Ramp identification method and device, electronic equipment and storage medium

Abstract

The application is suitable for the field of electronic technology, and provides a ramp recognition method and device, electronic equipment and a storage medium, comprising: acquiring target trajectories of each target on a target road; dividing the target road into N road sections according to the target trajectories; wherein N is a positive integer greater than 1; for each road section, determining a target feature of the road section according to the target trajectories; the target feature comprises any one or more of a target flow, a target lane-changing feature and a target speed change feature; and determining a road ramp recognition result according to the target features of each road section. The embodiment of the application can efficiently and accurately recognize ramps on a road.

Description

TECHNICAL FIELD

[0001] The present application belongs to the technical field of electronics, and particularly relates to a ramp identification method and device, an electronic device, and a storage medium. BACKGROUND

[0002] A ramp generally refers to a small section of road that provides vehicles to enter or exit a main line (a highway, an elevated road, a bridge, a driving tunnel, etc.) and an adjacent auxiliary road. On a highway, there are usually ramp entrances and exits for vehicles to enter or exit ramps. In order to realize intelligentization of a highway, effectively guide or assist lane marking of vehicles, there is currently a demand for identifying ramp entrances and exits of a highway. SUMMARY

[0003] Therefore, the present application provides a ramp identification method, device, and electronic device to solve the problem of how to accurately identify ramp entrances and exits on a road.

[0004] A first aspect of the present application provides a ramp identification method, comprising:

[0005] obtaining target trajectories of each target on a target road;

[0006] dividing the target road into N road segments according to the target trajectories, wherein N is a positive integer greater than 1;

[0007] for each road segment, determining a target feature of the road segment according to the target trajectories; the target feature includes any one or more of a target flow, a target lane-changing feature, and a target speed change feature;

[0008] determining a road ramp entrance and exit identification result according to the target features of each road segment.

[0009] Optionally, the target flow includes a target entry quantity and a target exit quantity, and correspondingly, the determining of the target feature of each road segment according to the target trajectories includes:

[0010] detecting a target flow of each road segment according to the target trajectories to determine the target entry quantity and the target exit quantity of the road segment;

[0011] Correspondingly, the determining of the road ramp entrance and exit identification result according to the target features of each road segment includes:

[0012] for each road segment, determining a target entry-exit ratio of the road segment according to the target entry quantity and the target exit quantity of the road segment, and determining a road segment ramp entrance and exit identification result corresponding to the road segment according to the target entry-exit ratio of the road segment;

[0013] According to the ramp identification result corresponding to each of the road segments, a ramp identification result corresponding to the target road is determined.

[0014] Optionally, the ramp identification result of the road segment includes a ramp type, and the determining of the ramp identification result corresponding to the road segment according to the target in-out ratio of the road segment includes:

[0015] If the target in-out ratio of the road segment is less than a minimum threshold of a preset road target in-out ratio range, the ramp type of the road segment is determined as a ramp exit;

[0016] If the target in-out ratio of the road segment is greater than a maximum threshold of the road target in-out ratio range, the ramp type of the road segment is determined as a ramp entrance;

[0017] If the target in-out ratio of the road segment is within the road target in-out ratio range, the ramp type of the road segment is determined as no ramp.

[0018] Optionally, the ramp identification result of the road segment further includes a ramp position, and the determining of the target feature of the road segment according to each of the target trajectories for each of the road segments further includes:

[0019] For each of the trajectory segments of each of the target trajectories included in the road segment, a lane-changing feature of each of the trajectory points of the trajectory segment is determined, and / or a speed change feature of each of the trajectory points of the trajectory segment is determined;

[0020] Correspondingly, after the determining of the ramp identification result corresponding to the road segment according to the target in-out ratio of the road segment, the method further includes:

[0021] According to the ramp type corresponding to the road segment and the lane-changing feature and / or the speed change feature of each of the trajectory points corresponding to each of the trajectory segments included in the road segment, a ramp position in the road segment is determined.

[0022] Optionally, the determining of the lane-changing feature of each of the trajectory points of the trajectory segment includes:

[0023] For each of the trajectory segments, a lane-changing feature of each of the trajectory points of the trajectory segment is determined through a lane-changing feature detection step.

[0024] The lane-changing feature detection step includes, for each of the trajectory points of the trajectory segment, obtaining the lane-changing feature of the trajectory point through a trajectory point lane-changing feature determination procedure.

[0025] The trajectory point lane-changing feature determination procedure includes:

[0026] A set of adjacent trajectory points of the trajectory point is obtained.

[0027] determining a trajectory direction of the trajectory point, and a trajectory direction of each adjacent trajectory point in the set of adjacent trajectory points;

[0028] determining an average driving direction of the current road segment according to the trajectory direction of each adjacent trajectory point in the set of adjacent trajectory points;

[0029] if an included angle between the trajectory direction of the trajectory point and the average driving direction of the current road segment is greater than a preset angle threshold, determining that a lane-changing feature of the trajectory point is trajectory deviation; otherwise, determining that the lane-changing feature of the trajectory point is trajectory non-deviation.

[0030] Optionally, the determining the speed change feature of each trajectory point of the trajectory segment comprises:

[0031] for each trajectory segment, determining a speed change feature of each trajectory point of the trajectory segment through a speed change feature detection step;

[0032] The speed change feature detection step comprises: for each trajectory point of the trajectory segment, obtaining the speed change feature of the trajectory point through a trajectory point speed change feature determination procedure.

[0033] The trajectory point speed change feature determination procedure comprises:

[0034] determining a current speed of the trajectory point according to current frame detection data;

[0035] determining a speed average of the trajectory point according to preset speed detection frame number detection data;

[0036] if an absolute value of a difference between the current speed of the trajectory point and the speed average is less than a preset speed fluctuation threshold, determining that the speed change feature of the trajectory point is uniform speed; otherwise:

[0037] if the current speed of the trajectory point is greater than the speed average, determining that the speed change feature of the trajectory point is acceleration; if the current speed of the trajectory point is less than the speed average, determining that the speed change feature of the trajectory point is deceleration.

[0038] Optionally, the determining the ramp location in the road segment according to the ramp type corresponding to the road segment, and the lane-changing feature and / or the speed change feature of each trajectory point corresponding to each trajectory segment contained in the road segment comprises:

[0039] determining a sub-road segment type of a first sub-road segment of the road segment according to the ramp type corresponding to the road segment, the sub-road segment type comprising any one of a normal sub-road segment, a speed-changing sub-road segment, and a ramp;

[0040] For each trajectory segment included in the road segment, a distribution of the road segment corresponding to the trajectory segment is determined according to the lane-changing feature and the speed change feature of each trajectory point of the trajectory segment, and the sub-road segment type of the first sub-road segment of the road segment and a preset hidden Markov model; the distribution of the road segment is a distribution of the road segment estimated based on the trajectory segment, and the distribution of the road segment includes a distribution of the sub-road segment type in the road segment.

[0041] A final distribution of the road segment is determined according to the distribution of the road segment corresponding to each trajectory segment included in the road segment.

[0042] A ramp location in the road segment is determined according to the final distribution of the road segment.

[0043] A second aspect of an embodiment of the present application provides a ramp location recognition device, including:

[0044] A trajectory tracking module is configured to acquire target trajectories of targets on a target road.

[0045] A road segmentation module is configured to divide the target road into N road segments according to the target trajectories; N is a positive integer greater than 1.

[0046] A feature detection module is configured to determine target features of each road segment according to the target trajectories; the target features include any one or more of target traffic flow, target lane-changing feature and target speed change feature.

[0047] A ramp recognition module is configured to determine a road ramp location recognition result according to the target features of each road segment.

[0048] A third aspect of an embodiment of the present application provides an electronic device including 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 electronic device implements steps of the ramp location recognition method.

[0049] A fourth aspect of an embodiment of the present application provides a computer readable storage medium, the computer readable storage medium stores a computer program, when the computer program is executed by a processor, the electronic device implements steps of the ramp location recognition method.

[0050] A fifth aspect of an embodiment of the present application provides a computer program product, when the computer program product is executed on an electronic device, the electronic device executes the ramp location recognition method of any one of the first aspect.

[0051] The beneficial effects of the embodiments of the present application compared with the prior art are: in the embodiments of the present application, the target trajectories of each target on the target road are acquired, and the target road is divided into N road segments according to the target trajectories; then, for each road segment, the target feature of the road segment is determined according to the target trajectories, and the final road ramp recognition result is determined according to the target feature. Since only the target trajectories are needed, the analysis of the target features such as the target flow, the target lane changing feature, and the target speed change feature of each road segment of the road can be efficiently and accurately realized, so that the ramp recognition result of the entire road, i.e., the road ramp recognition result, can be accurately determined based on the target features of each road segment, and thus the ramp recognition can be efficiently and accurately realized without relying on complex and less stable ramp image feature analysis. BRIEF DESCRIPTION OF DRAWINGS

[0052] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings needed to be used in the embodiments or the prior art description will be briefly introduced as follows.

[0053] Figure 1 is an implementation flow diagram of a ramp recognition method provided by the embodiments of the present application;

[0054] Figure 2 is a schematic diagram of a road segment provided by the embodiments of the present application;

[0055] Figure 3 is a single-point direction schematic diagram and a trajectory offset detection schematic diagram provided by the embodiments of the present application;

[0056] Figure 4 is a state transition schematic diagram of a hidden Markov model provided by the embodiments of the present application;

[0057] Figure 5 is a schematic diagram of an observation probability matrix provided by the embodiments of the present application;

[0058] Figure 6 is a decoding schematic diagram of a road segment provided by the embodiments of the present application;

[0059] Figure 7 is a schematic diagram of a ramp recognition device provided by the embodiments of the present application;

[0060] Figure 8 is a schematic diagram of an electronic device provided by the embodiments of the present application. DETAILED DESCRIPTION

[0061] In the following description, for purposes of explanation and not limitation, specific details are set forth such as particular architectures, technologies, techniques, etc. in order to provide 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 can be practiced in other embodiments that depart from these specific details. In other instances, detailed descriptions of well-known systems, devices, circuits, and methods are omitted so as not to obscure the description of the present application with unnecessary detail.

[0062] In order to illustrate the technical solutions described in the present application, the following will be described by specific embodiments.

[0063] In order to realize the intelligentization of the expressway, effectively guide the vehicles or assist in lane line calibration, there is a need to identify the ramp of the expressway. At present, there is a method of identifying the ramp, which is based on the image collected by the vehicle-mounted camera to identify the image feature information of the ramp exit sign, the flow guide line, the anti-collision barrel or the expressway guardrail and the like, so as to realize the identification of the ramp. However, the data acquisition mode of this ramp identification method is limited, and has the disadvantages of limited detection distance, weak anti-rain, snow and fog ability, and great influence of light, that is, the accuracy of the ramp identification method is poor. Moreover, the image collected by the vehicle-mounted camera can only be limited to judging whether there is a ramp in front of the vehicle, and cannot determine the position of the ramp relative to the whole road from the global perspective, so it cannot be used to guide the automatic calibration of the expressway lane line, and has certain limitations.

[0064] In order to solve the above technical problems, the present application provides a ramp identification method and device, electronic equipment and storage medium, which comprises: acquiring target trajectories of each target on a target road; dividing the target road into N road segments according to each target trajectory; wherein N is a positive integer greater than 1; for each road segment, determining the target characteristics of the road segment according to each target trajectory; the target characteristics include any one or more of target flow, target lane changing characteristics and target speed change characteristics; determining the road ramp identification result according to the target characteristics of each road segment.

[0065] Since only the target trajectory is needed, the analysis of the target characteristics of each road segment, such as target flow, target lane changing characteristics and target speed change characteristics, can be efficiently and accurately realized, so that the target characteristics of each road segment can be used to accurately determine the ramp identification result of the whole road, that is, the road ramp identification result, and the complex and less stable ramp image feature analysis is not needed, and the ramp identification can be efficiently and accurately realized from the global perspective of the road.

[0066] Example One

[0067] Figure 1A flowchart of a ramp identification method provided by an embodiment of the present application is shown, and is described in detail as follows.

[0068] In S101, target trajectories of each target on a target road are obtained.

[0069] In an embodiment of the present application, the target road is a road currently requiring ramp identification, and can be a specified expressway. The target generally refers to a vehicle driving on the target road, and the target trajectory is the driving trajectory of the vehicle on the target road.

[0070] In an embodiment, a sensor installed on the target road can be used to continuously track each target appearing on the target road, and obtain the target trajectory of each target. The sensor can be a radar sensor or an image sensor (the image sensor of the present application only needs to track the target to determine the target trajectory, and does not need to perform complex ramp image feature recognition). In some embodiments, each frame of detection data collected by the sensor can include the identification number (ID) of the target and the target position, and the continuous tracking of the target position of each target can be realized through the information contained in each frame of detection data, so as to obtain the target trajectory of each target.

[0071] In some embodiments, the target trajectories of each target obtained can be stored in a storage module. When the number of trajectories exceeds a first preset value (for example, 80% of the preset upper limit of the number of trajectories, which can be 1000), the storage module can delete the target trajectories stored before a preset time, for example, the target trajectories before 20 minutes. Through the storage and timely deletion of the target trajectories, the storage of the target trajectories can be ensured to be sufficient for the subsequent calling of the identification steps, while the storage space can be saved, the number of target trajectories to be processed in the subsequent steps can be reduced, and the subsequent calculation amount can be reduced.

[0072] In S102, the target road is divided into N road segments according to each target trajectory; wherein N is a positive integer greater than 1.

[0073] In an embodiment of the present application, after obtaining the target trajectories of each target, the road direction of the target road can be determined according to each target trajectory. Then, according to the road direction and a preset length L (for example, 100 meters), the target road is equally divided into N road segments along a direction perpendicular to the road direction (referred to as a division direction), and the length of each road segment is the preset length L. Wherein, N is a positive integer greater than 1, and the N road segments can be denoted as: R1, R2, …, RN. N .

[0074] In an embodiment, determining the road direction of the target road according to each target trajectory can include:

[0075] When it is detected that the total number of stored target trajectories is greater than a second preset value, curve fitting is performed according to each target trajectory to obtain the road direction.

[0076] In the embodiments of the present application, the second preset value can be set according to actual needs, and the second preset value can be equal to the first preset value described above. When the total number of target trajectories is greater than the second preset value, it means that the storage amount of the current target trajectory is sufficient for accurate curve fitting, thereby ensuring the accuracy of the determined road direction. In some embodiments, a trajectory point set can be formed according to all trajectory points contained in each target trajectory; a target straight line can be obtained by performing straight line fitting on the trajectory point set based on the least square method; and the road direction of the target road can be determined according to the slope of the target straight line. In other embodiments, a target curve can be obtained by performing polynomial fitting on the trajectory point set based on a polynomial fitting method; and the road direction of the target road can be obtained by connecting the start and end points of the target curve.

[0077] In S103, for each of the road segments, a target feature of the road segment is determined according to each target trajectory; and the target feature includes any one or more of a target flow, a target lane-changing feature, and a target speed change feature.

[0078] After the N road segments are divided, for each road segment, a target feature of the road segment is obtained by performing feature analysis on the trajectory segment of each target trajectory falling on the road segment. The target feature can include any one or more of a target flow, a target lane-changing feature, and a target speed change feature. The target flow is used to indicate the target quantity entering or leaving the road segment. If the target quantity entering the road segment is large, it can be indicated that there is a large probability that the road segment has a ramp entrance. If the target quantity leaving the road segment is large, it can be indicated that there is a large probability that the road segment has a ramp exit. The target lane-changing feature is used to indicate whether the trajectory of the target is deviated. Generally, if the trajectory of the target is deviated, it can be indicated that there is a certain probability that the road segment has a ramp. The target speed change feature is used to indicate whether the speed of the target is changed. Generally, if the speed of the target is changed on the road segment, it can be indicated that there is a certain probability that the road segment has a ramp.

[0079] In S104, a road ramp identification result is determined according to the target feature of each road segment.

[0080] After the target feature of each road segment is determined, for each road segment, a ramp identification result of each road segment can be determined according to the target feature of the road segment. Then, the ramp identification results of the road segments are summarized to obtain a complete road ramp identification result of the target road. The road ramp identification result can include any one or more of the type, position, and number of the ramps existing on the target road.

[0081] In the embodiments of the present application, the target trajectories of each target on the target road are acquired, and the target road is divided into N road segments according to the target trajectories. Then, for each road segment, the target feature of the road segment is determined according to the target trajectories, and the final road ramp recognition result is determined according to the target feature. Since the target flow, target lane-changing feature, target speed change feature and other target features of each road segment of the road can be efficiently and accurately analyzed based on the target trajectories, the target features of each road segment can be used to accurately determine the ramp recognition result of the entire road, i.e., the road ramp recognition result, so that the ramp recognition can be efficiently and accurately realized without relying on complex and less stable ramp image feature analysis.

[0082] Optionally, the target flow includes a target entering quantity and a target leaving quantity, and correspondingly, the determining, for each road segment, the target feature of the road segment according to the target trajectories includes:

[0083] target flow detection is performed on each road segment according to the target trajectories to determine the target entering quantity and the target leaving quantity of the road segment;

[0084] Correspondingly, the determining the road ramp recognition result according to the target features of each road segment includes:

[0085] For each road segment, the target in-out ratio of the road segment is determined according to the target entering quantity and the target leaving quantity of the road segment, and the road segment ramp recognition result corresponding to the road segment is determined according to the target in-out ratio of the road segment;

[0086] The road ramp recognition result corresponding to the target road is determined according to the road segment ramp recognition results corresponding to each road segment.

[0087] In the embodiments of the present application, the target entering quantity is the number of targets entering the road segment, and the target leaving quantity is the number of targets leaving the road segment. For each road segment, the target entering quantity and the target leaving quantity of the road segment can be determined by counting the targets passing through a specified cross section of the road segment. The specified cross section can include a starting road surface and / or an ending road surface in the road segment, where the starting road surface is a road surface range within a preset length d from the starting position of the road segment, and the ending road surface is a road surface range within a preset length d from the ending position of the road segment, as shown in FIG. 1. Figure 2 d is less than the length L of the road segment, for example, when L = 100 meters, d can be 10 meters.

[0088] In one embodiment, for each designated section of each road segment, the number of targets falling into that designated section is cumulatively counted based on each frame of detection data collected by the sensors. This count begins after the ramp entrance identification device is installed and is independent of the target trajectories stored in the storage module; the number does not decrease as target trajectories in the storage module disappear. Subsequently, for any road segment R... i It can be used to prevent people from falling into this section of road R. i The target number of the starting road surface IN i (where i = 1, 2, ..., N) represents the target entry quantity for this road segment; when the road segment R i If it is not the last segment of the target road, you can land on segment R. i The next segment R i+1 The number of targets on the starting road surface is used as the number of people leaving R. i Target departure number OUT for the road segment i (where i = 1, 2, ..., N-1), and for the last road segment R... N This section of road R can be... N The target quantity counted at the end of the road surface is used as the R of that road segment. N Target number of people leaving OUT N .

[0089] Correspondingly, in step S104 above, for each road segment, the target entry number IN for that road segment can be calculated based on the statistics. i and the number of people leaving the target (OUT) i Determine the inbound / outbound ratio of this road segment: OUT i / IN i Then, based on the entry / exit ratio of the road segment, the ramp identification result for that road segment is determined. The ramp identification result may include information indicating whether a ramp exists on the road segment, and may also include information about the type of ramp on the road segment.

[0090] After determining the ramp identification results for each road segment, information can be summarized to obtain which road segment on the target road has ramps and the types of ramps present on that road segment, thus obtaining the ramp identification results for the entire target road. In one embodiment, the ramp identification results may include the identification information of the road segment with ramps and the ramp type of that road segment.

[0091] In the embodiments of the present application, the target entering quantity and the target leaving quantity of each road section on the target road are accurately counted, the target in-out ratio of each road section is determined, the target in-out ratio is used to accurately determine whether the ramp exists on the target road, and the road section ramp identification result of each road section is accurately obtained. Then, the road ramp identification result of the target road is accurately determined according to the road section ramp identification result of each road section.

[0092] Optionally, the road section ramp identification result includes a ramp type, and the road section ramp identification result corresponding to the road section point is determined according to the target in-out ratio of the road section, including:

[0093] The ramp type of the road section is determined according to the target in-out ratio of the road section and a preset road target in-out ratio range.

[0094] In the embodiments of the present application, the preset road target in-out ratio range is the normal vehicle in-out ratio when the road section does not have a ramp, for example, 0.95-1.05. For each road section, the ramp type of the road section can be determined according to the comparison result of the target in-out ratio of the road section and the road target in-out ratio range. In one embodiment, the ramp type of the road section includes two types, i.e., “ramp exists” and “no ramp”. When the target in-out ratio of the road section is within the preset road target in-out ratio range, the road section is determined to not have a ramp, and the road section ramp identification result can be “no ramp”. When the target in-out ratio of the road section is outside the preset road target in-out ratio range, the road section is determined to have a ramp, and the road section ramp identification result can be “ramp exists”. In another embodiment, the ramp type of the road section includes three types, i.e., ramp exit, ramp entrance and no ramp. When the road section has a ramp, the ramp identification result specifically distinguishes the type of the ramp in the road section through “ramp exit” or “ramp entrance”.

[0095] In the embodiments of the present application, the ramp type of each road section is accurately determined based on the target in-out ratio of the road section with reference to the preset road target in-out ratio range, and thus the ramp identification can be accurately implemented.

[0096] Optionally, the ramp type of the road section is determined according to the target in-out ratio of the road section and the preset road target in-out ratio range, including:

[0097] If the target in-out ratio of the road section is less than the minimum threshold value of the road target in-out ratio range, the ramp type of the road section is determined to be a ramp exit.

[0098] If the target in-out ratio of the road section is greater than the maximum threshold value of the road target in-out ratio range, the ramp type of the road section is determined to be a ramp exit.

[0099] If the target ramp access ratio of the road section is within the range of the road target access ratio, the ramp type of the road section is determined as no ramp.

[0100] In the embodiments of the present application, the ramp type specifically includes three types of ramp exit, ramp entrance and no ramp.

[0101] After determining the target access ratio of the road section, the target access ratio is compared with the minimum threshold (for example, 0.95 described above) of the preset range of the road target access ratio. If the target access ratio is less than the minimum threshold, the target exit quantity of the current road section is less than the target entry quantity, that is, the quantity of targets leaving the road section from the end of the road section is less, which indicates that the road section may have a ramp exit, resulting in that part of the targets entering the road section are leaving the road section through the ramp exit. At this time, the ramp type of the road section is determined as ramp exit.

[0102] In addition, the target access ratio is compared with the maximum threshold (for example, 1.05 described above) of the preset range of the road target access ratio. If the target access ratio is greater than the maximum threshold, the target exit quantity of the current road section is greater than the target entry quantity, that is, in addition to the targets entering the road section from the starting pavement of the road section, the road section may have a ramp entrance, so that part of the targets leaving the road section are entering through the ramp entrance. At this time, the ramp type of the road section is determined as ramp entrance.

[0103] If the target access ratio of the road section is greater than the minimum threshold described above and less than the maximum threshold described above, it indicates that the target access ratio of the current road section falls within the range of the road target access ratio described above, that is, the target entry and exit of the road section is consistent with the target entry and exit of the normal road section. At this time, the ramp type of the road section is determined as no ramp.

[0104] In the embodiments of the present application, by comparing the target access ratio of the road section with the minimum threshold and the maximum threshold of the range of the road target access ratio, the three different ramp types of ramp exit, ramp entrance and no ramp can be accurately determined, so that the accuracy of the ramp identification can be improved.

[0105] Optionally, the road section ramp identification result further includes a ramp position, and the target feature of each road section is determined according to each target trajectory, and further includes:

[0106] For each trajectory segment of each target trajectory included in each road section, the lane change feature of each trajectory point of the trajectory segment is determined, and / or the speed change feature of each trajectory point of the trajectory segment is determined.

[0107] Correspondingly, after the identification result of the ramp of the road section is determined according to the target in-out ratio of the road section, the method further comprises:

[0108] According to the ramp type of the road section and the lane-changing feature and / or speed change feature of each trajectory point corresponding to each trajectory segment contained in the road section, the position of the ramp in the road section is determined.

[0109] In the embodiments of the present application, for each trajectory segment in the target trajectory contained in each road section, lane-changing detection is performed on the trajectory segment to determine the lane-changing feature of each trajectory point, which can include trajectory deviation or no trajectory deviation. In addition, speed change detection is performed on each trajectory segment in the target trajectory contained in each road section, and according to the positions of each trajectory point recorded between frames, the speed change feature of each trajectory point is determined, which can include acceleration, deceleration and constant speed.

[0110] After the ramp type of the road section is determined to be a ramp according to the target in-out ratio of the road section, or the ramp type is directly determined to be a ramp exit or a ramp entrance, according to the ramp type, the lane-changing feature and / or speed change feature of each trajectory point corresponding to each trajectory segment contained in the road section, the probability of the existence of the ramp at each position of the road section is evaluated, so as to determine the position of the ramp in the road section.

[0111] In the embodiments of the present application, since the ramp type of the road section, the lane-changing feature and / or speed change feature of the trajectory points of each trajectory segment in the road section are combined, the position of the ramp in the road section can be further accurately determined, and thus the accuracy of the ramp identification method can be improved.

[0112] Optionally, the determination of the lane-changing feature of each trajectory point of the trajectory segment comprises:

[0113] For each trajectory segment, the lane-changing feature of each trajectory point of the trajectory segment is determined through a lane-changing feature detection step.

[0114] The lane-changing feature detection step comprises:

[0115] For each trajectory point of the trajectory segment, the lane-changing feature of the trajectory point is obtained through a trajectory point lane-changing feature determination process.

[0116] The trajectory point lane-changing feature determination process comprises:

[0117] A1: Obtain a set of adjacent trajectory points of the trajectory point;

[0118] A2: Determine the trajectory direction of the trajectory point and the trajectory direction of each adjacent trajectory point in the set of adjacent trajectory points;

[0119] A3: determining an average driving direction of the current road segment according to the trajectory directions of each adjacent trajectory point in the adjacent trajectory point set;

[0120] A4: if the included angle between the trajectory direction of the trajectory point and the average driving direction of the current road segment is greater than a preset angle threshold, determining that the trajectory feature of the trajectory point is trajectory deviation; otherwise, determining that the trajectory feature of the trajectory point is trajectory non-deviation.

[0121] In the embodiments of the present application, for each trajectory segment included in a road segment, the trajectory feature of each trajectory point in the trajectory segment is determined through the trajectory feature detection step. For a determined trajectory segment, the trajectory feature detection step for the trajectory segment includes: for each trajectory point of the trajectory segment, the trajectory feature of the trajectory point is determined according to the trajectory point trajectory feature determination process. For a trajectory point in a determined trajectory segment, the corresponding trajectory point trajectory feature determination process includes steps A1 to A4.

[0122] In step A1, the adjacent trajectory point set of the trajectory point is obtained. The adjacent trajectory point set is a set of adjacent points determined according to an adjacent box centered on the current trajectory point. The shape of the adjacent box can be circular or rectangular. For example, the adjacent box can be a circular box with a radius of about two meters, or a square box with a side length of 3. In an embodiment, according to the position of the current trajectory point and the preset adjacent box, a set of adjacent trajectory points contained in the adjacent box is determined as an initial adjacent point set; then, for the initial adjacent point set, if there are different adjacent points from the same target trajectory, only the adjacent point closest to the current trajectory point in the target trajectory is retained as an adjacent trajectory point, thereby obtaining the final adjacent trajectory point set.

[0123] In step A2, the trajectory direction of the current trajectory point and the trajectory directions of each adjacent trajectory point in the adjacent trajectory point set are determined by a trajectory point direction determination method. For a single trajectory point, the trajectory point direction determination method includes: determining the position information of the trajectory point in the current frame detection data and the position information of the trajectory point in the N frame1 frame detection data after the trajectory point according to a preset trajectory direction discrimination frame number N frame1 , determining the position vector difference of the trajectory point, and taking the unit vector of the same direction of the position vector difference as the trajectory direction of the trajectory point. The trajectory direction is shown in the single point direction diagram in Figure 3 .

[0124] In step A3, after the trajectory directions of each adjacent trajectory point in the adjacent trajectory point set are determined, the vector average of the trajectory directions of each adjacent trajectory point is calculated to obtain the average driving direction of the current road segment.

[0125] In step A4, an included angle between a trajectory direction of the current trajectory point and the average driving direction of the current road segment determined in step A3 is calculated. If the included angle is greater than a preset angle threshold, it indicates that the current trajectory point has deviated from the normal driving direction of the road segment, and thus the trajectory shift feature of the trajectory point is determined as: trajectory shift. Exemplarily, Figure 3 A trajectory shift detection schematic diagram provides an example of trajectory shift. Conversely, if the included angle is less than or equal to the preset angle threshold, it indicates that the current trajectory point has not deviated from the normal driving direction of the road segment, and the trajectory shift feature of the trajectory point is determined as: no trajectory shift.

[0126] In the embodiments of the present application, the trajectory shift feature of each trajectory point in each trajectory segment can be accurately determined based on the trajectory shift feature detection step and the trajectory point trajectory shift feature determination process, so as to improve the accuracy of subsequent determination of ramp location based on the trajectory shift feature, and thus the accuracy of ramp recognition can be improved.

[0127] Optionally, the determination of the speed change feature of each trajectory point of the trajectory segment comprises:

[0128] For each trajectory segment, the speed change feature of each trajectory point of the trajectory segment is determined through a speed change feature detection step;

[0129] The speed change feature detection step comprises: for each trajectory point of the trajectory segment, the speed change feature of the trajectory point is obtained through a trajectory point speed change feature determination process.

[0130] The trajectory point speed change feature determination process comprises:

[0131] B1: determining the current speed of the trajectory point according to the current frame detection data;

[0132] B2: determining the speed average of the trajectory point according to the detection data of a preset speed detection frame number;

[0133] B3: if the absolute value of the difference between the current speed of the trajectory point and the speed average is less than a preset speed fluctuation threshold, the speed change feature of the trajectory point is determined as: uniform speed; otherwise:

[0134] B4: if the current speed of the trajectory point is greater than the speed average, the speed change feature of the trajectory point is determined as: acceleration; if the current speed of the trajectory point is less than the speed average, the speed change feature of the trajectory point is determined as: deceleration.

[0135] In this embodiment of the application, for each trajectory segment included in the road segment, the speed change characteristics of each trajectory point in the trajectory segment are determined through a speed change feature detection step. For a determined trajectory segment, the speed change feature detection step for that trajectory segment includes: for each trajectory point in the trajectory segment, detection is performed according to the trajectory point speed change feature determination process to obtain the speed change characteristics of that trajectory point. For a trajectory point in a determined trajectory segment, the corresponding trajectory point speed change feature determination process includes steps B1 to B4.

[0136] In step B1, the current velocity of the current trajectory point is determined based on the velocity information contained in the current frame detection data. This current frame detection data can be the currently acquired frame of radar data.

[0137] In step B2, the number of frames N is detected according to the preset speed. frame2 Obtain N data preceding the current frame's detection data. frame2 Frame detection data, calculate N frame2 The average velocity of the trajectory point in the frame detection data is used to obtain the average velocity of the trajectory point.

[0138] In step B3, the difference between the current velocity of the trajectory point determined in step B1 and the average velocity of the trajectory point determined in step B2 is calculated. If the absolute value of this difference is less than a preset velocity fluctuation threshold, it can be determined that the velocity of the trajectory point has not changed significantly, and therefore the velocity change characteristic of the trajectory point is determined to be uniform.

[0139] In step B4, if the absolute value of the difference determined in step B3 is greater than or equal to the preset speed fluctuation threshold, it is determined that the speed of the trajectory point has changed significantly. Specifically, when the current speed of the trajectory point is greater than the average speed, the speed change characteristic of the trajectory point is determined to be acceleration. When the current speed of the trajectory point is less than the average speed, the speed change characteristic of the trajectory point is determined to be deceleration.

[0140] In this embodiment of the application, since the speed change characteristics of each trajectory point in each trajectory segment can be accurately determined based on the speed change feature detection step and the trajectory point speed change feature determination process, the accuracy of subsequent determination of the ramp entrance position based on the speed change characteristics can be improved, thus improving the accuracy of ramp entrance identification.

[0141] Optionally, determining the location of the ramps in the road segment based on the ramp type corresponding to the road segment, and the lane-changing characteristics and / or speed change characteristics of each trajectory point corresponding to each trajectory segment included in the road segment, includes:

[0142] determine a sub-road section type of a first sub-road section of the road section according to the ramp type corresponding to the road section, the sub-road section type including any one of a normal sub-road section, a speed-changing sub-road section, and a ramp;

[0143] for each trajectory section included in the road section, determine a predicted road section distribution corresponding to the trajectory section according to a lane-changing feature and a speed-changing feature of each trajectory point of the trajectory section, the sub-road section type of the first sub-road section of the road section, and a preset hidden Markov model, the predicted road section distribution being a road section distribution of the road section predicted according to the trajectory section, the road section distribution including a distribution of the sub-road section type in the road section;

[0144] determine a final road section distribution of the road section according to the predicted road section distribution corresponding to each trajectory section included in the road section;

[0145] determine a ramp position in the road section according to the final road section distribution of the road section.

[0146] In the embodiments of the present application, for a road section with a ramp, a sub-road section type of a first sub-road section of the road section can be determined according to a ramp type corresponding to the road section. The sub-road section type can be any one of a normal sub-road section, a speed-changing sub-road section (deceleration sub-road section or acceleration sub-road section), and a ramp (ramp exit or ramp entrance). The normal sub-road section is a sub-road section in which a vehicle travels normally, the speed-changing sub-road section is a sub-road section in which a vehicle is required to change speed, and the ramp is a position at which a vehicle enters or exits a ramp. In an embodiment, when the ramp type corresponding to the road section is a ramp exit, the sub-road section type of the first sub-road section can be a normal sub-road section by default. When the ramp type corresponding to the road section is a ramp entrance, a probability that the sub-road section type of the first sub-road section is a ramp is m / L (where m is a length of the first sub-road section, which can be equal to the length d of the specified section, and L is a total length of the road), and a probability that the sub-road section type of the first sub-road section is a normal sub-road section is 1-(m / L).

[0147] After determining the type of the first sub-road section of the road section, for each trajectory section included in the road section, the lane-changing feature and the speed-changing feature of each trajectory point of the trajectory section are taken as the observation state, and the sub-road section type of the first sub-road section of the trajectory section is taken as the initial state, and a preset Hidden Markov Model is used to determine the estimated road section distribution corresponding to the trajectory section, i.e., the distribution of the sub-road section type of the current road section estimated based on the trajectory section. The Hidden Markov Model (HMM) is a statistical model used to describe a Markov process with hidden unknown parameters. The difficulty is to determine the hidden parameters of the process from the observable parameters, and then use these parameters for further analysis. In this application, the hidden unknown parameters are the sub-road section types of each sub-road section in the road section.

[0148] After determining the estimated road section distribution corresponding to each trajectory section included in the current road section, one of the estimated road section distributions with the highest occurrence probability is determined as the final road section distribution.

[0149] After determining the final road section distribution of the road section, the ramp position in the road section is determined according to the road section distribution. In one embodiment, the position of the sub-road section with the sub-road section type of the ramp (ramp entrance or ramp exit) in the road section is determined as the ramp position according to the road section distribution. In another embodiment, if no sub-road section with the sub-road section type of the ramp can be found according to the final road section distribution, the current detection is marked as a failure, and the ramp detection for the road section is performed again.

[0150] In one embodiment, if the ramp type corresponding to the road section is a ramp exit, the hidden state set in the preset Hidden Markov Model is H = {h1, h2, h3}, where h1 represents a normal sub-road section, h2 represents a deceleration road section, and h3 represents a ramp exit; the observation state set is S = {S1 = [uniform speed, no trajectory deviation], S2 = [deceleration, no trajectory deviation], S3 = [uniform speed, trajectory deviation], S4 = [deceleration, trajectory deviation]}. In addition, the Hidden Markov Model has a corresponding state transition matrix A = {aij | i = 1, 2, 3, j = 1, 2, 3}, where aij represents the probability of the hidden state transitioning from hi to hj; and has a corresponding observation probability matrix B = {bij | i = 1, 2, 3, j = 1, 2, 3, 4}, where bij represents the probability of the observation state being Sj when the hidden state is hi. The two matrices can be obtained according to actual experiments or theoretical analysis. Exemplarily, Figure 4 A state transition diagram of a Hidden Markov Model of an embodiment of the application is provided, Figure 5A schematic diagram of an observation probability matrix is provided. For each trajectory segment of a current road segment, an observation state sequence is formed based on the speed variation characteristics and the lane change characteristics of each trajectory point included in the trajectory segment, an initial state is set as the sub-road segment type of the first sub-road segment of the road segment, and the hidden Markov model is input. The hidden Markov model decodes the trajectory segment based on the initial state, the observation sequence state, and a preset state transition matrix A and observation probability transition matrix B by using a Viterbi algorithm to obtain a predicted road segment distribution corresponding to the trajectory segment. Exemplarily, Figure 7 A decoding schematic diagram of a road segment is provided, in which the decoded road segment distribution with the highest probability is the predicted road segment distribution corresponding to the trajectory segment.

[0151] In another embodiment, if the ramp type corresponding to the road segment is a ramp entrance, the hidden state set in the preset hidden Markov model is H = {h1, h2, h3}, in which h1 represents a normal sub-road segment, h2 represents an acceleration road segment, and h3 represents a ramp entrance; the observation state set is S = {S1 = [uniform speed, no trajectory deviation], S2 = [acceleration, no trajectory deviation], S3 = [uniform speed, trajectory deviation], S4 = [acceleration, trajectory deviation]}. In addition, the hidden Markov model has a corresponding state transition matrix A = {aij | i = 1, 2, 3, j = 1, 2, 3}, in which aij represents the probability of transition from hidden state hi to hj; and a corresponding observation probability matrix B = {bij | i = 1, 2, 3, j = 1, 2, 3, 4}, in which bij represents the probability of observing state Sj when the hidden state is hi. The two matrices can be obtained based on actual experimental statistics or theoretical analysis. Moreover, since the ramp type of the road segment is a ramp entrance, the sub-road segment type of the first sub-road segment of the road segment is set as a normal sub-road segment with a probability of 1-m / L and as a ramp entrance with a probability of m / L. Subsequently, for each trajectory segment of a current road segment, an observation state sequence is formed based on the speed variation characteristics and the lane change characteristics of each trajectory point included in the trajectory segment, an initial state is set as the sub-road segment type of the first sub-road segment of the road segment, and the hidden Markov model is input. The hidden Markov model decodes the trajectory segment based on the initial state, the observation sequence state, and a preset state transition matrix A and observation probability transition matrix B by using a Viterbi algorithm to obtain a predicted road segment distribution corresponding to the trajectory segment.

[0152] In the embodiments of the present application, for each road segment, the predicted road segment distribution can be determined based on the ramp type, the lane change characteristics and the speed variation characteristics of each trajectory point of the trajectory segment by using the preset hidden Markov model, and the actual road segment distribution of the road segment can be accurately determined, so that the position of the ramp in the road segment can be accurately located, and the accuracy of ramp recognition can be improved.

[0153] In some embodiments, after determining the ramp type and location of each road segment, the ramp location and type of each ramp on the target road can be output as the road ramp identification result. For example, the ramp location can be represented as (x1, x2, y1, y2), where x1, x2, y1, and y2 are the coordinates of the four vertices of the ramp. For example, the ramp type can be ENTRANCE (representing ramp entrance) or EXIT (representing ramp exit).

[0154] It should be understood that the sequence number of each step in the above embodiments does not imply 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.

[0155] Example Two

[0156] Figure 7 The diagram shows a schematic representation of a ramp entrance recognition device according to an embodiment of this application. For ease of explanation, only the parts relevant to the embodiment of this application are shown:

[0157] The ramp identification device includes: a trajectory tracking module 71, a road segmentation module 72, a feature detection module 73, and a ramp identification module 74. Among them:

[0158] The trajectory tracking module 71 is used to acquire the target trajectory of each target on the target road.

[0159] The road segmentation module 72 is used to divide the target road into N road segments according to each target trajectory; where N is a positive integer greater than 1.

[0160] The feature detection module 73 is used to determine the target features of each road segment based on each target trajectory; the target features include any one or more of the target flow, target lane change features, and target speed change features.

[0161] The ramp identification module 74 is used to determine the road ramp identification result based on the target features of each road segment.

[0162] In some embodiments, the ramp identification device further includes:

[0163] The storage module is used to store the trajectories of each target.

[0164] In some embodiments, the feature detection module may specifically include a flow detection module for detecting target flow, a lane change detection module for detecting target lane change features, and a speed change detection module for detecting target speed change features.

[0165] Optionally, the target flow includes a target entering quantity and a target leaving quantity, and the feature detection module 73 is specifically configured to detect the target flow of each road segment according to the target trajectories, and determine the target entering quantity and the target leaving quantity of the road segment.

[0166] Correspondingly, the ramp identification module 74 includes:

[0167] A road segment ramp identification result determination unit is configured to determine a target in-out ratio of each road segment according to the target entering quantity and the target leaving quantity of the road segment, and determine a road segment ramp identification result corresponding to the road segment according to the target in-out ratio of the road segment.

[0168] A road ramp identification result determination unit is configured to determine a road ramp identification result corresponding to the target road according to the road segment ramp identification results corresponding to each road segment.

[0169] Optionally, the road segment ramp identification result includes a ramp type, and in the road segment ramp identification result determination unit, the road segment ramp identification result corresponding to the road segment is determined according to the target in-out ratio of the road segment, including:

[0170] If the target in-out ratio of the road segment is less than a minimum threshold of a preset road target in-out ratio range, it is determined that the ramp type of the road segment is a ramp exit.

[0171] If the target in-out ratio of the road segment is greater than a maximum threshold of the road target in-out ratio range, it is determined that the ramp type of the road segment is a ramp entrance.

[0172] If the target in-out ratio of the road segment is within the road target in-out ratio range, it is determined that the ramp type of the road segment is no ramp.

[0173] Optionally, the road segment ramp identification result further includes a ramp location, and the feature detection module 73 is further configured to determine a lane change feature of each trajectory point of each trajectory segment of each target trajectory included in each road segment, and / or determine a speed change feature of each trajectory point of each trajectory segment.

[0174] Correspondingly, the road segment ramp identification result determination unit is further configured to determine a ramp location in the road segment according to the ramp type corresponding to the road segment and the lane change feature and / or the speed change feature of each trajectory point corresponding to each trajectory segment included in the road segment.

[0175] Optionally, in the feature detection module 73, the lane change feature of each trajectory of the trajectory point includes:

[0176] For each of the trajectory segments, a lane-changing feature of each trajectory point of the trajectory segment is determined by a lane-changing feature detection step.

[0177] The lane-changing feature detection step comprises: for each trajectory point of the trajectory segment, a lane-changing feature of the trajectory point is determined by a trajectory point lane-changing feature determination procedure.

[0178] The trajectory point lane-changing feature determination procedure comprises:

[0179] A set of neighboring trajectory points of the trajectory point is obtained.

[0180] A trajectory direction of the trajectory point and a trajectory direction of each neighboring trajectory point in the set of neighboring trajectory points are determined.

[0181] An average driving direction of the current road segment is determined according to the trajectory direction of each neighboring trajectory point in the set of neighboring trajectory points.

[0182] If an angle between the trajectory direction of the trajectory point and the average driving direction of the current road segment is greater than a preset angle threshold, it is determined that the lane-changing feature of the trajectory point is trajectory deviation; otherwise, it is determined that the lane-changing feature of the trajectory point is trajectory non-deviation.

[0183] Optionally, in the feature detection module 73, the determination of the speed change feature of each trajectory point of the trajectory point comprises:

[0184] For each of the trajectory segments, a speed change feature of each trajectory point of the trajectory segment is determined by a speed change feature detection step.

[0185] The speed change feature detection step comprises: for each trajectory point of the trajectory segment, a speed change feature of the trajectory point is determined by a trajectory point speed change feature determination procedure.

[0186] The trajectory point speed change feature determination procedure comprises:

[0187] A current speed of the trajectory point is determined according to current frame detection data.

[0188] A speed average of the trajectory point is determined according to preset speed detection frame number detection data.

[0189] If an absolute value of a difference between the current speed of the trajectory point and the speed average is less than a preset speed fluctuation threshold, it is determined that the speed change feature of the trajectory point is uniform speed; otherwise:

[0190] If the current speed of the trajectory point is greater than the average speed, it is determined that the speed change feature of the trajectory point is acceleration; if the current speed of the trajectory point is less than the average speed, it is determined that the speed change feature of the trajectory point is deceleration.

[0191] Optionally, in the ramp portal recognition result determination unit, the ramp portal position in the road section is determined according to the ramp portal type corresponding to the road section, and the lane change feature and / or the speed change feature of each trajectory point corresponding to each trajectory segment contained in the road section, including:

[0192] According to the ramp portal type corresponding to the road section, the sub-road section type of the first sub-road section of the road section is determined, and the sub-road section type includes any one of a normal sub-road section, a speed-changing sub-road section, and a ramp portal.

[0193] For each trajectory segment contained in the road section, according to the lane change feature and the speed change feature of each trajectory point of the trajectory segment, and the sub-road section type of the first sub-road section of the road section and the preset hidden Markov model, a predicted road section distribution condition corresponding to the trajectory segment is determined; the predicted road section distribution condition is a road section distribution condition of the road section predicted based on the trajectory segment, and the road section distribution condition includes a distribution condition of the sub-road section type in the road section.

[0194] According to the predicted road section distribution condition corresponding to each trajectory segment contained in the road section, a final road section distribution condition of the road section is determined.

[0195] According to the final road section distribution condition of the road section, a ramp portal position in the road section is determined.

[0196] It should be noted that the information interaction, execution process and the like between the above apparatuses / units are based on the same concept as the method embodiments of the present application, and the specific functions and technical effects brought by the same can be referred to the method embodiments part, which will not be described here.

[0197] Example Three

[0198] Figure 8 is a schematic diagram of an electronic device provided by an embodiment of the present application. As shown in Figure 8 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, such as a ramp portal recognition program. The processor 80 implements the steps in each of the above ramp portal recognition method embodiments when executing the computer program 82, such as Figure 1The steps S101-S104 are shown. Alternatively, the processor 80 implements the functions of the modules / units in each of the above apparatus embodiments when executing the computer program 82, for example Figure 7 The functions of the trajectory tracking module 71 to the ramp identification module 74 are shown.

[0199] For example, the computer program 82 can be divided into one or more modules / units, which are stored in the memory 81 and executed by the processor 80 to complete the present application. The one or more modules / units can be a series of computer program instruction segments capable of completing a specific function, which are used to describe the execution process of the computer program 82 in the electronic device 8.

[0200] The electronic device 8 can be a desktop computer, a notebook, a palm computer, and a cloud server, etc. The electronic device can include, but is not limited to, the processor 80, the memory 81. Those skilled in the art can understand that Figure 8 The electronic device 8 is only an example and does not constitute a limitation on the electronic device 8, and can include more or fewer components than shown, or combine certain components, or different components, for example, the electronic device can also include an input / output device, a network access device, a bus, etc.

[0201] The processor 80 can be a central processing unit (CPU), and can also be other general-purpose processors, digital signal processors (DSP), application specific integrated circuits (ASIC), field programmable gate arrays (FPGA) or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components, etc. The general-purpose processor can be a microprocessor or the processor can also be any conventional processor.

[0202] The memory 81 can be an internal storage unit of the electronic device 8, for example, a hard disk or a memory of the electronic device 8. The memory 81 can also be an external storage device of the electronic device 8, for example, 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. Further, the memory 81 can also include both the internal storage unit and the external storage device of the electronic device 8. The memory 81 is used to store the computer program and other programs and data required by the electronic device. The memory 81 can also be used to temporarily store data that has been output or will be output.

[0203] Those skilled in the art can clearly understand that, for the convenience and brevity of description, only the division of the above functional units and modules is exemplified, and in actual application, the above functions can be completed by different functional units and modules according to needs, that is, the internal structure of the apparatus is divided into different functional units or modules to complete all or part of the functions described above. Each functional unit and module in the embodiment can be integrated in one processing unit, or each unit can be physically present separately, or two or more units can be integrated in one unit. The integrated unit can be realized in the form of hardware or in the form of software functional unit. In addition, the specific names of each functional unit and module are only for easy distinction, and do not limit the protection scope of the application. The specific working process of the units and modules in the above system can refer to the corresponding process in the foregoing method embodiments, which will not be described here.

[0204] In the above embodiments, the description of each embodiment has its own emphasis, and the parts not described or recorded in detail in a certain embodiment can be referred to the relevant description of other embodiments.

[0205] Those of ordinary skill in the art can appreciate that the units and algorithm steps of each example described in combination with the embodiments disclosed herein can be realized by electronic hardware or a combination of computer software and electronic hardware. Whether the functions are executed in hardware or software depends on the specific application and design constraints of the technical solution. A person skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of the application.

[0206] In the embodiments of the present application, it should be understood that the disclosed apparatus / equipment and method can be implemented by other manners. For example, the apparatus / equipment embodiments described above are merely illustrative, for example, the division of the modules or units is merely a logical function division, and actual implementation can have another division manner, for example, a plurality of units or components can be combined or integrated into another system, or some features can be ignored or not executed. In addition, the coupling or direct coupling or communication connection between the units or components shown or discussed can be indirect coupling or communication connection through some interfaces, apparatuses or units, which can be electrical, mechanical or other forms.

[0207] The units described as separate components can or can not be physically separate, and the components shown as units can or can not be physical units, i.e. can be located in one place or distributed on a plurality of network units. Part or all of the units can be selected to achieve the purpose of the embodiment according to actual needs.

[0208] In addition, the functional units in each embodiment of the present application can be integrated in one processing unit, or each unit can be physically present separately, or two or more units can be integrated in one unit. The integrated unit can be realized in the form of hardware or in the form of a software functional unit.

[0209] The integrated module / unit, if realized in the form of a software functional unit and sold or used as an independent product, can be stored in a computer readable storage medium. Based on such understanding, all or part of the processes in the above-mentioned embodiment methods can also be completed by a computer program instructing related hardware, and the computer program can be stored in a computer readable storage medium. When the processor executes the computer program, the steps of each method embodiment described above can be implemented. The computer program includes computer program code, which can be in the form of source code, object code, executable file or some intermediate form. The computer readable medium can include any entity or device capable of carrying the computer program code, recording medium, U disk, mobile hard disk, magnetic disk, optical disk, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signal, telecommunication signal and software distribution medium, etc. It should be noted that the computer readable medium can include or exclude contents according to the requirements of legislation and patent practice in the jurisdiction, for example, in some jurisdictions, according to legislation and patent practice, the computer readable medium does not include electrical carrier signals and telecommunication signals.

[0210] The above-described embodiments are only used to illustrate the technical solutions of the present application, but not limit the present application; although the present application is described in detail with reference to the foregoing embodiments, those skilled in the art should understand that: it can still modify the technical solutions recorded in the foregoing embodiments, or make equivalent replacement for part of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present application, and should be included in the protection scope of the present application.

Claims

1. A ramp identification method characterized by, The method comprises: acquiring target trajectories of targets on a target road; in a case where a number of the target trajectories is greater than a preset trajectory number, determining a road direction of the target road according to the target trajectories; dividing the target road into N road segments along a direction perpendicular to the road direction according to the target trajectories, wherein N is a positive integer greater than 1; for each of the road segments, determining target features of the road segment according to the target trajectories, wherein the target features comprise any one or more of target flow, target lane-changing feature and target speed change feature, and the target flow is used to represent a target number entering or leaving the road segment; determining a road ramp recognition result according to the target features of the road segments.

2. The ramp identification method of claim 1, wherein, The target flow comprises target entering number and target leaving number, and correspondingly, the determining the target features of the road segment according to the target trajectories comprises: detecting target flow of each of the road segments according to the target trajectories to determine the target entering number and the target leaving number of the road segment; correspondingly, the determining the road ramp recognition result according to the target features of the road segment comprises: for each of the road segments, determining a target in-out ratio of the road segment according to the target entering number and the target leaving number of the road segment, and determining a road segment ramp recognition result corresponding to the road segment according to the target in-out ratio of the road segment; determining a road ramp recognition result corresponding to the target road according to the road segment ramp recognition results corresponding to the road segments.

3. The ramp identification method of claim 2, wherein, The road segment ramp recognition result comprises a ramp type, and the determining the road segment ramp recognition result corresponding to the road segment according to the target in-out ratio of the road segment comprises: if the target in-out ratio of the road segment is less than a minimum threshold of a preset road target in-out ratio range, determining that the ramp type of the road segment is ramp exit; if the target in-out ratio of the road segment is greater than a maximum threshold of the road target in-out ratio range, determining that the ramp type of the road segment is ramp entrance; and if the target in-out ratio of the road segment is within the road target in-out ratio range, determining that the ramp type of the road segment is no ramp.

4. The ramp identification method of claim 3, wherein, The road segment ramp recognition result further comprises a ramp position, and the determining the target features of the road segment according to the target trajectories further comprises: for each of the target trajectory segments included in each of the road segments, determining a lane-changing feature of each of the trajectory points of the target trajectory segment, and / or determining a speed change feature of each of the trajectory points of the target trajectory segment; correspondingly, after the determining the road segment ramp recognition result corresponding to the road segment according to the target in-out ratio of the road segment, the method further comprises: determining a ramp position in the road segment according to the ramp type corresponding to the road segment and the lane-changing feature and / or the speed change feature of each of the trajectory points corresponding to each of the target trajectory segments included in the road segment.

5. The ramp identification method of claim 4, wherein, The determining the lane-changing feature of each of the trajectory points of the target trajectory segment comprises: For each of the trajectory segments, a lane-changing feature of each trajectory point of the trajectory segment is determined through a lane-changing feature detection step; The lane-changing feature detection step comprises: for each trajectory point of the trajectory segment, a trajectory point lane-changing feature determination procedure is performed to obtain the lane-changing feature of the trajectory point. The trajectory point lane-changing feature determination procedure comprises: obtaining a set of adjacent trajectory points of the trajectory point; determining a trajectory direction of the trajectory point and a trajectory direction of each adjacent trajectory point in the set of adjacent trajectory points; determining an average driving direction of the current road segment according to the trajectory direction of each adjacent trajectory point in the set of adjacent trajectory points; if an angle between the trajectory direction of the trajectory point and the average driving direction of the current road segment is greater than a preset angle threshold, determining that the lane-changing feature of the trajectory point is trajectory deviation; otherwise, determining that the lane-changing feature of the trajectory point is trajectory non-deviation.

6. The ramp identification method of claim 4, wherein, The determination of the speed change feature of each trajectory point of the trajectory segment comprises: For each of the trajectory segments, a speed change feature of each trajectory point of the trajectory segment is determined through a speed change feature detection step; The speed change feature detection step comprises: for each trajectory point of the trajectory segment, a trajectory point speed change feature determination procedure is performed to obtain the speed change feature of the trajectory point. The trajectory point speed change feature determination procedure comprises: determining a current speed of the trajectory point according to current frame detection data; determining a speed average of the trajectory point according to detection data of a preset speed detection frame number; if an absolute value of a difference between the current speed of the trajectory point and the speed average is less than a preset speed fluctuation threshold, determining that the speed change feature of the trajectory point is uniform speed; otherwise: if the current speed of the trajectory point is greater than the speed average, determining that the speed change feature of the trajectory point is acceleration; if the current speed of the trajectory point is less than the speed average, determining that the speed change feature of the trajectory point is deceleration.

7. The ramp identification method of claim 4, wherein, The determination of the ramp location in the road segment comprises: determining a sub-road segment type of a first sub-road segment of the road segment according to the ramp type corresponding to the road segment, the sub-road segment type comprising any one of a normal sub-road segment, a speed-changing sub-road segment, and a ramp; for each trajectory segment included in the road segment, determining a predicted road segment distribution condition corresponding to the trajectory segment according to the lane-changing feature and the speed change feature of each trajectory point of the trajectory segment, the sub-road segment type of the first sub-road segment of the road segment, and a preset hidden Markov model, the predicted road segment distribution condition being a road segment distribution condition of the road segment predicted based on the trajectory segment, the road segment distribution condition comprising a distribution condition of the sub-road segment type in the road segment; determining a final road segment distribution condition of the road segment according to the predicted road segment distribution condition corresponding to each trajectory segment included in the road segment; and determining the ramp location in the road segment according to the final road segment distribution condition of the road segment.

8. A ramp identification device, characterized by The method comprises: a trajectory tracking module, configured to acquire target trajectories of targets on a target road; a road segmenting module, configured to, when a number of the target trajectories is greater than a preset trajectory number, determine a road direction of the target road according to the target trajectories, and divide the target road into N road segments along a direction perpendicular to the road direction according to the target trajectories; N is a positive integer greater than 1; a feature detecting module, configured to, for each of the road segments, determine target features of the road segment according to the target trajectories; the target features include any one or more of a target flow, a target lane-changing feature, and a target speed change feature, and the target flow is used to represent a target number entering or leaving the road segment; a ramp identifying module, configured to determine a road ramp mouth identifying result according to the target features of each of the road segments.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, The computer program, when executed by the processor, causes the electronic device to implement the steps of the method of any one of claims 1 to 7.

10. A computer-readable storage medium storing a computer program, the computer program comprising instructions that, when executed by a computer, cause the computer to perform the method of any one of claims 1 to 9. The computer program, when executed by the processor, causes the electronic device to implement the steps of the method of any one of claims 1 to 7.

Images