Vehicle stop event information detection method and system based on radar and vision fusion

Abstract

The application discloses a vehicle stopping event information detection method and system based on radar and vision fusion. The method comprises the following steps: acquiring the driving track of each vehicle according to space-time information, and judging whether the state of the vehicle at the previous moment of the current moment is a non-stopping state; if the state of the previous moment is a non-stopping state, then judging whether the instantaneous speed at the current moment is less than a low speed threshold value; if yes, then the state of the current moment is a stopping state; if no, then the state of the current moment is a non-stopping state; if the state of the previous moment is a stopping state, then judging whether the instantaneous speed at the current moment is less than a high speed threshold value; if yes, then the state of the current moment is a non-stopping state; if no, then the state of the current moment is a stopping state; judging whether the vehicle exists a transition from a non-stopping state to a stopping state according to the state of each vehicle at all moments; if yes, then calculating the actual stopping times of each vehicle according to the number of transitions from a non-stopping state to a stopping state.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of intelligent transportation, in particular to a vehicle stopping event information detection method and system based on radar and vision fusion. BACKGROUND

[0002] With the rapid development of society, traffic congestion has become one of the problems in traffic construction, and the solution to traffic congestion plays a key role in improving urban road traffic. Vehicle stopping event information is related information about the vehicle in the stopping state, including vehicle parking duration, vehicle parking frequency, vehicle queue length and queue quantity and other information. Vehicle stopping event information is an important indicator for evaluating road traffic capacity, and plays a key role in improving traffic congestion.

[0003] The traditional method divides the detection area on the road based on the video stream, and obtains the image of the corresponding detection area, and counts the vehicles according to the image to obtain the related information of the vehicle stopping event. However, the traditional method needs to accurately calibrate the camera to obtain the position change, but the actual traffic scene has diversity in vehicle driving change, which makes it difficult to determine the actual position of the vehicle through the image, so that the detection accuracy of the traditional method is low. SUMMARY

[0004] The purpose of the present application is to solve the technical problem of low detection accuracy of the traditional method. In order to achieve the above purpose, the present application provides a vehicle stopping event information detection method and system based on radar and vision fusion.

[0005] The present application provides a vehicle stopping event information detection method based on radar and vision fusion, comprising:

[0006] Obtain the space-time information of all vehicles passing through each intersection, and the space-time information is obtained based on radar and vision fusion data;

[0007] According to the space-time information, obtain the driving track of each vehicle, and set the state of each vehicle at the zeroth moment as non-parking state, and the driving track includes the speed and latitude and longitude position of the vehicle at each moment;

[0008] According to the driving track of each vehicle, judge whether the state of the vehicle at the previous moment of the current moment is non-parking state;

[0009] If the state of the previous moment is non-parking state, judge whether the instantaneous speed of the vehicle at the current moment is less than the low speed threshold, if yes, the state of the current moment is parking state, if no, the state of the current moment is non-parking state;

[0010] if the state of the previous time is the parking state, determining whether the instantaneous speed of the vehicle at the current time is less than a high speed threshold, if yes, the state of the current time is the non-parking state, if no, the state of the current time is the parking state;

[0011] determining whether the vehicle has a transition from the non-parking state to the parking state according to the state of all times of each vehicle;

[0012] if yes, calculating the actual parking times of each vehicle according to the number of transitions from the non-parking state to the parking state.

[0013] In one embodiment, after the if the state of the previous time is the parking state, determining whether the instantaneous speed of the vehicle at the current time is less than a high speed threshold, if yes, the state of the current time is the non-parking state, if no, the state of the current time is the parking state, the method further comprises:

[0014] summing up the time lengths corresponding to the continuous parking states according to the state of all times of each vehicle, to obtain a plurality of single parking time lengths of each vehicle;

[0015] obtaining the total parking time length corresponding to the driving track of each vehicle according to the plurality of single parking time lengths.

[0016] In one embodiment, after the determining whether the vehicle has a transition from the non-parking state to the parking state according to the state of all times of each vehicle, the method further comprises:

[0017] if the vehicle has a transition from the non-parking state to the parking state, obtaining the parking time corresponding to the parking state;

[0018] obtaining the latitude and longitude position corresponding to the parking time according to the driving track, and positioning the latitude and longitude position to the lane, direction and intersection.

[0019] In one embodiment, the system further comprises:

[0020] a vehicle classification module, configured to divide the vehicle corresponding to each time into a stopped vehicle and a non-stopped vehicle according to the state of all times of each vehicle;

[0021] a queuing team generation module, configured to obtain a plurality of queuing teams formed by a plurality of stopped vehicles on each lane according to the stopped vehicle and the non-stopped vehicle corresponding to each time;

[0022] a queue length obtaining module, configured to obtain a queue length of each of the queue groups according to a length of a road segment between a front position of a front vehicle of the queue group and a rear position of a rear vehicle of the queue group;

[0023] a queue number length obtaining module, configured to obtain a queue number length of each of the queue groups according to a number of vehicles between the front position of the front vehicle and the rear vehicle of the queue group.

[0024] In an embodiment, the system further comprises:

[0025] a lane queue length obtaining module, configured to obtain a lane queue length according to a sum of the queue lengths of the queue groups located in a same lane;

[0026] a direction queue length obtaining module, configured to obtain a direction queue length according to a sum of the lane queue lengths of all lanes located in a same direction;

[0027] an intersection queue length obtaining module, configured to obtain an intersection queue length according to a sum of the direction queue lengths of all directions located in a same intersection;

[0028] or, a lane queue number length obtaining module, configured to obtain a lane queue number length according to a sum of the queue number lengths of the queue groups located in a same lane;

[0029] a direction queue number length obtaining module, configured to obtain a direction queue number length according to a sum of the lane queue number lengths of all lanes located in a same direction;

[0030] an intersection queue number length obtaining module, configured to obtain an intersection queue number length according to a sum of the direction queue number lengths of all directions located in a same intersection.

[0031] In an embodiment, the system further comprises:

[0032] a timing plan information obtaining module, configured to obtain intersection timing plan information;

[0033] a green-on time queue length obtaining module, configured to obtain a queue length corresponding to a green-on time and a queue length corresponding to a red-on time according to the intersection timing plan information and the lane queue length or the direction queue length or the intersection queue length;

[0034] or, a green-on time queue number length obtaining module, configured to obtain a queue number length corresponding to the green-on time and a queue number length corresponding to the red-on time according to the intersection timing plan information and the lane queue number length or the direction queue number length or the intersection queue number length.

[0035] In the vehicle stop event information detection method and system based on radar and video fusion, the video image data and the radar detection data are fused based on the space-time information of each vehicle, the vehicle state at the previous moment is combined with the vehicle speed at the current moment to jointly determine the vehicle state at the current moment, and the problem of low detection accuracy caused by single image determination is avoided. By combining the vehicle state at the previous moment with the vehicle speed at the current moment and performing threshold interval filtering based on the low speed threshold and the high speed threshold, the mutation between the parking state and the non-parking state is reduced, and the vehicle state at each moment can be accurately determined. Thus, according to the state of each vehicle at all moments, the number of changes from the non-parking state to the parking state is counted, and the actual parking number of each vehicle is obtained. The vehicle stop event information detection method based on radar and video fusion provided in the application can accurately and comprehensively describe the actual parking number of each vehicle from multiple dimensions. The actual parking number of each vehicle can be used as the driving feedback of the vehicle at the intersection to optimize the signal control scheme. BRIEF DESCRIPTION OF DRAWINGS

[0036] Figure 1 is a step flow diagram of the vehicle stop event information detection method based on radar and video fusion provided in the application.

[0037] Figure 2 is a structural diagram of the vehicle stop event information detection system based on radar and video fusion provided in the application. DETAILED DESCRIPTION

[0038] The technical solutions of the application will be further described in detail below with reference to the drawings and examples.

[0039] Please refer to Figure 1 The application provides a vehicle stop event information detection method based on radar and video fusion, which comprises the following steps:

[0040] S10, acquiring the space-time information of all vehicles passing through each intersection, the space-time information being obtained based on radar and video fusion data;

[0041] S20, acquiring the driving trajectory of each vehicle according to the space-time information, and setting the state of each vehicle at the zeroth moment as a non-parking state, the driving trajectory comprising the speed and the latitude and longitude position of the vehicle at each moment;

[0042] S30, judging whether the state of the vehicle at the previous moment at the current moment is a non-parking state according to the driving trajectory of each vehicle;

[0043] S40, if the state of the last time is the non-stopping state, it is judged whether the instantaneous speed of the vehicle at the current time is less than the low speed threshold, if yes, the state of the current time is the stopping state, if no, the state of the current time is the non-stopping state;

[0044] S50, if the state of the last time is the stopping state, it is judged whether the instantaneous speed of the vehicle at the current time is less than the high speed threshold, if yes, the state of the current time is the non-stopping state, if no, the state of the current time is the stopping state;

[0045] S60, according to the state of all times of each vehicle, it is judged whether there is a transition from the non-stopping state to the stopping state for the vehicle;

[0046] S70, if there is, according to the number of transitions from the non-stopping state to the stopping state, the actual stopping number of each vehicle is calculated.

[0047] In the embodiment, the radar and video integrated machine is arranged at the intersection, and is used to acquire video image data and radar detection data. The video image data and the radar detection data are fused to obtain relative space-time information of each vehicle relative to the intersection. The vehicle is captured through the video image data, and vehicle attribute information such as a license plate number, a license plate color, and a vehicle color is recognized. The vehicle distance and the vehicle speed are recognized through the radar detection data. The video image data and the radar detection data are transmitted through an optical fiber, demodulated by a demodulator, and then transmitted to a platform end for fusion.

[0048] When the camera-based vehicle information and the radar-based vehicle information are fused in space-time, it can be divided into time fusion and space fusion. The radar coordinate system and the camera coordinate system are converted, the radar coordinate system is converted to the pixel coordinate system, the space fusion of the video image data and the radar detection data is completed, the video image data and the radar detection data are synchronously collected in time, and the time fusion is realized. The high-precision map can provide stop line contour position information of the intersection, lane line contour position information, and latitude and longitude information of the intersection. The relative space-time information includes a relative spatial position of the vehicle relative to the intersection. According to the distance of the vehicle relative to the intersection, the relative spatial position can be converted into a spatial position in the geographic coordinate system to obtain the latitude and longitude position corresponding to the vehicle in the geographic coordinate system. Thus, the relative space-time information of each vehicle relative to the intersection is converted into space-time information, and then the driving trajectory of each vehicle is obtained. The space-time information includes the latitude and longitude position of a certain vehicle at a certain time, time information, vehicle position information, vehicle speed information, vehicle driving direction information, vehicle model information, vehicle license plate information, and vehicle color information. According to the latitude and longitude position corresponding to the vehicle, the corresponding vehicle can be marked on the map to provide related geographic position information support.

[0049] The driving trajectory of each vehicle is analyzed. The driving trajectory includes the speed and latitude and longitude position of the vehicle at each time. For each time corresponding to the trajectory, the state of the zeroth time is set as a non-stopping state. According to the driving trajectory of each vehicle, the speed of each vehicle is obtained, and the motion state of the vehicle is known through the analysis of the speed.

[0050] A low speed threshold and a high speed threshold are set. The range of the low speed threshold can be 2 km / h to 4 km / h. The range of the high speed threshold can be 9 km / h to 11 km / h. Taking the zeroth time as the initial state, the vehicle reaches the intersection from the first time, and the current time can be the first time, the second time, the third time, and so on. From the zeroth time to the current time, if the state corresponding to the last time of the current time is a non-stopping state, it is further judged whether the instantaneous speed of the current time is lower than the low speed threshold, if it is lower, the state of the current time is a stopping state, if it is not lower, the state of the current time is a non-stopping state. It is further judged whether the instantaneous speed of the current time is lower than the high speed threshold, if it is lower, the state of the current time is a non-stopping state, if it is not lower, the state of the current time is a stopping state. The vehicle state of the last time and the vehicle instantaneous speed of the current time are combined to jointly determine the vehicle state of the current time.

[0051] After traversing all the times, the vehicle states of all the times are judged, and the vehicle states of all the times are obtained. For each time state from non-stopping state to stopping state, it is regarded as 1 actual stopping time. The vehicle state transition of all times is calculated to obtain the actual stopping time of each vehicle.

[0052] The vehicle stopping event information detection method based on radar and video fusion provided in the application is based on the space-time information of each vehicle, and the video image data and the radar detection data are fused. Based on the multi-dimensional information of the vehicle, the vehicle state of the last time and the vehicle speed of the current time are combined to jointly determine the vehicle state of the current time, which avoids the problem of low detection accuracy caused by single image judgment. By combining the vehicle state of the last time and the vehicle speed of the current time, and taking the low speed threshold and the high speed threshold as the judgment basis for threshold interval filtering, the mutation between the stopping state and the non-stopping state is reduced, and the vehicle state of each time can be accurately determined. Therefore, according to the state of each vehicle at all times, the change number from non-stopping state to stopping state is counted to obtain the actual stopping time of each vehicle. Through the vehicle stopping event information detection method based on radar and video fusion provided in the application, the actual stopping time of each vehicle can be accurately and comprehensively described from multiple dimensions. The actual stopping time of each vehicle can be used as the driving feedback of the vehicle at the intersection to optimize the signal control scheme.

[0053] In one embodiment, S50, if the state of the previous moment is the parking state, it is judged whether the instantaneous speed of the vehicle at the current moment is less than the high speed threshold, if yes, the state of the current moment is the non-parking state, if no, the state of the current moment is the parking state, and then the vehicle stopping event information detection method based on radar and vision fusion further comprises:

[0054] According to the state of each vehicle at all moments, the time lengths corresponding to the continuous multiple parking states are summed up to obtain the multiple single parking time lengths of each vehicle;

[0055] According to the multiple single parking time lengths, the parking total time length corresponding to the driving trajectory of each vehicle is obtained.

[0056] In this embodiment, the continuous multiple parking states can be understood as the vehicle states corresponding to the continuous moments being parking states, the moments corresponding to all the continuous parking states are summed up to obtain the single parking time length corresponding to each parking of each vehicle. By adding up the multiple single parking time lengths, the total parking time length corresponding to all moments of each vehicle, i.e., the parking total time length corresponding to the driving trajectory of each vehicle, can be obtained. The single parking time length and the parking total time length can be used as the information of the stopping event, as the driving feedback of the vehicle at the intersection, and to optimize the signal control scheme.

[0057] In one embodiment, according to the state of each vehicle at all moments, it is judged whether the vehicle has a transition from the non-parking state to the parking state, and the vehicle stopping event information detection method based on radar and vision fusion further comprises:

[0058] If the vehicle has a transition from the non-parking state to the parking state, the parking moment corresponding to the parking state is obtained;

[0059] According to the driving trajectory, the latitude and longitude position corresponding to the parking moment is obtained, and the latitude and longitude position is positioned to the lane, direction and intersection.

[0060] In this embodiment, each moment corresponds to a vehicle state, which can be a parking state or a non-parking state. After traversing the states of all moments, the moment corresponding to the transition from the non-parking state to the parking state can be obtained, and then the parking moment corresponding to the parking state can be obtained. According to the space-time information, the driving trajectory can be obtained, and then the latitude and longitude position corresponding to the parking moment can be obtained. According to the latitude and longitude position corresponding to the parking moment, the stopping event corresponding to the parking moment can be compared in the high-definition map and regressed to a specific lane, a specific direction and a specific intersection, which can be used as the information of the stopping event, as the driving feedback of the vehicle at the intersection, and to optimize the signal control scheme.

[0061] In one embodiment, the vehicle stopping event information detection method based on radar and vision fusion further comprises:

[0062] determining whether the actual number of stops of each vehicle is 0;

[0063] if not, the number of stops in a period is 1.

[0064] In this embodiment, the timing plan information is obtained through the signal control machine, the period control time of the timing plan information is combined with the actual number of stops of each vehicle, the start and end time of each period is obtained, and it is determined whether the actual number of stops of each vehicle in a period is 0. If not, the number of stops in a period is 1, and the number of stops in a period is obtained. The actual number of stops, the number of stops in a period, the single stop time, the total stop time, the latitude and longitude position, the lane, the direction, and the intersection are used as index data, which are recorded and stored for use by various applications as driving feedback of vehicles at intersections to optimize the signal control scheme.

[0065] In one embodiment, S50, if the state of the last time is the stop state, it is determined whether the instantaneous speed of the vehicle at the current time is less than the high speed threshold. If yes, the state of the current time is the non-stop state, and if not, the state of the current time is the stop state. After that, the vehicle stop event information detection method based on radar and vision fusion further includes:

[0066] According to the state of each vehicle at each time, the vehicle corresponding to each time is divided into stopped vehicles and non-stopped vehicles;

[0067] According to the stopped vehicles and non-stopped vehicles corresponding to each time, a plurality of queuing teams formed by a plurality of consecutive stopped vehicles on each lane are obtained;

[0068] According to the road segment length between the head position of the head vehicle of each queuing team and the tail position of the tail vehicle, the queuing team length of each queuing team is obtained;

[0069] According to the number of vehicles between the head position of the head vehicle and the tail vehicle, the queuing number length of each queuing team is obtained.

[0070] In this embodiment, each vehicle corresponds to a driving track. The state of each time can be a parking state or a non-parking state. The parking state corresponds to a stopped vehicle, and the non-parking state corresponds to a non-stopped vehicle. According to the state of each vehicle at each time, the state of all vehicles at a certain time can be obtained, and further, it can be known whether each vehicle at a certain time is a stopped vehicle or a non-stopped vehicle, so as to divide each vehicle at each time into two categories of stopped vehicles and non-stopped vehicles. For example, at 10:20, A vehicle is a stopped vehicle, B vehicle is a stopped vehicle, C vehicle is a stopped vehicle, D vehicle is a non-stopped vehicle, E vehicle is a non-stopped vehicle, F vehicle is a stopped vehicle, G vehicle is a stopped vehicle, H vehicle is a stopped vehicle, and I vehicle is a non-stopped vehicle.

[0071] After the plurality of vehicles on each lane are divided into stopped vehicles or non-stopped vehicles, the plurality of vehicles on a certain lane at a certain time are traversed from the center of the intersection as a starting point, and a plurality of queuing teams are formed by the plurality of consecutive stopped vehicles on the lane. The plurality of consecutive stopped vehicles can be understood as that there is no non-stopped vehicle between the consecutive stopped vehicles. The plurality of consecutive stopped vehicles form a queuing team, for example, A vehicle, B vehicle, and C vehicle form a queuing team, and F vehicle, G vehicle, and H vehicle form a queuing team. The first stopped vehicle of each queuing team is set as a head vehicle, and the last stopped vehicle is set as a tail vehicle. The road segment length between the head vehicle and the tail vehicle of each queuing team is calculated as the queuing team length of each queuing team. The road segment length includes the non-straight distance corresponding to the road curvature. In this embodiment, the vehicles are divided into stopped vehicles or non-stopped vehicles by the vehicle state corresponding to each vehicle at each time, the consecutive stopped vehicles are queued, and a plurality of queuing teams corresponding to a lane are formed. The vehicle state at the current time is obtained according to the vehicle state at the last time and the vehicle speed at the current time, which avoids the problem of low detection accuracy caused by the judgment from a single image, and the queuing team length and the queuing number length of each queuing team can be more accurately obtained.

[0072] In one embodiment, after the queuing number length of each queuing team is obtained according to the number of vehicles between the head vehicle and the tail vehicle of the queuing team, the vehicle stopping event information detection method based on radar and vision fusion further includes:

[0073] obtaining a lane queuing length according to the sum of the queuing team lengths of the plurality of queuing teams located in the same lane;

[0074] obtaining a direction queuing length according to the sum of the lane queuing lengths of all lanes located in the same direction;

[0075] obtaining an intersection queuing length according to the sum of the direction queuing lengths of all directions located in the same intersection;

[0076] Or, according to the sum of the queue length of the plurality of queue teams located in the same lane, the lane queue length is obtained;

[0077] According to the sum of the lane queue length of all lanes located in the same direction, the direction queue length is obtained;

[0078] According to the sum of the direction queue length of all directions located in the same intersection, the intersection queue length is obtained;

[0079] In this embodiment, the same lane corresponds to at least one queue team, and the queue length of each queue team corresponding to the queue team is summed up to obtain the lane queue length corresponding to the lane. The same direction corresponds to at least one lane, and the lane queue length corresponding to each lane is summed up to obtain the direction queue length corresponding to the direction. The same intersection corresponds to multiple directions, and the direction queue length corresponding to each direction is summed up to obtain the intersection queue length.

[0080] The same lane corresponds to at least one queue team, and the queue length of each queue team corresponding to the queue team is summed up to obtain the lane queue length corresponding to the lane. The same direction corresponds to at least one lane, and the lane queue length corresponding to each lane is summed up to obtain the direction queue length corresponding to the direction. The same intersection corresponds to multiple directions, and the direction queue length corresponding to each direction is summed up to obtain the intersection queue length.

[0081] In one embodiment, after obtaining the intersection queue length according to the sum of the direction queue length of all directions located in the same intersection, or after obtaining the intersection queue length according to the sum of the direction queue length of all directions located in the same intersection, the vehicle stop event information detection method based on radar and vision fusion further comprises:

[0082] Obtain intersection timing plan information;

[0083] According to the intersection timing plan information and the lane queue length or the direction queue length or the intersection queue length, the queue length corresponding to the red light start time and the queue length corresponding to the green light start time are obtained;

[0084] Or, according to the intersection timing plan information and the lane queue length or the direction queue length or the intersection queue length, the queue length corresponding to the red light start time and the queue length corresponding to the green light start time are obtained.

[0085] In this embodiment, the platform end is connected with the signal control machine, and the timing scheme information can be obtained through the message queue. Through the timing scheme information, the time of each red light starting to light up and the time of each green light starting to light up can be calculated and obtained. The time of red light starting to light up can be understood as the time of red light starting to brighten. The time of green light starting to light up can be understood as the time of green light starting to brighten. The lane queue length, the direction queue length, the intersection queue length, the lane queue number length, the direction queue number length, and the intersection queue number length are all calculated from a plurality of queue teams formed by a plurality of stopped vehicles corresponding to a certain time, and correspond to the length of a certain time. According to the intersection timing scheme information, the red light starting time and the green light starting time can be obtained, and then the lane queue length, the direction queue length, the intersection queue length, the lane queue number length, the direction queue number length, and the intersection queue number length calculated from a plurality of queue teams corresponding to a certain time can be obtained.

[0086] The queue length corresponding to the red light starting time can be the lane queue length corresponding to the red light starting time or the direction queue length corresponding to the red light starting time or the intersection queue length corresponding to the red light starting time. The queue length corresponding to the green light starting time can be the lane queue length corresponding to the green light starting time or the direction queue length corresponding to the green light starting time or the intersection queue length corresponding to the green light starting time.

[0087] The queue number length corresponding to the red light starting time can be the lane queue number length corresponding to the red light starting time or the direction queue number length corresponding to the red light starting time or the intersection queue number length corresponding to the red light starting time. The queue number length corresponding to the green light starting time can be the lane queue number length corresponding to the green light starting time or the direction queue number length corresponding to the green light starting time or the intersection queue number length corresponding to the green light starting time.

[0088] According to the vehicle state of each vehicle at all times, the stopped vehicles and the non-stopped vehicles are distinguished, and then the queue teams are formed according to the vehicle state, and the team length and the number are calculated. According to the spatial relationship of the intersection, the direction and intersection indexes are calculated, and the length and number of the direction and the intersection corresponding to a certain time are obtained. According to the signal control time, the queue length indexes at a certain time are calculated, such as the queue length corresponding to the green light starting time and the red light starting time. Therefore, the vehicle stopping event information detection method based on radar and vision fusion provided in the present application can accurately, quasi-real-time, quickly and efficiently give the queue length information of different lanes, different times and different directions of the intersection, which is helpful for analyzing the running situation of the intersection and providing data support for signal control optimization and evaluation.

[0089] Please refer to Figure 2The application provides a vehicle stop event information detection system 100 based on radar and vision fusion. The vehicle stop event information detection system 100 based on radar and vision fusion comprises a space-time information acquisition module 10, a driving track acquisition module 20, a vehicle state judgment module 30, a low speed threshold judgment module 40, a high speed threshold judgment module 50, a state transition judgment module 60 and an actual parking times calculation module 70. The space-time information acquisition module 10 is used for acquiring space-time information of all vehicles passing through each intersection, and the space-time information is obtained based on radar and vision fusion data. The driving track acquisition module 20 is used for acquiring a driving track of each vehicle according to the space-time information, and setting a state of each vehicle at a zero time as a non-parking state, and the driving track comprises a speed and a latitude and longitude position of the vehicle at each time.

[0090] The vehicle state judgment module 30 is used for judging whether a state of the vehicle at a previous time of a current time is a non-parking state according to the driving track of each vehicle. The low speed threshold judgment module 40 is used for judging whether an instantaneous speed of the vehicle at the current time is less than a low speed threshold if the state of the previous time is the non-parking state, and if yes, the state of the current time is a parking state, and if no, the state of the current time is the non-parking state. The high speed threshold judgment module 50 is used for judging whether the instantaneous speed of the vehicle at the current time is less than a high speed threshold if the state of the previous time is the parking state, and if yes, the state of the current time is the non-parking state, and if no, the state of the current time is the parking state.

[0091] The state transition judgment module 60 is used for judging whether there is a transition from the non-parking state to the parking state of the vehicle according to the state of each vehicle at all times. The actual parking times calculation module 70 is used for calculating actual parking times of each vehicle according to a transition times from the non-parking state to the parking state if there is.

[0092] In the embodiment, the related description of the space-time information acquisition module 10 can refer to the related description of S10 in the above embodiment. The related description of the driving track acquisition module 20 can refer to the related description of S20 in the above embodiment. The related description of the vehicle state judgment module 30 can refer to the related description of S30 in the above embodiment. The related description of the low speed threshold judgment module 40 can refer to the related description of S40 in the above embodiment. The related description of the high speed threshold judgment module 50 can refer to the related description of S50 in the above embodiment. The related description of the state transition judgment module 60 can refer to the related description of S60 in the above embodiment. The related description of the actual parking times calculation module 70 can refer to the related description of S70 in the above embodiment.

[0093] In an embodiment, the radar and vision fusion based vehicle stop event information detection system 100 further comprises a single parking duration calculation module and a total parking duration calculation module. The single parking duration calculation module is configured to sum up durations corresponding to continuous multiple parking states to obtain multiple single parking durations of each vehicle according to states of each vehicle at all time points. The total parking duration calculation module is configured to obtain a total parking duration corresponding to the driving track of each vehicle according to the multiple single parking durations. In this embodiment, the related descriptions of the single parking duration calculation module and the total parking duration calculation module can refer to the descriptions of the related methods in the above embodiments.

[0094] In an embodiment, the radar and vision fusion based vehicle stop event information detection system 100 further comprises a parking time point acquisition module and a position positioning module. The parking time point acquisition module is configured to acquire a parking time point corresponding to the parking state if the vehicle exists a transition from the non-parking state to the parking state. The position positioning module is configured to acquire a latitude and longitude position corresponding to the parking time point according to the driving track, and position the latitude and longitude position to a lane, a direction and an intersection.

[0095] In this embodiment, the related descriptions of the parking time point acquisition module and the position positioning module can refer to the descriptions of the related methods in the above embodiments.

[0096] In an embodiment, the radar and vision fusion based vehicle stop event information detection system 100 further comprises a vehicle classification module, a queuing team generation module, a queuing team length acquisition module and a queuing number length acquisition module. The vehicle classification module is configured to divide vehicles corresponding to each time point into stop vehicles and non-stop vehicles according to states of each vehicle at all time points. The queuing team generation module is configured to obtain multiple queuing teams formed by continuous multiple stop vehicles on each lane according to stop vehicles and non-stop vehicles corresponding to each time point. The queuing team length acquisition module is configured to obtain a queuing team length of each queuing team according to a road segment length between a head position of a head vehicle of each queuing team and a tail position of a tail vehicle. The queuing number length acquisition module is configured to obtain a queuing number length of each queuing team according to a number of vehicles between the head position of the head vehicle and the tail vehicle.

[0097] In this embodiment, the related descriptions of the vehicle classification module, the queuing team generation module, the queuing team length acquisition module and the queuing number length acquisition module can refer to the descriptions of the related methods in the above embodiments.

[0098] In one embodiment, the radar and vision fusion based vehicle stop event information detection system 100 further comprises a lane queue length obtaining module, a direction queue length obtaining module, and an intersection queue length obtaining module, or a lane queue number length obtaining module, a direction queue number length obtaining module, and an intersection queue number length obtaining module. The lane queue length obtaining module is configured to obtain a lane queue length according to a sum of queue vehicle fleet lengths of a plurality of queue teams located in a same lane. The direction queue length obtaining module is configured to obtain a direction queue length according to a sum of lane queue lengths of all lanes located in a same direction. The intersection queue length obtaining module is configured to obtain an intersection queue length according to a sum of direction queue lengths of all directions located in a same intersection.

[0099] Alternatively, the lane queue number length obtaining module is configured to obtain a lane queue number length according to a sum of queue number lengths of a plurality of queue teams located in a same lane. The direction queue number length obtaining module is configured to obtain a direction queue number length according to a sum of lane queue number lengths of all lanes located in a same direction. The intersection queue number length obtaining module is configured to obtain an intersection queue number length according to a sum of direction queue number lengths of all directions located in a same intersection.

[0100] In this embodiment, the lane queue length obtaining module, the direction queue length obtaining module, and the intersection queue length obtaining module, or the lane queue number length obtaining module, the direction queue number length obtaining module, and the intersection queue number length obtaining module, can refer to the descriptions of the related methods in the above embodiments.

[0101] In one embodiment, the radar and vision fusion based vehicle stop event information detection system 100 further comprises a timing plan information obtaining module, a queue length at start-up time obtaining module, or a queue number length at start-up time obtaining module. The timing plan information obtaining module is configured to obtain intersection timing plan information. The queue length at start-up time obtaining module is configured to obtain queue lengths corresponding to a start-up time of a red light and a start-up time of a green light according to the intersection timing plan information and a lane queue length or a direction queue length or an intersection queue length. Alternatively, the queue number length at start-up time obtaining module is configured to obtain queue number lengths corresponding to a start-up time of a red light and a start-up time of a green light according to the intersection timing plan information and a lane queue number length or a direction queue number length or an intersection queue number length.

[0102] In this embodiment, the timing plan information obtaining module, the queue length at start-up time obtaining module, or the queue number length at start-up time obtaining module can refer to the descriptions of the related methods in the above embodiments.

[0103] In the various embodiments described above, the particular order or hierarchy of steps in processes disclosed should not be understood to represent a limitation but shall be understood to be an example of an illustrative order. Based upon design preferences, it should be understood that the particular order or hierarchy of steps in the processes could be rearranged, or otherwise modified, without departing from the scope of the disclosure. The accompanying method claims present elements of the various steps in a sample order, and are not meant to be limited to the specific order or hierarchy presented.

[0104] Those of skill would further appreciate that the various illustrative logical blocks, modules, and steps described in connection with the embodiments disclosed herein can be implemented by electronic hardware, computer software, or combinations of both. To clearly illustrate this interchangeability of hardware and software, various illustrative components, blocks, modules, and steps have been described above generally in terms of their functionality. Whether such functionality is implemented as hardware or software depends upon the particular application and design constraints imposed on the overall system. Skilled artisans can implement the described functionality in varying ways for each particular application, but such implementation decisions should not be interpreted as causing a departure from the scope of the present embodiments.

[0105] The various illustrative logical blocks, modules, and steps described in connection with the embodiments disclosed herein can be implemented or performed by a general purpose processor, a digital signal processor (DSP), an application specific integrated circuit (ASIC), a field programmable gate array (FPGA) or other programmable logic device, discrete gate or transistor logic, discrete hardware components, or any combination thereof designed to perform the functions described herein. A general purpose processor can be a microprocessor, but in the alternative, the general purpose processor can be any conventional processor, controller, microcontroller, or state machine. A processor can also be implemented as a combination of computing devices, e.g., a combination of a DSP and a microprocessor, a plurality of microprocessors, one or more microprocessors in conjunction with a DSP core, or any other such configuration.

[0106] The steps of a method or algorithm described in connection with the embodiments disclosed herein can be embodied directly in hardware, in a software module executed by a processor, or in a combination of the two. A software module can reside in RAM memory, flash memory, ROM memory, EPROM memory, EEPROM memory, registers, hard disk, a removable disk, a CD-ROM, or any other form of storage medium known in the art. An exemplary storage medium can be coupled to the processor, such that the processor can read information from, and write information to, the storage medium. In the alternative, the storage medium can be integral to the processor. The processor and the storage medium can reside in an ASIC. The ASIC can reside in a user terminal. In the alternative, the processor and the storage medium can reside as discrete components in a user terminal. The previous description of the disclosed embodiments is provided to enable any person skilled in the art to make or use the present application. The various methods described herein can be implemented on a computer using software having suitable instructions or programming code applied as needed. The embodiments described herein are not inherently related to any particular computer or other apparatus. Various hardware and software

[0107] The above detailed description has been given to the purpose of further explaining the purpose, technical solutions and beneficial effects of the present application. It should be understood that the above description is only a specific implementation of the present application and is not intended to limit the protection scope of the present application. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application shall be included in the protection scope of the present application.

Claims

1. A vehicle stop event information detection method based on radar and vision fusion, characterized by, The method comprises the following steps: acquiring the space-time information of all vehicles passing through each intersection, the space-time information being obtained based on radar and vision fusion data; acquiring the driving trajectory of each vehicle according to the space-time information, and setting the state of each vehicle at the zeroth moment as a non-stopping state, the driving trajectory comprising the speed and latitude-longitude position of the vehicle at each moment; judging whether the state of the vehicle at the previous moment of the current moment is a non-stopping state according to the driving trajectory of each vehicle; if the state of the previous moment is a non-stopping state, judging whether the instantaneous speed of the vehicle at the current moment is less than a low speed threshold, if yes, the state of the current moment is a stopping state, if no, the state of the current moment is a non-stopping state; if the state of the previous moment is a stopping state, judging whether the instantaneous speed of the vehicle at the current moment is less than a high speed threshold, if yes, the state of the current moment is a non-stopping state, if no, the state of the current moment is a stopping state; judging whether the vehicle has a transition from the non-stopping state to the stopping state according to the state of each vehicle at all moments; if yes, calculating the actual stopping times of each vehicle according to the number of transitions from the non-stopping state to the stopping state; after the step of if the state of the previous moment is a stopping state, judging whether the instantaneous speed of the vehicle at the current moment is less than a high speed threshold, if yes, the state of the current moment is a non-stopping state, if no, the state of the current moment is a stopping state, the method further comprises: dividing the vehicle corresponding to each moment into a stopped vehicle and a non-stopped vehicle according to the state of each vehicle at all moments; obtaining a plurality of queuing teams formed by a plurality of stopped vehicles on each lane according to the stopped vehicle and the non-stopped vehicle corresponding to each moment; obtaining the queuing team length of each queuing team according to the road segment length between the head position of the head vehicle of the queuing team and the tail position of the tail vehicle of the queuing team; obtaining the queuing number length of each queuing team according to the number of vehicles between the head position of the head vehicle and the tail vehicle of the queuing team; after the step of obtaining the queuing number length of each queuing team according to the number of vehicles between the head position of the head vehicle and the tail vehicle of the queuing team, the method further comprises: obtaining the lane queuing length according to the sum of the queuing team lengths of the plurality of queuing teams located in the same lane; obtaining the direction queuing length according to the sum of the lane queuing lengths of all lanes located in the same direction; obtaining the intersection queuing length according to the sum of the direction queuing lengths of all directions located in the same intersection; or, obtaining the lane queuing number length according to the sum of the queuing number lengths of the plurality of queuing teams located in the same lane; obtaining the direction queuing number length according to the sum of the lane queuing number lengths of all lanes located in the same direction; obtaining the intersection queuing number length according to the sum of the direction queuing number lengths of all directions located in the same intersection.

2. The method of claim 1, wherein the method is based on a fusion of visual and radar information. if the state of the previous time is the non-stopping state, determining whether the instantaneous speed of the vehicle at the current time is less than a low speed threshold, if yes, the state of the current time is the non-stopping state, if no, the state of the current time is the stopping state, after determining whether the state of the vehicle at the current time is the stopping state or the non-stopping state, the method further comprises: summing up the time lengths corresponding to the stopping states of the continuous multiple stopping states according to the states of all times of each vehicle, to obtain the single stopping time length of each vehicle; obtaining the total stopping time length corresponding to the driving trajectory of each vehicle according to the single stopping time length.

3. The method of claim 1, wherein the method further comprises: if the state of the previous time is the non-stopping state, determining whether the instantaneous speed of the vehicle at the current time is less than a low speed threshold, if yes, the state of the current time is the non-stopping state, if no, the state of the current time is the stopping state, after determining whether the state of the vehicle at the current time is the stopping state or the non-stopping state, the method further comprises: if the state of the previous time is the non-stopping state, determining whether the instantaneous speed of the vehicle at the current time is less than a low speed threshold, if yes, the state of the current time is the non-stopping state, if no, the state of the current time is the stopping state, after determining whether the state of the vehicle at the current time is the stopping state or the non-stopping state, the method further comprises: obtaining the longitude and latitude position corresponding to the stopping time according to the driving trajectory, and positioning the longitude and latitude position to the lane, the direction and the intersection.

4. The method of claim 1, wherein the method further comprises: the method further comprises: obtaining the intersection timing scheme information; obtaining the queue length corresponding to the red light starting time and the queue length corresponding to the green light starting time according to the intersection timing scheme information and the lane queue length or the direction queue length or the intersection queue length; or, obtaining the queue number length corresponding to the red light starting time and the queue number length corresponding to the green light starting time according to the intersection timing scheme information and the lane queue number length or the direction queue number length or the intersection queue number length.

5. A vehicle stop event information detection system based on radar and vision fusion, characterized by, comprises: a space-time information acquisition module, configured to acquire space-time information of all vehicles passing through each intersection, the space-time information being obtained based on radar and vision fusion data; a driving trajectory acquisition module, configured to acquire a driving trajectory of each vehicle according to the space-time information, and set the state of each vehicle at the zeroth time as the non-stopping state, the driving trajectory comprising a speed and a longitude and latitude position of the vehicle at each time; a vehicle state judgment module, configured to determine whether the state of the vehicle at the previous time of the current time is the non-stopping state according to the driving trajectory of each vehicle; a low speed threshold judgment module, configured to determine whether the instantaneous speed of the vehicle at the current time is less than a low speed threshold if the state of the previous time is the non-stopping state, if yes, the state of the current time is the stopping state, if no, the state of the current time is the non-stopping state; a high speed threshold judgment module, configured to determine whether the instantaneous speed of the vehicle at the current time is less than a high speed threshold if the state of the previous time is the stopping state, if yes, the state of the current time is the non-stopping state, if no, the state of the current time is the stopping state; a state transition judgment module, configured to determine whether the vehicle has a transition from the non-stopping state to the stopping state according to the states of all times of each vehicle. An actual parking frequency calculation module is configured to calculate the actual parking frequency of each vehicle according to the number of transitions from the non-parking state to the parking state if the number of transitions exists; The system further comprises: A vehicle classification module is configured to divide each vehicle at each time into a stopped vehicle and a non-stopped vehicle according to the state of each vehicle at each time; A queuing team generation module is configured to obtain a plurality of queuing teams formed by a plurality of stopped vehicles at each lane according to the stopped vehicles and the non-stopped vehicles at each time; A queuing team length acquisition module is configured to obtain the queuing team length of each queuing team according to the road segment length between the head position of the head vehicle and the tail position of the tail vehicle of each queuing team; A queuing number length acquisition module is configured to obtain the queuing number length of each queuing team according to the number of vehicles between the head position of the head vehicle and the tail vehicle; The system further comprises: A lane queuing length acquisition module is configured to obtain the lane queuing length according to the sum of the queuing team lengths of the plurality of queuing teams located at the same lane; A direction queuing length acquisition module is configured to obtain the direction queuing length according to the sum of the lane queuing lengths of all lanes located at the same direction; An intersection queuing length acquisition module is configured to obtain the intersection queuing length according to the sum of the direction queuing lengths of all directions located at the same intersection; Alternatively, a lane queuing number length acquisition module is configured to obtain the lane queuing number length according to the sum of the queuing number lengths of the plurality of queuing teams located at the same lane; A direction queuing number length acquisition module is configured to obtain the direction queuing number length according to the sum of the lane queuing number lengths of all lanes located at the same direction; An intersection queuing number length acquisition module is configured to obtain the intersection queuing number length according to the sum of the direction queuing number lengths of all directions located at the same intersection.

6. The storm-based vehicle stoppage event information detection system of claim 5, wherein, The system further comprises: A single parking duration calculation module is configured to obtain a plurality of single parking durations of each vehicle by summing the durations corresponding to a plurality of consecutive parking states according to the state of each vehicle at all times; A total parking duration calculation module is configured to obtain the total parking duration corresponding to the driving trajectory of each vehicle according to the plurality of single parking durations.

7. The storm-based vehicle stoppage event information detection system of claim 5, wherein, The system further comprises: A parking time acquisition module is configured to obtain the parking time corresponding to the parking state if the vehicle has a transition from the non-parking state to the parking state; A position positioning module is configured to obtain the latitude and longitude position corresponding to the parking time according to the driving trajectory, and position the latitude and longitude position to a lane, a direction, and an intersection.

8. The storm-based vehicle stoppage event information detection system of claim 7, wherein, The system further comprises: A timing plan information acquisition module is configured to obtain intersection timing plan information; An illumination time queuing length acquisition module is configured to obtain the queuing length corresponding to the red light illumination time and the queuing length corresponding to the green light illumination time according to the intersection timing plan information and the lane queuing length or the direction queuing length or the intersection queuing length. Alternatively, the green light on time queue length obtaining module is configured to obtain the queue length corresponding to the green light on time and the queue length corresponding to the red light on time according to the intersection timing plan information and the lane queue length or the direction queue length or the intersection queue length.

Images