camera 3 is used to photograph the vehicle and collect a first vehicle image frame sequence. In the embodiment of the present application, the first camera 3 has an angle of 60 to 90 degrees with respect to the horizontal direction, a focal length of 4 to 8 mm, and an image refresh frequency of not less than 50 Hz.

[0047] Exemplarily, the image refresh frequency of the first camera 3 is 100 Hz, the focal length is 4 mm, and the first camera 3 captures the first vehicle image vertically downward.

[0048] In an embodiment of the present application, the shooting angle of the first camera can be freely adjusted. By adjusting the shooting angle of the first camera, the viewing angle range of the camera can be expanded, and effective feature points can be matched in more adjacent image frames, thereby improving the stability of the vehicle image displacement calculation.

[0049] The data processing device 4 is used to obtain a first reference image frame, and a first point cloud frame and a second point cloud frame matching the first reference image frame, where the first reference image frame is any first vehicle image frame in the first vehicle image frame sequence, the first point cloud frame is a point cloud frame scanned by the first scanning laser rangefinder 1, and the second point cloud frame is a point cloud frame scanned by the second scanning laser rangefinder 2.

[0050] The data processing device 4 is also used to determine the first vehicle displacement of the vehicle based on each of the first reference image frames and their matching first point cloud frames and second point cloud frames; construct a three-dimensional point cloud of the vehicle based on the first vehicle displacement, the first point cloud frame and the second point cloud frame; and determine the size information of the vehicle based on the three-dimensional point cloud.

[0051] The data processing device 4 may be a mobile device, a server or other electronic device with communication capabilities.

[0052] As a possible implementation of the present application, the vehicle outline dimension detection system can use Bluetooth technology, WIFI technology, or 3G / 4G / 5G technology to establish a wireless connection between the data processing device 4 and the first scanning laser rangefinder 1, the second scanning laser rangefinder 2, and the first camera 3. Of course, in some possible implementations, serial port technology or USB interface technology can also be used to establish a wired connection between the data processing device 4 and the first scanning laser rangefinder 1, the second scanning laser rangefinder 2, and the first camera 3.

[0053] As a possible implementation of the present application, the vehicle outline size detection system further includes a second camera 7, which is in communication with the data processing device 4. Figure 1.2As shown, in this embodiment of the present application, the first camera 3 is mounted on the mast crossbar, and the second camera 7 is mounted on a side vertical bar of the mast 5. The first camera 3 faces downward at a 30-degree angle to the vertical, and is used to capture images of the vehicle's roof. The second camera 7 is placed horizontally, perpendicular to the vehicle's direction of travel, and faces the vehicle body, capturing images of the second vehicle vertically downward.

[0054] For example, the first camera 3 is installed on the horizontal bar of the gantry at a height of 5.5 meters from the ground, and the second camera 7 is installed on a side vertical bar of the gantry at a height of 1.5 meters from the ground. The image refresh frequencies of the first camera 3 and the second camera 7 are both 100 Hz, and the focal lengths are both 6 mm. The first camera 3 faces downward and has an angle of 30 degrees with the vertical direction, and is used to capture the roof image of the vehicle; the second camera 7 is placed horizontally, perpendicular to the driving direction of the vehicle, and horizontally facing the vehicle body, and is used to capture the second vehicle image.

[0055] The above-mentioned data processing device 4 is also used to obtain a second reference image frame, which is any second vehicle image frame in a second vehicle image frame sequence collected by shooting the vehicle with the second camera 7, and the second reference image frame is matched with the first point cloud frame, the second point cloud frame and the first reference image frame; based on each second reference image frame and its matching first point cloud frame and second point cloud frame, the second vehicle displacement of the vehicle is determined; based on the second vehicle displacement, the three-dimensional point cloud is corrected; and based on the corrected three-dimensional point cloud, the size information of the vehicle is determined.

[0056] In an embodiment of the present application, by adjusting the shooting angle of the first camera and expanding the viewing angle range, effective feature points can be matched in more adjacent image frames. By adding a second camera on the side, the range of feature point extraction and matching is expanded, and effective feature points can be matched in more adjacent image frames, thereby making the constructed three-dimensional point cloud more accurate and effective, and more stable.

[0057] In an embodiment of the present application, each device in the above-mentioned vehicle contour dimension detection system is simple to install and has low requirements for the detection site. The vehicle image frame captured by the camera is obtained through a data processing device, and the first point cloud frame and the second point cloud frame are captured by two scanning laser rangefinders. The vehicle image frame is matched with the point cloud frame to determine the reference image frame and its matching first point cloud frame and second point cloud frame. Then, based on each of the first reference image frames and its matching first point cloud frame and second point cloud frame, the first vehicle displacement of the vehicle is determined. Then, based on the first vehicle displacement, the first point cloud frame and the second point cloud frame, the three-dimensional point cloud of the vehicle can be accurately constructed. At the same time, the driving state of the vehicle is not required. Finally, the size information of the vehicle is determined based on the three-dimensional point cloud, and the vehicle contour dimension detection can be completed efficiently, accurately and non-contact.

[0058] It should be noted that the information interaction, execution process, etc. between the above-mentioned systems / units are based on the same concept as the method embodiment of this application. Their specific functions and technical effects can be found in the method embodiment section and will not be repeated here.

[0059] Figure 2 The implementation process of the vehicle outline size detection method provided by the embodiment of the present application is shown. The execution subject of the process is Figure 1 The data processing device 4 shown in FIG. 4 includes steps S101 to S104. The specific implementation principles of each step are as follows:

[0060] S101: Acquire a first reference image frame, and a first point cloud frame and a second point cloud frame that match the first reference image frame.

[0061] The above-mentioned first reference image frame is any first vehicle image frame in the first vehicle image frame sequence collected by shooting the vehicle with the first camera, the above-mentioned first point cloud frame is the point cloud frame scanned by the first scanning laser rangefinder, and the above-mentioned second point cloud frame is the point cloud frame scanned by the second scanning laser rangefinder.

[0062] In one embodiment, a data processing device obtains a first vehicle image frame sequence, a first point cloud frame sequence, and a second point cloud frame sequence within the same vehicle detection period, selects any one frame from the first vehicle image frame sequence as the first reference image frame, and based on the acquisition time of the first reference image frame, selects a first point cloud frame with the same acquisition time as the first reference image frame from the first point cloud frame sequence, and selects a second point cloud frame with the same acquisition time as the first reference image frame from the second point cloud frame sequence to complete the matching.

[0063] By matching the first reference image frame with the first point cloud frame and the second point cloud frame, calculation errors can be reduced and detection accuracy can be improved.

[0064] The matching of the reference image frame with the first and second point cloud frames is based on the acquisition time. In this embodiment of the present application, the first vehicle image frame sequence is a sequence of image frames acquired by the first camera during a vehicle detection period, the first scanning laser rangefinder scans the vehicle during the vehicle detection period to obtain the first point cloud frame sequence, and the second scanning laser rangefinder scans the vehicle during the vehicle detection period to obtain the second point cloud frame sequence. The first vehicle image frame sequence is matched with the first and second point cloud frame sequences based on the acquisition time, that is, the first vehicle image frame, first point cloud frame, and second point cloud frame at the same acquisition time are matched.

[0065] Exemplarily, an arbitrary frame is selected from the above-mentioned first vehicle image frame sequence as the reference image frame, the above-mentioned first point cloud frame is a point cloud frame in the above-mentioned first point cloud frame sequence with the same acquisition time as the reference image frame, and the above-mentioned second point cloud frame is a point cloud frame in the above-mentioned second point cloud frame sequence with the same acquisition time as the reference image frame.

[0066] In a possible implementation, the vehicle detection period starts when the first scanning laser rangefinder detects a vehicle and ends when the second scanning laser rangefinder fails to detect a vehicle.

[0067] For example, in one application scenario, a vehicle travels along a lane between toll booth safety islands. When the front of the vehicle triggers the first scanning laser rangefinder, detection begins. The first and second scanning laser rangefinders respectively capture point clouds of the vehicle's left and right cross-sections, while the first camera captures the vehicle's body image. Detection ceases when the rear of the vehicle leaves the second scanning laser rangefinder. During the vehicle detection period, based on the timestamps of the scanning laser rangefinder and the first camera, the first vehicle image frame captured by the first camera during that period is synchronously matched with the first point cloud frame captured by the first scanning laser rangefinder and the second point cloud frame captured by the second scanning laser rangefinder.

[0068] S102: Determine a first vehicle displacement of the vehicle according to each of the first reference image frames and the matched first point cloud frames and the second point cloud frames.

[0069] As a possible implementation of this application, Figure 3 The specific implementation process of step S102 of the vehicle outline dimension detection method provided in the embodiment of the present application is shown and is detailed as follows:

[0070] A1: Perform feature extraction on the first reference image frame according to the first point cloud frame and the second point cloud frame to obtain first feature points.

[0071] The first point cloud frame includes the first vehicle body point cloud, and the second point cloud frame includes the second vehicle body point cloud. As a possible implementation of the present application, Figure 3a As shown, the specific process of the above step A1 includes:

[0072] A11: Obtain first pixel coordinates of the first vehicle body point cloud in the first reference image frame, and second pixel coordinates of the second vehicle body point cloud in the first reference image frame.

[0073] In the embodiment of the present application, the pixel coordinates of the vehicle body point cloud in the first reference image frame are calculated through projective transformation between the vehicle body point cloud and the first reference image frame.

[0074] The single-point laser ranging of a scanning laser rangefinder has spatial three-dimensional coordinate information, while the camera image pixels are two-dimensional plane coordinate information. The laser ranging point must first be transformed from the world coordinate system to the camera coordinates through rotation, translation, scaling, and other transformations, and then multiplied by the camera intrinsic parameter matrix to convert it to the image coordinates. This embodiment uses a scanning laser rangefinder. The point cloud frame data obtained is a two-dimensional cross-section of the vehicle, and the transformation from two-dimensional laser radar coordinates to image coordinates can be more simply viewed as a projective transformation. The laser point of the two-dimensional laser is equivalent to the top view, and the image captured by the camera can be imagined as the main view. The process of transforming the top view to the main view by multiplying a homography matrix is ​​the projective transformation. The general form of the transformation is as follows (1):

[0075]

[0076] where [xy 1] T is the homogeneous coordinate of the lidar, [uv 1] T is the homogeneous coordinate of the image pixel after the laser is converted to the image coordinate system. h9 in our h matrix is ​​used as the scaling factor λ and extracted to the left side of the equation. The equation can be written as follows (2):

[0077]

[0078] Eliminating the scaling factor λ, we get formula (3):

[0079]

[0080] By selecting k (k>4) groups of lidar points and corresponding pixel coordinate points, an overdetermined equation can be formed and the optimal solution of the n vector can be obtained using least squares.

[0081] A12: Determine a mapping area according to the first pixel coordinates and the second pixel coordinates.

[0082] In this embodiment, according to the first pixel coordinates and the second pixel coordinates, a circumscribed envelope rectangle or parallelogram with a width of 50 pixels corresponding to the image points of the first vehicle body point cloud and the second vehicle body point cloud in the first reference image frame is selected as the mapping area.

[0083] A13: Extract features from the mapped area to obtain first feature points. Feature points are selected from SIFT and / or HOG.

[0084] In an embodiment of the present application, since the pixel points corresponding to the point cloud are discrete, a connected mapping area is determined based on the pixel coordinates corresponding to the vehicle body point cloud in the first vehicle image frame to improve the number and quality of feature points, thereby improving the accuracy of vehicle contour detection.

[0085] A2: Acquire a first adjacent image frame sequence, where a capture time of the first adjacent image frame sequence is adjacent to a capture time of the first reference image frame, and the first adjacent image frame sequence includes a plurality of first adjacent image frames.

[0086] In this embodiment, the first adjacent image frame is the first vehicle image frame in the first vehicle image frame sequence whose capture time is adjacent to the first reference image frame. The first adjacent image frame sequence includes at least two first vehicle image frames whose capture time is adjacent to the first reference image frame. The specific number of frames in the first adjacent image frame sequence can be determined based on computing requirements.

[0087] In one embodiment, the first adjacent image frame sequence includes a first adjacent subsequence and a second adjacent subsequence, wherein the first adjacent subsequence includes a number of first vehicle image frames whose consecutive frames were captured before the first reference image frame, and the second adjacent subsequence includes a number of first vehicle image frames whose consecutive frames were captured after the first reference image frame. It should be noted that when the first reference image frame is the first frame in the first vehicle image frame sequence, the first adjacent subsequence is empty. When the first reference image frame is the last frame in the first vehicle image frame sequence, the second adjacent subsequence is empty.

[0088] A3: Perform feature extraction on the first adjacent image frames to obtain a plurality of first adjacent feature points.

[0089] In an embodiment of the present application, feature extraction is performed on each frame of the vehicle image in the first adjacent image frame sequence, and the above-mentioned plurality of first adjacent feature points respectively correspond to the first adjacent image frames in the above-mentioned first adjacent image frame sequence.

[0090] In this embodiment, the algorithm for extracting features from the first vehicle image frame in the first sequence of adjacent image frames refers to the feature extraction algorithm for the first reference image frame in step S102 , and is not described in detail here.

[0091] A4: Calculate the relative pixel displacement between the first reference image frame and each of the first adjacent image frames based on the first feature point and the plurality of first adjacent feature points to obtain a plurality of relative pixel displacements.

[0092] As a possible implementation of this application, Figure 3b The specific implementation process of calculating the relative pixel displacement between the first reference image frame and each of the first adjacent image frames in the embodiment of the present application is shown, including:

[0093] A41: Obtain an image pixel coordinate difference, where the image pixel coordinate difference is a coordinate difference between the pixel coordinates of the first feature point and the pixel coordinates of the first adjacent feature point. In this embodiment, the coordinate difference is an absolute value.

[0094] A42: Obtain the pixel size of the first camera. In this embodiment, the pixel size of the first camera can be determined by consulting the camera manual corresponding to the first camera. A pixel is an imaging unit and is the smallest unit that constitutes a digital image. During scanning imaging, a pixel is the smallest unit used by a sensor to scan and sample ground scenes. In digital image processing, it is the sampling point when scanning and digitizing an analog image. Generally, the pixel size is L*W, where W is the physical width of a single pixel in the direction of vehicle travel and perpendicular to the direction of travel, and L is the physical height of a single pixel in the direction of vehicle travel and perpendicular to the direction of travel.

[0095] A43: Determine the relative pixel displacement according to the product of the image pixel coordinate difference and the pixel size.

[0096] In the embodiment of the present application, the relative pixel displacement between the reference image frame and its adjacent image frame can be determined based on the product of the image pixel coordinate difference between the first feature point and its first adjacent feature point and the pixel size.

[0097] A5: Determine the first vehicle displacement according to the plurality of relative pixel displacements corresponding to each of the first reference images.

[0098] As a possible implementation method of this application, Figure 4 The specific implementation process of determining the displacement of the first vehicle is shown, including:

[0099] B1: Obtain a first point cloud distance and the focal length of the first camera. The first point cloud distance refers to the distance from the vehicle body point cloud corresponding to the first feature point within the mapping area to the focal point of the first camera. The vehicle body point cloud may be the first vehicle body point cloud or the second vehicle body point cloud. The focal length of the first camera can be determined by consulting the camera manual corresponding to the first camera.

[0100] In an embodiment of the present application, after extracting a first feature point from a first reference image frame, if, in the first reference image frame, there are pixels mapped to a vehicle body point cloud within a range of no more than 15 pixels surrounding the first feature point, the distance from the vehicle body point cloud to the first camera focus is determined as the first point cloud distance, where the vehicle body point cloud may be the first vehicle body point cloud or the second vehicle body point cloud. If the first feature point corresponds to more than one vehicle body point cloud, the average distance from the vehicle body point clouds corresponding to the first feature point to the camera focus is obtained, and this average distance is determined as the first point cloud distance. In some embodiments, if there are no pixels mapped to a vehicle body point cloud within a range of no more than 15 pixels surrounding the first feature point, the first feature point is discarded.

[0101] B2: determining the relative displacements between the first reference image frame and its plurality of first adjacent point cloud frames in sequence according to the first point cloud distance, the focal length of the first camera, and the plurality of relative pixel displacements.

[0102] Exemplarily, the relative displacement between the first reference image frame and the first adjacent image frame with sequence number 1 is determined based on the first point cloud distance, the first camera focal length, and the relative pixel displacement between the first reference image frame and the first adjacent image frame with sequence number 1 in the first adjacent image frame sequence; the relative displacement between the first reference image frame and the first adjacent image frame with sequence number 2 is determined based on the first point cloud distance, the first camera focal length, and the relative pixel displacement between the first reference image frame and the first adjacent image frame with sequence number 2 in the first adjacent image frame sequence; and so on, the relative displacement between the first reference image frame and several of its first adjacent point cloud frames is determined in sequence.

[0103] B3: Determine the first vehicle displacement according to the relative displacement corresponding to each of the first reference image frames.

[0104] In the embodiment of the present application, a displacement curve is fitted according to the relative displacement corresponding to each of the first reference image frames, so that the relative displacement information between the first reference image frame and several first adjacent image frames can be corrected, thereby determining the first vehicle displacement.

[0105] S103: Constructing a three-dimensional point cloud of the vehicle according to the first vehicle displacement, the first point cloud frame, and the second point cloud frame.

[0106] When each first vehicle image frame is obtained as a first reference image frame, the corresponding first vehicle displacement is obtained, and a three-dimensional point cloud of the vehicle is constructed according to the first vehicle displacement, the first point cloud frame, and the second point cloud frame.

[0107] Since the camera is a two-dimensional projection of a real object on the imaging plane, the image size of the object is related to the real size of the object and the distance from the imaging plane. Figure 5 As shown, Figure 5 It includes feature points 51 and 52 on the vehicle body, an imaging plane 55, the corresponding image point 54 of the feature point 51 on the imaging plane 55, the corresponding image point 53 of the feature point 52 on the imaging plane 55, the equivalent focus o56 of the pinhole imaging model, a feature point 57, and the corresponding image point 58 of the feature point 57 on the imaging plane 55.

[0108] The vertical distances from the feature point 51 and the feature point 52 on the vehicle body to the imaging plane are the same. According to the pinhole imaging model, in the direction of vehicle travel, the following conditions are satisfied:

[0109] d=n*W*D / f

[0110] Where d is the distance difference between feature point 51 and feature point 52 in the vehicle's travel direction, n is the pixel coordinate difference between corresponding image points 54 and 53 on imaging plane 55, W is the physical size of a single pixel in the vehicle's travel direction, f is the distance from equivalent focal point o56 to imaging plane 55, and D is the vertical distance from feature point 51 to equivalent focal point o56. This formula can be used to calibrate the camera's focal length f. If feature point 51 and feature point 52 are the same point, d can be used to calculate their displacement in the travel direction.

[0111] If two feature points with different distances to the imaging plane are located directly below the focus, their corresponding image points on the imaging plane will coincide in the vertical direction of the vehicle. Figure 5 The positions 52 and 57 shown in the figure have different movement distances on the imaging plane. Because the top or side of the vehicle body is often not a flat surface, the distance from each point on the vehicle body to the camera varies. During image template matching or image feature point matching, the distance from the successfully matched area to the camera cannot be determined. Therefore, the resulting pixel displacement does not reflect the actual vehicle displacement. A common solution is to move the vehicle away from the camera or reduce the camera's field of view. Considering the limited distance from the camera to the vehicle body in toll lanes, due to the viewing angle, more points on the vehicle body with inconsistent distances from the camera will be mapped into the image. Furthermore, the image changes will be greater for the same displacement of the vehicle body. Therefore, reducing the field of view cannot reduce the error caused by the different vehicle body distances. Therefore, vehicle image stitching based on image template matching or image feature point matching will have certain errors in toll lanes, where the camera is close to the vehicle body. This embodiment uses a scanning laser rangefinder to measure the distance from the vehicle body feature points to the camera. This can calculate the vehicle's actual displacement information and thus correct for these feature point or template matching errors.

[0112] S104: Determine the size information of the vehicle according to the three-dimensional point cloud.

[0113] In the embodiment of the present application, the vehicle's size information includes the vehicle's width, height, and length. Based on the three-dimensional point cloud, the circumscribed cube of the vehicle's three-dimensional point cloud is calculated, thereby obtaining the vehicle's size information such as width, height, and length.

[0114] Exemplarily, since the point cloud frame matches the vehicle image frame, the point cloud in the point cloud frame is taken as the displacement coordinate 0 (the vehicle's travel direction coordinate). The displacement coordinates of the point cloud in each subsequent point cloud frame are the displacement coordinates of the point cloud in the previous point cloud frame plus the relative displacement of the corresponding reference image frame and the next adjacent image frame. Based on the relative displacement, the three-dimensional point cloud of the vehicle is established, and then the circumscribed cube of the three-dimensional point cloud of the vehicle is calculated to obtain the vehicle's width, height, and length and other dimensional information. The point cloud frame includes a first point cloud frame and a second point cloud frame. The vehicle image frame is specifically the first vehicle image frame, and the reference image frame is specifically the first reference image frame.

[0115] In an embodiment of the present application, by acquiring a first reference image frame, and a first point cloud frame and a second point cloud frame that match the first reference image frame, and then determining the first vehicle displacement of the vehicle based on each of the first reference image frame and its matching first point cloud frame and second point cloud frame, the three-dimensional point cloud of the vehicle can be accurately constructed based on the first vehicle displacement, the first point cloud frame, and the second point cloud frame, and the requirements for the vehicle's driving state are not high. Finally, based on the three-dimensional point cloud, the size information of the vehicle is determined, thereby enabling efficient, accurate, and non-contact detection of the vehicle's contour dimensions. The present application solution has low requirements for the detection site and strong anti-pedestrian interference capabilities, which can effectively improve the efficiency of vehicle contour detection.

[0116] As a possible implementation of this application, Figure 6 Another vehicle outline dimension detection method provided by an embodiment of the present application is shown, and is described in detail as follows:

[0117] S201: Acquire a first reference image frame, and a first point cloud frame and a second point cloud frame that match the first reference image frame.

[0118] The first reference image frame is any first vehicle image frame in a first vehicle image frame sequence captured by a first camera shooting the vehicle, the first point cloud frame is a point cloud frame scanned by a first scanning laser rangefinder, and the second point cloud frame is a point cloud frame scanned by a second scanning laser rangefinder.

[0119] S202: Determine a first vehicle displacement of the vehicle according to each of the first reference image frames and the matched first point cloud frames and the second point cloud frames.

[0120] S203: Constructing a three-dimensional point cloud of the vehicle according to the first vehicle displacement, the first point cloud frame, and the second point cloud frame.

[0121] In this embodiment, the specific steps of steps S201 to S203 refer to steps S101 to S103 in the above embodiment, which will not be repeated here.

[0122] S204: Acquire a second reference image frame.

[0123] The second reference image frame is any second vehicle image frame in a second vehicle image frame sequence captured by a second camera, the second reference image frame matches the first point cloud frame, the second point cloud frame, and the first reference image frame. The second reference image frame and the first reference image frame are captured at the same time.

[0124] In an embodiment of the present application, a first camera is mounted on a horizontal bar of the gantry, and a second camera is mounted on a vertical bar on one side of the gantry. The first camera faces downward and has an angle with the vertical direction, for example, a 30-degree angle, and is used to capture an image of the first vehicle, where the first vehicle image is an image of the roof of the vehicle; the second camera is placed horizontally, perpendicular to the driving direction of the vehicle, and horizontally facing the vehicle body, and is used to capture an image of the second vehicle.

[0125] The second reference image frame is any second vehicle image frame in a second vehicle image frame sequence collected by photographing the vehicle with a second camera.

[0126] In one embodiment, a data processing device obtains a second vehicle image frame sequence, a first point cloud frame sequence, and a second point cloud frame sequence within the same vehicle detection period, selects any one frame from the second vehicle image frame sequence as the second reference image frame, and according to the acquisition time of the second reference image frame, selects a first point cloud frame with the same acquisition time as the first reference image frame from the first point cloud frame sequence, and selects a second point cloud frame with the same acquisition time as the second reference image frame from the second point cloud frame sequence to complete the matching.

[0127] S205: Determine a second vehicle displacement of the vehicle according to each of the second reference image frames and the matched first point cloud frames and the second point cloud frames.

[0128] In this embodiment, feature extraction is performed on the second reference image frame based on the first point cloud frame and the second point cloud frame to obtain second feature points. A second adjacent image frame sequence is acquired, where the acquisition time of the second adjacent image frame sequence is adjacent to the acquisition time of the second reference image frame, and the second adjacent image frame sequence includes a plurality of second adjacent image frames. Feature extraction is performed on the second adjacent image frames to obtain a plurality of second adjacent feature points. Based on the second feature points and the plurality of second adjacent feature points, the relative pixel displacement between the second reference image frame and each of the second adjacent image frames is calculated to obtain a plurality of relative pixel displacements. The second vehicle displacement is determined based on the plurality of relative pixel displacements corresponding to each of the second reference images.

[0129] As a possible implementation of this application, Figure 7The specific implementation process of step S205 of the vehicle outline size detection method provided in the embodiment of the present application is shown and detailed as follows:

[0130] C1: Obtain the second point cloud distance and the focal length of the second camera. The second point cloud distance refers to the distance from the vehicle body point cloud corresponding to the second feature point within the mapping area of ​​the second reference image frame to the focal point of the second camera. The vehicle body point cloud may be the first vehicle body point cloud or the second vehicle body point cloud. The focal length of the second camera can be determined by consulting the corresponding camera manual.

[0131] C2: determining, in sequence, the corrected displacements between the second reference image frame and the plurality of second adjacent point cloud frames thereof according to the second point cloud distance, the focal length of the second camera, and the plurality of corrected pixel displacements.

[0132] Exemplarily, the corrected displacement between the second reference image frame and the second adjacent image frame with sequence number 1 is determined based on the second point cloud distance, the second camera focal length, and the corrected pixel displacement between the second reference image frame and the second adjacent image frame with sequence number 1 in the second adjacent image frame sequence; the corrected displacement between the second reference image frame and the second adjacent image frame with sequence number 2 is determined based on the second point cloud distance, the second camera focal length, and the corrected pixel displacement between the second reference image frame and the second adjacent image frame with sequence number 2 in the second adjacent image frame sequence; and so on, the corrected displacements between the second reference image frame and several of its second adjacent point cloud frames are determined in sequence.

[0133] C3: Determine the second vehicle displacement according to the corrected displacement corresponding to each of the second reference image frames.

[0134] In this embodiment, the specific algorithm for determining the second vehicle displacement of the vehicle can refer to the specific algorithm for determining the first vehicle displacement of the vehicle in step S102, which will not be repeated here.

[0135] S206: Correcting the three-dimensional point cloud according to the second vehicle displacement.

[0136] In an embodiment of the present application, by adjusting the shooting angle of the first camera and expanding the camera's viewing angle, effective feature points can be matched in more adjacent image frames, thereby making the constructed three-dimensional point cloud more accurate and effective. By adding a second camera on the side and expanding the range of feature point extraction and matching, effective feature points can be matched in more adjacent image frames, and the constructed three-dimensional point cloud can be effectively corrected by using the second vehicle displacement, thereby further improving the accuracy and stability of the constructed three-dimensional point cloud.

[0137] S207: Determine the size information of the vehicle according to the corrected three-dimensional point cloud.

[0138] In the embodiment of the present application, the circumscribed cube of the corrected three-dimensional point cloud is obtained, and the dimensional information of the vehicle, such as width, height and length, can be obtained.

[0139] In an embodiment of the present application, by using the first camera and the second camera to respectively capture vehicle images of the same vehicle at different angles, the range of feature point extraction and matching is expanded, so that effective feature points are matched in more adjacent image frames, and the constructed three-dimensional point cloud is more accurate and effective, thereby further completing the detection of vehicle contour dimensions efficiently, accurately and contactlessly.

[0140] It should be understood that the size of the serial numbers of the steps in the above embodiments does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application.

[0141] The embodiment of the present application further provides a computer-readable storage medium storing a computer program, wherein the computer program is executed by a processor to implement the following Figures 2 to 7 The steps of any vehicle contour dimension detection method are represented.

[0142] The embodiment of the present application further provides an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, Figures 2 to 7 The steps of any vehicle contour dimension detection method are represented.

[0143] The embodiment of the present application also provides a computer program product, which, when executed on a server, enables the server to execute the following Figures 2 to 7 The steps of any vehicle contour dimension detection method are represented.

[0144] Figure 8 Schematic diagram of an electronic device provided by an embodiment of the present application. Figure 8 As shown, the electronic device 8 of this embodiment includes: a processor 80, a memory 81, and a computer program 82 stored in the memory 81 and executable on the processor 80. When the processor 80 executes the computer program 82, the steps in the above-mentioned vehicle outline dimension detection method embodiments are implemented, such as Figure 2 Alternatively, when the processor 80 executes the computer program 82, the functions of the modules / units in the above-mentioned device embodiments are realized, for example, Figure 8 The functions of the units 81 to 84 are shown.

[0145] For example, the computer program 82 may be divided into one or more modules / units, which are stored in the memory 81 and executed by the processor 80 to implement the present application. The one or more modules / units may be a series of computer-readable instruction segments capable of performing specific functions, and the instruction segments are used to describe the execution process of the computer program 82 in the electronic device 8.

[0146] The electronic device 8 may be a mobile smart device. The electronic device 8 may include, but is not limited to, a processor 80 and a memory 81. Those skilled in the art will understand that Figure 8 It is only an example of the electronic device 8 and does not constitute a limitation of the electronic device 8. It may include more or fewer components than shown in the figure, or a combination of certain components, or different components. For example, the electronic device 8 may also include input and output devices, network access devices, buses, etc.

[0147] The processor 80 may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor may be a microprocessor or any conventional processor.

[0148] The memory 81 may be an internal storage unit of the electronic device 8, such as a hard disk or memory of the electronic device 8. The memory 81 may also be an external storage device of the electronic device 8, such as a plug-in hard disk, a smart media card (SMC), a secure digital (SD) card, a flash card, etc. equipped on the electronic device 8. Furthermore, the memory 81 may include both an internal storage unit of the electronic device 8 and an external storage device. The memory 81 is used to store the computer program and other programs and data required by the electronic device. The memory 81 may also be used to temporarily store data that has been output or is about to be output.

[0149] It should be noted that the information interaction, execution process, etc. between the above-mentioned devices / units are based on the same concept as the method embodiment of this application. Their specific functions and technical effects can be found in the method embodiment section and will not be repeated here.

[0150] Those skilled in the art can clearly understand that, for the convenience and brevity of description, only the division of the above-mentioned functional units and modules is used as an example for illustration. In actual applications, the above-mentioned functions can be distributed and completed by different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above. The functional units and modules in the embodiment can be integrated into one processing unit, or each unit can exist physically alone, or two or more units can be integrated into one unit. The above-mentioned integrated unit can be implemented in the form of hardware or in the form of software functional units. In addition, the specific names of the functional units and modules are only for the convenience of distinguishing each other, and are not used to limit the scope of protection of this application. The specific working process of the units and modules in the above-mentioned system can refer to the corresponding process in the aforementioned method embodiment, and will not be repeated here.

[0151] 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 present application implements all or part of the processes in the above-mentioned embodiment method, which can be completed by instructing the relevant hardware through a computer program. The computer program can be stored in a computer-readable storage medium, and when the computer program is executed by the processor, it can implement the steps of the above-mentioned various method embodiments. Wherein, the computer program includes computer program code, and the computer program code can be in source code form, object code form, executable file or some intermediate form. The computer-readable medium may at least include: any entity or device capable of carrying the computer program code to the device / terminal device, a recording medium, a computer memory, a read-only memory (ROM), a random access memory (RAM), an electric carrier signal, a telecommunication signal and a software distribution medium. For example, a USB flash drive, a mobile hard disk, a magnetic disk or an optical disk. In some jurisdictions, according to legislation and patent practice, a computer-readable medium cannot be an electric carrier signal or a telecommunication signal.

[0152] In the above embodiments, the description of each embodiment has its own focus. For parts that are not described or recorded in detail in a certain embodiment, reference can be made to the relevant description of other embodiments.

[0153] The above-described embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. 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 various embodiments of the present application, and should all be included in the scope of protection of the present application.

Claims

1. A vehicle outline size detection method, characterized in that: include: Acquire a first reference image frame, and a first point cloud frame and a second point cloud frame that match the first reference image frame, wherein the first reference image frame is any first vehicle image frame in a sequence of first vehicle image frames captured by a first camera; the first point cloud frame is a point cloud frame scanned by a first scanning laser rangefinder; and the second point cloud frame is a point cloud frame scanned by a second scanning laser rangefinder; the first point cloud frame includes a first vehicle body point cloud, and the second point cloud frame includes a second vehicle body point cloud; Acquire a first pixel coordinate of the first vehicle body point cloud in the first reference image frame, and a second pixel coordinate of the second vehicle body point cloud in the first reference image frame; determining a mapping area according to the first pixel coordinates and the second pixel coordinates; Performing feature extraction on the mapping area to obtain a first feature point; Acquire a first adjacent image frame sequence, where a capture time of the first adjacent image frame sequence is adjacent to a capture time of the first reference image frame, and the first adjacent image frame sequence includes a plurality of first adjacent image frames; Performing feature extraction on the first adjacent image frames to obtain a plurality of first adjacent feature points; calculating a relative pixel displacement between the first reference image frame and each of the first adjacent image frames based on the first feature point and the plurality of first adjacent feature points to obtain a plurality of relative pixel displacements; Obtaining a first point cloud distance and a focal length of the first camera, wherein the first point cloud distance refers to the distance from the vehicle body point cloud corresponding to the first feature point within the mapping area to the focal point of the first camera, and the vehicle body point cloud is the first vehicle body point cloud or the second vehicle body point cloud; determining, in sequence, the relative displacements between the first reference image frame and the plurality of first adjacent image frames thereof according to the first point cloud distance, the focal length of the first camera, and the plurality of relative pixel displacements; Determining the first vehicle displacement according to the relative displacement corresponding to each of the first reference image frames; and constructing a three-dimensional point cloud of the vehicle according to the first vehicle displacement, the first point cloud frame, and the second point cloud frame; Determine the size information of the vehicle based on the three-dimensional point cloud.

2. The vehicle outline dimension detection method according to claim 1, characterized in that: The acquiring of a first reference image frame, and a first point cloud frame and a second point cloud frame matching the first reference image frame, includes: Acquire a first vehicle image frame sequence, a first point cloud frame sequence, and a second point cloud frame sequence within the same vehicle detection period; Select any frame from the first vehicle image frame sequence as the first reference image frame, and based on the acquisition time of the first reference image frame, select a first point cloud frame with the same acquisition time as the first reference image frame from the first point cloud frame sequence, and select a second point cloud frame with the same acquisition time as the first reference image frame from the second point cloud frame sequence to complete the matching.

3. The vehicle outline dimension detection method according to claim 1, characterized in that: The calculating, based on the first feature point and the plurality of first adjacent feature points, a relative pixel displacement between the first reference image frame and each of the first adjacent image frames includes: Obtaining an image pixel coordinate difference, where the image pixel coordinate difference is a coordinate difference between a pixel coordinate of the first feature point and a pixel coordinate of the first adjacent feature point; Obtaining the pixel size of the first camera; The relative pixel displacement is determined according to the product of the image pixel coordinate difference and the pixel size.

4. The vehicle outline dimension detection method according to claim 1, characterized in that: The vehicle outline size detection method further includes: Acquire a second reference image frame, where the second reference image frame is any second vehicle image frame in a second vehicle image frame sequence captured by a second camera, and the second reference image frame matches the first point cloud frame, the second point cloud frame, and the first reference image frame; determining a second vehicle displacement of the vehicle based on each of the second reference image frames and the matched first point cloud frame and the second point cloud frame; Correcting the three-dimensional point cloud according to the second vehicle displacement; The size information of the vehicle is determined based on the corrected three-dimensional point cloud.

5. A vehicle outline dimension detection system, characterized in that: include: a first scanning laser rangefinder, a first camera, a second scanning laser rangefinder, and a data processing device; The first camera, the first scanning laser rangefinder, and the second scanning laser rangefinder are communicatively connected to the data processing device; The first scanning laser rangefinder and the second scanning laser rangefinder are used to scan the vehicle to obtain a point cloud frame; The first camera is used to photograph the vehicle and collect a first vehicle image frame sequence; The data processing device is configured to acquire a first reference image frame, and a first point cloud frame and a second point cloud frame matching the first reference image frame, wherein the first reference image frame is any first vehicle image frame in the first vehicle image frame sequence, the first point cloud frame is a point cloud frame scanned by a first scanning laser rangefinder, and the second point cloud frame is a point cloud frame scanned by a second scanning laser rangefinder; the first point cloud frame includes a first vehicle body point cloud, and the second point cloud frame includes a second vehicle body point cloud; The data processing device is further configured to obtain a first pixel coordinate of the first vehicle body point cloud in the first reference image frame, and a second pixel coordinate of the second vehicle body point cloud in the first reference image frame; and determine a mapping area based on the first pixel coordinate and the second pixel coordinate; performing feature extraction on the mapping area to obtain a first feature point; acquiring a first adjacent image frame sequence, where the acquisition time of the first adjacent image frame sequence is adjacent to the acquisition time of the first reference image frame, and the first adjacent image frame sequence includes a plurality of first adjacent image frames; performing feature extraction on the first adjacent image frames to obtain a plurality of first adjacent feature points; calculating a relative pixel displacement between the first reference image frame and each of the first adjacent image frames based on the first feature point and the plurality of first adjacent feature points to obtain a plurality of relative pixel displacements; acquiring a first point cloud distance and a focal length of the first camera, wherein the first point cloud distance refers to the distance from a vehicle body point cloud corresponding to the first feature point in the mapping area to the focus of the first camera, the vehicle body point cloud being the first vehicle body point cloud or the second vehicle body point cloud; sequentially determining a relative displacement between the first reference image frame and its plurality of first adjacent image frames based on the first point cloud distance, the focal length of the first camera, and the plurality of relative pixel displacements; determining the first vehicle displacement based on the relative displacement corresponding to each of the first reference image frames; and constructing a three-dimensional point cloud of the vehicle based on the first vehicle displacement, the first point cloud frame, and the second point cloud frame. Determine the size information of the vehicle based on the three-dimensional point cloud.

6. The vehicle outline dimension detection system according to claim 5, characterized in that: The first camera is mounted on the crossbar of the gantry. The gantry includes a crossbar and vertical poles on both sides. The gantry is mounted on the safety island of the toll station. The first scanning laser rangefinder and the second scanning laser rangefinder are respectively mounted above the vertical poles on both sides of the gantry and are mounted facing each other.

7. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the computer program, the vehicle outline dimension detection method according to any one of claims 1 to 4 is implemented.

Images