CN113034963B - Vision parking stall tracking system - Google Patents
Vision parking stall tracking system Download PDFInfo
- Publication number
- CN113034963B CN113034963B CN202110231087.3A CN202110231087A CN113034963B CN 113034963 B CN113034963 B CN 113034963B CN 202110231087 A CN202110231087 A CN 202110231087A CN 113034963 B CN113034963 B CN 113034963B
- Authority
- CN
- China
- Prior art keywords
- parking space
- parking
- visual
- vehicle
- tracking
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G08—SIGNALLING
- G08G—TRAFFIC CONTROL SYSTEMS
- G08G1/00—Traffic control systems for road vehicles
- G08G1/14—Traffic control systems for road vehicles indicating individual free spaces in parking areas
-
- E—FIXED CONSTRUCTIONS
- E04—BUILDING
- E04H—BUILDINGS OR LIKE STRUCTURES FOR PARTICULAR PURPOSES; SWIMMING OR SPLASH BATHS OR POOLS; MASTS; FENCING; TENTS OR CANOPIES, IN GENERAL
- E04H6/00—Buildings for parking cars, rolling-stock, aircraft, vessels or like vehicles, e.g. garages
- E04H6/42—Devices or arrangements peculiar to garages, not covered elsewhere, e.g. securing devices, safety devices, monitoring and operating schemes; centering devices
-
- G—PHYSICS
- G08—SIGNALLING
- G08G—TRAFFIC CONTROL SYSTEMS
- G08G1/00—Traffic control systems for road vehicles
- G08G1/01—Detecting movement of traffic to be counted or controlled
- G08G1/04—Detecting movement of traffic to be counted or controlled using optical or ultrasonic detectors
Abstract
The invention discloses a visual parking space tracking system, which adopts the technical scheme that the system comprises a processing module, a camera module and a vehicle mileage module, wherein the processing module comprises a global model unit and a parking space matching unit, a global coordinate system is configured in the global model unit, the vehicle mileage module is arranged on a vehicle, the global coordinate system is used for displaying a visual parking space and a vehicle mileage route, and the visual parking space is compared with a tracking parking space recorded in a tracking list through the parking space matching unit, so that whether the visual parking space is matched with the tracking parking space or not is judged, and the effects of tracking rancour parking spaces and accurately judging whether the parking spaces can be used for vehicles to park are achieved.
Description
Technical Field
The invention relates to the technical field of automobile illumination, in particular to a visual parking space tracking system.
Background
With the increase of the scale and complexity of the parking lot, the construction of an indoor parking lot and an outdoor parking lot is particularly important, wherein the indoor parking lot often has the conditions of narrow road, complex environment and narrow visual field, the outdoor parking lot has a relatively wide visual field, but when parking, a driver is difficult to quickly find a parking space needing parking, and although the intelligent degree in daily life is continuously improved, an intelligent auxiliary parking system is also developed.
Currently, parking space detection is adopted in an auxiliary parking system, and a vehicle is matched with the parking space, so that the vehicle can be conveniently parked into the matched parking space, at present, a plurality of methods for detecting the parking space are provided, wherein a detection mode through a fisheye camera is a currently common mode, the detection mode is generally detected through a traditional image algorithm or a deep learning algorithm, deep learning can detect whether the parking space is an empty parking space, but the detection effect of the deep learning mainly depends on a learned sample, namely a model sample of a parking lot, but parking space scenes are very complicated and changeable, and empty parking space detection of all parking spaces is difficult to support through the deep learning method, so that the parking space state cannot be accurately tracked in real time; the traditional image algorithm is used for identifying the parking space by detecting the line clip angle of the parking space, generally cannot detect the state of the parking space, and cannot accurately judge whether the parking space is an empty parking space; at present, the parking space is detected through the fisheye camera, and when a parking space image is detected, the tracking detection effect on the parking space is poor, so that the parking space cannot be accurately tracked, whether the parking space can be used for parking a vehicle or not can be accurately judged, and the auxiliary parking efficiency of the vehicle is reduced.
Disclosure of Invention
Aiming at the defects in the prior art, the invention aims to provide a visual parking space tracking system which has the effects of tracking a parking space and accurately judging whether the parking space can be used for parking a vehicle.
In order to achieve the purpose, the invention provides the following technical scheme:
a visual parking space tracking system comprises a processing module, a camera module and a vehicle mileage module, wherein the processing module comprises a global model unit and a parking space matching unit, a global coordinate system is configured in the global model unit, the vehicle mileage module is arranged on a vehicle, the vehicle mileage module comprises a parking trigger unit, a parking instruction and vehicle coordinate information are generated and sent to the global model unit when the parking trigger unit is triggered, the vehicle coordinate information represents a coordinate position of a rear axle center of the vehicle in the global coordinate system when the vehicle starts to park, and the global model unit takes the coordinate position of the vehicle coordinate information as an origin of the global coordinate system;
the system comprises a camera module, a global model unit and a parking space matching unit, wherein the camera module shoots a parking space in real time and generates a visual parking space, the camera module sends the visual parking space to the global model unit and the parking space matching unit, and when the global model unit receives the visual parking space, the global model unit displays the visual parking space in a global coordinate system with vehicle coordinate information as an origin so that a coordinate position at a corner of the visual parking space is displayed in the global model unit;
a parking space tracking list is configured in the parking space matching unit, when the parking space matching unit receives a visual parking space for the first time, the parking space tracking list records the visual parking space received for the first time and takes the visual parking space as a tracking parking space, a judgment strategy and a preset matching threshold value are configured in the parking space matching unit, the judgment strategy comprises comparing the type of the visual parking space with the type of the tracking parking space, if the type of the visual parking space is different from the type of the tracking parking space, the judgment strategy generates a preset matching value larger than the matching threshold value, if the type of the visual parking space is consistent with the type of the tracking parking space, whether the visual parking space is matched with the parking space or not is judged, a matching strategy and a matching difference value used for matching the parking space or not are further configured in the parking space matching unit, the matching difference value represents the Euclidean distance value of the visual parking space and the tracking parking space with a parking space entrance sideline, and the matching strategy comprises a difference algorithm for calculating a matching difference value, the matching difference value calculated by the difference algorithm is compared with a matching threshold value, if the matching difference value is greater than the matching threshold value, the visual parking space and the tracking parking space are not matched, and if the matching difference value is less than the matching threshold value, the visual parking space and the tracking parking space are matched.
As a further improvement of the invention, a visual parking space filtering strategy is also configured in the camera module, and the visual parking space filtering strategy comprises the steps of filtering the visual parking space shot by the camera module and having edge distortion and abnormal parking space size, and providing the abnormal visual parking space.
As a further improvement of the present invention, the processing module further includes a parking space processing unit, the parking space processing unit is configured with a visual parking space time stamp and a preset imaging time stamp from the shooting of the camera module to the receiving of the visual parking space, the visual parking space time stamp represents the time value generated by the camera module and transmitted to the global model unit for display, the vehicle mileage module further includes a mileage unit, the parking trigger unit further transmits a parking instruction to the mileage unit when triggered, the mileage unit records the moving route of the vehicle and generates mileage information when receiving the parking instruction, the mileage information includes a route and a mileage time stamp, the mileage route represents the route of the vehicle moving along the global coordinate system, the mileage time stamp represents the time value of the vehicle driving to the mileage route terminal, the mileage unit further generates the vehicle real-time coordinate system when receiving the parking instruction, the real-time vehicle coordinate system takes a rear axle center of a vehicle as an original point, the original point of the real-time vehicle coordinate system moves along a mileage route, a timestamp strategy and a compensation strategy are further configured in the parking space processing unit, the timestamp strategy comprises the steps of searching a mileage timestamp which is relatively displayed in a parking space model when a visual parking space timestamp is searched according to mileage information, generating a real parking space timestamp according to a difference value of the visual parking space timestamp and an imaging timestamp, and the compensation strategy comprises the steps of converting the position of the vehicle in the real-time vehicle coordinate system into a global coordinate system and calculating the real coordinate position of the parking space in the global coordinate system according to the real parking space timestamp and the mileage timestamp.
As a further improvement of the present invention, the mileage information further includes a heading angle of the vehicle and a vehicle speed, the vehicle speed includes a speed component along a coordinate system, the heading angle represents an included angle between a vehicle head and a global coordinate system when the vehicle is running, the compensation policy calculates a real coordinate position of the parking space in the global coordinate system according to a compensation algorithm, and the compensation algorithm is specifically configured to:
wherein: x is the number of temp Representing the X-direction coordinate, y, when converting a position located in the vehicle real-time coordinate system to the global coordinate system temp Representing the Y-direction coordinate, x, when converting a position located in the vehicle real-time coordinate system to the global coordinate system v X-direction coordinate, y, representing the vehicle in the vehicle real-time coordinate system v The coordinate of the vehicle in the Y direction in the real-time coordinate system of the vehicle is shown, alpha is a course angle, X is the coordinate of the X direction at the edge of the parking space, Y is the coordinate of the Y direction at the edge of the parking space, v x Representing the component of velocity along the X-axis, v y Representing the velocity component along the Y-axisAmount, t v Time stamp indicating real parking space, t 0 Representing a mileage timestamp, x 0 X-direction coordinate, y, indicating that the parking space is located in the global coordinate system 0 And the Y-direction coordinate of the parking space in the global coordinate system is shown.
As a further improvement of the present invention, the difference algorithm specifically includes:
x vp ＝(x v0 +x v1 )
y vp ＝(y v0 +y v1 )
x tp ＝(x t0 +x t1 )
y tp ＝(y t0 +y t1 )
wherein: x is the number of vp X-axis coordinate, X, of a parking space entry sideline representing a visual parking space v0 X-axis coordinate, X, of end point at one end of parking space inlet sideline representing visual parking space v1 X-axis coordinate, y, of another end point of parking space entrance sideline representing visual parking space vp Y-axis coordinate, Y, of parking space entrance sideline representing visual parking space v0 Y-axis coordinate, Y, of end point at one end of parking space inlet sideline representing visual parking space v1 Y-axis coordinate, x, of end point at other end of parking space inlet sideline for expressing visual parking space tp X-axis coordinate, X, representing a parking space entry sideline for tracking a parking space t0 X-axis coordinate, X, of a parking space entry sideline one end point representing a tracked parking space t1 X-axis coordinate, y, of another end point of a parking space entrance sideline representing a tracked parking space tp Y-axis coordinate, Y, representing a parking space entry sideline for tracking a parking space r0 X-axis coordinate, y, of an end point of a parking space entrance sideline representing a tracked parking space t1 And d represents a matching difference value.
As a further improvement of the present invention, the parking space matching unit further includes a reserve policy, where the reserve policy includes that when the visual parking space is not matched with a tracked parking space, the visual parking space that is not matched with the tracked parking space is determined as a new parking space, and the visual parking space is recorded in a tracking list.
As a further improvement of the present invention, the processing module further includes a parking space fusion unit, and a parking space fusion policy is further configured in the parking space fusion unit, and the fusion policy recalculates the tracked parking space according to the visual parking space matched with the tracked parking space.
As a further improvement of the present invention, the visual parking space further includes a parking space angle and a parking space size, the parking space angle represents an included angle between the parking space and the global coordinate system, the parking space size includes a parking space length and a parking space width, the fusion policy includes a fusion algorithm for recalculating the tracking parking space, and the fusion algorithm specifically includes:
s vp ＝0.5
s＝s f ×m f
s tp1 ＝(s tp0 +s)×(m f /t f )
c tp ＝s tp1 /(s tp0 +s vp )
c vp ＝s vp /(s tp0 +s vp )
x tp1 ＝c tp ×x tp +c vp ×x vp
y tp1 ＝c tp ×y tp +c vp ×y vp
a tp ＝atan2(y t0 -y t3 ,x t0 -x t3 )
wherein: s vp Score representing the number of times of successful matching, s，m f Number of times of successful matching, s f The matching score coefficient is expressed, the default value is 0.1, c tp Representing the weight coefficient of the tracked parking space, c vp Representing the visual parking space weighting coefficient, s tp1 Mean score, s, representing tracked parking space tp0 Score, t, of the tracking slot f Indicating the total number of matches, x, for the tracked space tp1 Indicating the recalculated X-axis coordinate, y of the tracked vehicle tp1 Indicating the recalculated Y-axis coordinate of the tracked parking space, l tp Indicating the length of the space after recalculating the tracked space, x t0 、x t1 、x t2 、x t3 Respectively, the recalculated X-axis coordinate, y, of the corner of the tracked parking space t0 、y t1 、y t2 、y t3 Respectively, the recalculated Y-axis coordinate, w, of the corner of the tracked vehicle tp Indicating the width of the space after recalculating the tracked space, α tp And the calculated parking space angle of the tracking parking space is represented, and alpha is the initial parking space angle of the tracking parking space.
As a further improvement of the present invention, the processing module further includes a parking space state management unit, and a management policy for managing a parking space state is configured in the parking space state management unit, where the management policy specifically includes:
if the matching times of the tracked parking spaces are 1 and the matching success times are zero, judging that the visual parking spaces are unknown parking spaces;
if the matching times of the tracking parking spaces are larger than 1 and the matching success times are 1, judging that the visual parking spaces are successfully matched once and are in tracking confirmation;
if the matching times of the tracked parking spaces are more than 1 and the matching success times are more than 1, judging that the visual parking spaces are in a stable tracking state and the parking spaces are reliable;
otherwise, the parking space is judged to be an unreliable parking space.
The invention has the beneficial effects that: the parking space is shot in real time through the camera module, a visual parking space which can be displayed in a global coordinate system is formed, when a parking trigger unit arranged in the vehicle mileage module is triggered, the global coordinate system can display vehicle coordinates and mileage lines, the types of the visual parking space and the tracking parking space are judged under the action of the parking space matching unit, whether the visual parking space and the tracking parking space are matched or not is accurately judged, the parking space is tracked, the state of the parking space is accurately judged, and the effect of accurately judging whether the parking space can be used for parking a vehicle or not is achieved.
Drawings
FIG. 1 is a logic diagram of the present invention.
Reference numerals: 1. a processing module; 2. a parking space processing unit; 3. a global model unit; 4. a parking space matching unit; 5. a parking space fusion unit; 6. a parking space state management unit; 7. a camera module; 8. a vehicle mileage module; 9. a parking trigger unit.
Detailed Description
The present invention will be described in further detail with reference to the accompanying drawings and examples. In which like parts are designated by like reference numerals. It should be noted that the terms "front," "back," "left," "right," "upper" and "lower" used in the following description refer to directions in the drawings, and the terms "bottom" and "top," "inner" and "outer" refer to directions toward and away from, respectively, the geometric center of a particular component.
Referring to fig. 1, the visual parking space tracking system according to the embodiment of the present invention includes a processing module 1, a camera module 7, and a vehicle mileage module 8, where the processing module 1 includes a global model unit 3 and a parking space matching unit 4, a global coordinate system is configured in the global model unit 3, the vehicle mileage module 8 is disposed on a vehicle, the vehicle mileage module 8 includes a parking trigger unit 9, when the parking trigger unit 9 is triggered, a parking instruction and vehicle coordinate information are generated, and the parking instruction and the vehicle coordinate information are sent to the global model unit 3, the vehicle coordinate information represents a coordinate position where a rear axle center of the vehicle is located in the global coordinate system when the vehicle starts parking, and the global model unit 3 uses the coordinate position of the vehicle coordinate information as an origin of the global coordinate system;
the camera module 7 shoots the parking space in real time and generates a visual parking space, the camera module 7 sends the visual parking space to the global model unit 3, and when the global model unit 3 receives the visual parking space, the global model unit 3 displays the visual parking space in a global coordinate system with the vehicle coordinate information as an original point, so that the coordinate position of the corner of the visual parking space is displayed in the global model unit 3; still dispose vision parking stall in the camera module 7 and filter the tactics, vision parking stall filters the tactics including the vision parking stall that appears edge distortion and parking stall size abnormity to the vision parking stall that camera module 7 shot and filters and rejects unusual vision parking stall for the vision parking stall that shows is normal parking stall in global coordinate system.
The processing module 1 further comprises a parking space processing unit 2, the parking space processing unit 2 is configured with a visual parking space time stamp and a preset imaging time stamp from shooting to receiving of a visual parking space by the camera module, the visual parking space time stamp represents a time value generated by the camera module 7 and sent to the global model unit for display, the vehicle mileage module 8 further comprises a mileage unit, the parking trigger unit further sends a parking instruction to the mileage unit when triggered, the mileage unit records a moving route of the vehicle and generates mileage information when receiving the parking instruction, the mileage information comprises a mileage route and a mileage time stamp, the mileage route represents a route of the vehicle moving along the global coordinate system, the mileage time stamp represents a time value of the vehicle driving to a mileage route terminal, the mileage unit further generates a vehicle real-time coordinate system when receiving the parking instruction, and the vehicle real-time coordinate system takes a rear axle center of the vehicle as an original point, the original point of the vehicle real-time coordinate system moves along a mileage route, a timestamp strategy and a compensation strategy are also configured in the parking space processing unit 2, the timestamp strategy comprises the steps of searching a mileage timestamp which displays a parking space model when the parking space timestamp is displayed relatively according to mileage information, generating a real parking space timestamp according to the difference value of the vision parking space timestamp and an imaging timestamp, and the compensation strategy comprises the steps of converting the position of a vehicle in the vehicle real-time coordinate system into a global coordinate system and calculating the real coordinate position of a parking space in the global coordinate system according to the real parking space timestamp and the mileage timestamp;
the mileage information also comprises a course angle of the vehicle and the vehicle speed, the vehicle speed comprises a speed component along a coordinate system, the speed component comprises a speed component along an X axis and a speed component along a Y axis, the course angle represents an included angle between a vehicle head and a global coordinate system when the vehicle runs, a compensation strategy calculates a real coordinate position of a parking space in the global coordinate system according to a compensation algorithm, and the compensation algorithm is specifically configured as follows:
wherein: x is the number of temp Representing the X-direction coordinate, y, when converting a position located in the vehicle real-time coordinate system to the global coordinate system temp Representing the Y-direction coordinate, x, when converting a position located in the vehicle real-time coordinate system to the global coordinate system v X-direction coordinate, y, representing the vehicle in the vehicle real-time coordinate system v The coordinate of the vehicle in the Y direction in the real-time coordinate system of the vehicle is shown, alpha is a course angle, X is the coordinate of the X direction at the edge of the parking space, Y is the coordinate of the Y direction at the edge of the parking space, v x Representing a component of velocity along the X-axis, v y Representing the component of velocity along the Y axis, t v Time stamp indicating real parking space, t 0 Representing a mileage timestamp, x 0 X-direction coordinate, y, indicating that the parking space is located in the global coordinate system 0 And the Y-direction coordinate of the parking space in the global coordinate system is represented, so that the accurate coordinates of the four corners of the visual parking space in the global coordinate system can be calculated.
The processing module 1 further comprises a parking space matching unit 4, a parking space tracking list is configured in the parking space matching unit 4, when the parking space matching unit 4 receives a visual parking space for the first time, the parking space tracking list records the visual parking space received for the first time and takes the visual parking space as a tracking parking space, a judgment strategy and a preset matching threshold value are configured in the parking space matching unit 4, the matching threshold value is 1.5, the judgment strategy comprises the step of comparing the type of the visual parking space with the type of the tracking parking space, if the type of the visual parking space is different from the type of the tracking parking space, the judgment strategy generates a preset matching value larger than the matching threshold value, and the preset matching value is 100;
if the type of the visual parking space is consistent with that of the tracking parking space, whether the parking spaces are matched or not is judged, a matching strategy and a matching difference value for judging whether the parking spaces are matched or not are also configured in the parking space matching unit 4, the matching difference value represents an Euclidean distance value of the visual parking space and the tracking parking space along a parking space inlet side line, the matching difference degree of the visual parking space and the tracking parking space is represented according to the Euclidean distance value, the matching strategy comprises a difference algorithm for calculating the matching difference value, and the matching difference value calculated through the difference algorithm is compared with a matching threshold value;
the difference algorithm specifically comprises the following steps:
x vp ＝(x v0 +x v1 )
y vp ＝(y v0 +y v1 )
x tp ＝(x t0 +x t1 )
y tp ＝(y t0 +y t1 )
wherein: x is the number of vp X-axis coordinate, X, of a parking space entry sideline representing a visual parking space v0 X-axis coordinate, X, of end point at one end of parking space inlet sideline representing visual parking space v1 X-axis coordinate, y, of another end point of parking space entrance sideline representing visual parking space vp Y-axis coordinate, Y, of parking space entrance sideline representing visual parking space v0 Y-axis coordinate, Y, of end point at one end of parking space inlet sideline representing visual parking space v1 Y-axis coordinate, x, of end point at other end of parking space inlet sideline for expressing visual parking space tp X-axis coordinate, X, representing a parking space entry sideline for tracking a parking space t0 X-axis coordinate, X, representing endpoint at one end of edge line at entrance of parking space to track parking space t1 X-axis coordinate, y, of another end point of a parking space entrance sideline representing a tracked parking space tp Y-axis coordinate, Y, representing a parking space entry sideline for tracking a parking space t0 X-axis coordinate, y, of an end point of a parking space entrance sideline representing a tracked parking space t1 Representing the X-axis coordinate of the end point at the other end of the side line of the parking space entrance of the tracked parking space, and d representing the matching difference value;
if the matching difference value is larger than the matching threshold value, the visual parking space and the tracking parking space are not matched, and if the matching difference value is smaller than the matching threshold value, the visual parking space and the tracking parking space are matched; the parking space matching unit 4 further comprises a storage strategy, wherein the storage strategy comprises the steps of judging the visual parking space which is not matched with the tracking parking space as a new parking space when the visual parking space is not matched with the tracking parking space, and recording the visual parking space into the tracking list to serve as a new tracking parking space.
Processing module 1 still includes parking stall fusion unit 5, still disposes the parking stall in the parking stall fusion unit 5 and fuses the strategy, fuses the strategy according to the vision parking stall recalculation tracking parking stall that matches with tracking parking stall, and the vision parking stall still includes parking stall angle and parking stall size, and the parking stall angle represents the contained angle of parking stall and global coordinate system, and the parking stall size includes parking stall length and parking stall width, fuses the strategy including the fusion algorithm that is used for recalculating tracking parking stall, and the fusion algorithm specifically is:
s vp ＝0.5
s＝s f ×m f
s tp1 ＝(s tp0 +s)×(m f /t f )
c tp ＝s tp1 /(s tp0 +s vp )
c vp ＝s vp /(s tp0 +s vp )
x tp1 ＝c tp ×x tp +c vp ×x vp
y tp1 ＝c tp ×y tp +c vp ×y vp
a tp ＝atan2(y t0 -y t3 ,x t0 -x t3 )
wherein: s vp A score representing the number of times of successful matching is assigned to the detected visual space, s represents a score representing the number of times of successful matching, m f Number of times of successful matching, s f The matching score coefficient is expressed, the default value is 0.1, c tp Representing the weight coefficient of the tracked parking space, c vp Representing the visual parking space weighting coefficient, s tp1 Mean score, s, representing tracked parking space tp0 Score, t, of the tracking slot f Indicating the total number of matches, x, for the tracked space tp1 Indicating the recalculated X-axis coordinate, y of the tracked vehicle tp1 Indicating the recalculated Y-axis coordinate of the tracked parking space, l tp Indicating the length of the space after recalculating the tracked space, x t0 、x t1 、x t2 、x t3 Respectively, the recalculated X-axis coordinate, y, of the corner of the tracked parking space t0 、y t1 、y t2 、y t3 Respectively, the recalculated Y-axis coordinate, w, of the corner of the tracked vehicle tp Indicating the width of the space after recalculating the tracked space, α tp Representing the recalculated parking space angle of the tracking parking space, wherein alpha is the initial parking space angle of the tracking parking space; therefore, the coordinate values of the four corners of the tracked parking space, the length of the tracked parking space, the width of the tracked parking space and the angle of the tracked parking space can be recalculated.
The processing module 1 further comprises a parking space state management unit 6, a management strategy for managing the parking space state is configured in the parking space state management unit 6, and the management strategy specifically comprises:
if the matching times of the tracked parking spaces are 1 and the matching success times are zero, judging that the visual parking spaces are unknown parking spaces; if the matching times of the tracking parking spaces are larger than 1 and the matching success times are 1, judging that the visual parking spaces are successfully matched once and are in tracking confirmation; if the matching times of the tracked parking spaces are more than 1 and the matching success times are more than 1, judging that the visual parking spaces are in a stable tracking state and the parking spaces are reliable; otherwise, judging the parking space as an unreliable parking space; therefore, the matching condition of the visual parking space and the tracking parking space is judged more accurately, and the judgment accuracy of the parking space is improved.
The working principle and the effect are as follows:
the parking space is shot in real time through the camera module 7, a visual parking space which can be displayed in a global coordinate system is formed, when a parking triggering unit 9 arranged in a vehicle mileage module 8 is triggered, the global coordinate system can display vehicle coordinates and mileage lines, real coordinates of four corners of the visual parking space are calculated through a compensation algorithm, the visual parking space displayed in the global coordinate system is a real visual parking space, types of the visual parking space and a tracking parking space are judged under the action of the parking space matching unit 4, whether the visual parking space is matched with the tracking parking space or not is accurately judged, the parking space is tracked, the state of the parking space is accurately judged, and the effect of accurately judging whether the parking space can be used for parking a vehicle or not is achieved.
The above description is only a preferred embodiment of the present invention, and the protection scope of the present invention is not limited to the above embodiments, and all technical solutions belonging to the idea of the present invention belong to the protection scope of the present invention. It should be noted that modifications and adaptations to those skilled in the art without departing from the principles of the present invention should also be considered as within the scope of the present invention.
Claims (9)
1. The utility model provides a vision parking stall tracking system which characterized in that: the system comprises a processing module (1), a camera module (7) and a vehicle mileage module (8), wherein the processing module (1) comprises a global model unit (3) and a parking space matching unit (4), a global coordinate system is configured in the global model unit (3), the vehicle mileage module (8) is arranged on a vehicle, the vehicle mileage module (8) comprises a parking trigger unit (9), a parking instruction and vehicle coordinate information are generated and sent to the global model unit (3) when the parking trigger unit (9) is triggered, the vehicle coordinate information represents a coordinate position of a vehicle rear axle center in the global coordinate system when the vehicle starts parking, and the global model unit (3) takes the coordinate position of the vehicle coordinate information as an origin of the global coordinate system;
the system comprises a camera module (7), a global model unit (3) and a parking space matching unit (4), wherein the camera module (7) shoots a parking space in real time and generates a visual parking space, the camera module (7) sends the visual parking space to the global model unit (3) and the parking space matching unit (4), and when the global model unit (3) receives the visual parking space, the global model unit (3) displays the visual parking space in a global coordinate system with vehicle coordinate information as an origin so that the coordinate position at the corner of the visual parking space is displayed in the global model unit (3);
a parking space tracking list is configured in the parking space matching unit (4), when the parking space matching unit (4) receives a visual parking space for the first time, the parking space tracking list records the visual parking space received for the first time and takes the visual parking space as a tracking parking space, a judgment strategy and a preset matching threshold value are configured in the parking space matching unit (4), the judgment strategy comprises comparing the type of the visual parking space with the type of the tracking parking space, if the type of the visual parking space is different from the type of the tracking parking space, the judgment strategy generates a preset matching value larger than the matching threshold value, the visual parking space and the tracking parking space are not matched, if the type of the visual parking space is consistent with the type of the tracking parking space, whether the parking space is matched or not is judged, and a matching strategy and a matching difference value used for matching the parking space or not are further configured in the parking space matching unit (4), the matching strategy comprises a difference algorithm used for calculating the matching difference value, the matching difference value calculated through the difference algorithm is compared with a matching threshold value, if the matching difference value is larger than the matching threshold value, the visual parking space and the tracking parking space are not matched, and if the matching difference value is smaller than the matching threshold value, the visual parking space and the tracking parking space are matched.
2. The visual space tracking system of claim 1, wherein: still dispose vision parking stall in camera module (7) and filter the tactics, vision parking stall filters the tactics and filters and rejects unusual vision parking stall including the vision parking stall that appears marginal distortion and parking stall size to the vision parking stall that camera module (7) was shot.
3. The visual space tracking system of claim 1, wherein: the processing module (1) further comprises a parking space processing unit (2), the parking space processing unit (2) is configured with a visual parking space time stamp and a preset imaging time stamp from the shooting of the camera module to the receiving of the visual parking space, the visual parking space time stamp represents the time value generated by the camera module (7) and is sent to the global model unit for display, the vehicle mileage module (8) further comprises a mileage unit, the parking trigger unit is further sent to the mileage unit when triggered, the mileage unit records the moving route of the vehicle and generates mileage information when receiving the parking instruction, the mileage information comprises a mileage route and a mileage time stamp, the mileage route represents the route of the vehicle moving along the global coordinate system, the mileage time stamp represents the time value of the vehicle driving to a mileage route terminal, and the mileage unit further generates a vehicle real-time coordinate system when receiving the parking instruction, the real-time coordinate system of the vehicle takes the center of a rear axle of the vehicle as an original point, the original point of the real-time coordinate system of the vehicle moves along a mileage route, a timestamp strategy and a compensation strategy are further configured in the parking space processing unit (2), the timestamp strategy comprises the steps of searching a mileage timestamp which relatively displays a parking space model when a vision parking space timestamp is obtained according to mileage information, generating a real parking space timestamp according to a difference value of the vision parking space timestamp and an imaging timestamp, and the compensation strategy comprises the steps of converting the position of the vehicle in the real-time coordinate system of the vehicle into a global coordinate system and calculating the real coordinate position of the parking space in the global coordinate system according to the real parking space timestamp and the mileage timestamp.
4. The visual space tracking system of claim 3, wherein: the mileage information further comprises a course angle of the vehicle and a vehicle speed, the vehicle speed comprises a speed component along a coordinate system, the course angle represents an included angle between a vehicle head and a global coordinate system when the vehicle runs, the compensation strategy calculates a real coordinate position of the parking space in the global coordinate system according to a compensation algorithm, and the compensation algorithm is specifically configured as follows:
wherein: x is the number of temp Representing the X-direction coordinate, y, when converting a position located in the vehicle real-time coordinate system to the global coordinate system temp Representing the Y-direction coordinate, x, when converting a position located in the vehicle real-time coordinate system to the global coordinate system v X-direction coordinate, y, representing the vehicle in the vehicle real-time coordinate system v The coordinate of the vehicle in the Y direction in the real-time coordinate system of the vehicle is shown, alpha is a course angle, X is the coordinate of the X direction at the edge of the parking space, Y is the coordinate of the Y direction at the edge of the parking space, v x Representing the component of velocity along the X-axis, v y Representing the component of velocity along the Y axis, t v Time stamp indicating real parking space, t 0 Representing a mileage timestamp, x 0 X-direction coordinate, y, indicating that the parking space is located in the global coordinate system 0 And the Y-direction coordinate of the parking space in the global coordinate system is shown.
5. The visual space tracking system of claim 1, wherein: the difference algorithm specifically comprises the following steps:
x vp ＝(x v0 +x v1 )
y vp ＝(y v0 +y v1 )
x tp ＝(x t0 +x t1 )
y tp ＝(y t0 +y t1 )
wherein: x is a radical of a fluorine atom vp X-axis coordinate, X, of a parking space entry sideline representing a visual parking space v0 X-axis coordinate, X, of end point at one end of parking space inlet sideline representing visual parking space v1 X-axis coordinate, y, of another end point of parking space entrance sideline representing visual parking space vp Parking space entrance for indicating visual parking spaceY-axis coordinate of edge line, Y v0 Y-axis coordinate, Y, of end point at one end of parking space inlet sideline representing visual parking space v1 Y-axis coordinate, x, of end point at other end of parking space inlet sideline for expressing visual parking space tp X-axis coordinate, X, representing a parking space entry sideline for tracking a parking space t0 X-axis coordinate, X, representing endpoint at one end of edge line at entrance of parking space to track parking space t1 X-axis coordinate, y, of another end point of a parking space entrance sideline representing a tracked parking space tp Y-axis coordinate, Y, representing a parking space entry sideline for tracking a parking space t0 X-axis coordinate, y, of an end point of a parking space entrance sideline representing a tracked parking space t1 And d represents a matching difference value.
6. The visual space tracking system of claim 1, wherein: the parking space matching unit (4) further comprises a storage strategy, wherein the storage strategy comprises the steps that when the visual parking space is not matched with the tracking parking space, the visual parking space which is not matched with the tracking parking space is judged as a new parking space, and the visual parking space is recorded into a tracking list.
7. The visual space tracking system of claim 1, wherein: the processing module (1) further comprises a parking space fusion unit (5), a parking space fusion strategy is further configured in the parking space fusion unit (5), and the tracking parking space is recalculated according to the fusion strategy and the visual parking space matched with the tracking parking space.
8. The visual space tracking system of claim 7, wherein: the vision parking stall still includes parking stall angle and parking stall size, the contained angle of parking stall angle representation parking stall and global coordinate system, the parking stall size includes parking stall length and parking stall width, it includes the fusion algorithm that is used for recalculating the tracking parking stall to fuse the strategy, it specifically is to fuse the algorithm:
s vp ＝0.5
s＝s f ×m f
s tp1 ＝(s tp0 +s)×(m f /t f )
c tp ＝s tp1 /(s tp0 +s vp )
c vp ＝s vp /(s tp0 +s vp )
x tp1 ＝c tp ×x tp +c vp ×x vp
y tp1 ＝c tp ×y tp +c vp ×y vp
a tp ＝atan2(y t0 -y t3 ,x t0 -x t3 )
wherein: s vp A score representing the number of times of successful matching is assigned to the detected visual space, s represents a score representing the number of times of successful matching, m f Number of times of successful matching, s f The matching score coefficient is expressed, the default value is 0.1, c tp Representing the weight coefficient of the tracked parking space, c vp Representing the visual parking space weighting coefficient, s tp1 Mean score, s, representing tracked parking space tp0 Score of the parking space, t f Indicating the total number of matches, x, for the tracked space tp1 Indicating the recalculated X-axis coordinate, y of the tracked vehicle tp1 Indicating the recalculated Y-axis coordinate of the tracked parking space, l tp Indicating the length of the space after recalculating the tracked space, x t0 、x t1 、x t2 、x t3 Respectively, the recalculated X-axis coordinate, y, of the corner of the tracked parking space t0 、y t1 、y t2 、y t3 Respectively, the recalculated Y-axis coordinate, w, of the corner of the tracked vehicle tp Indicating the width of the space after recalculating the tracked space, α tp And the calculated parking space angle of the tracking parking space is represented, and alpha is the initial parking space angle of the tracking parking space.
9. The visual space tracking system of claim 1, wherein: the processing module (1) further comprises a parking space state management unit (6), a management strategy for managing the parking space state is configured in the parking space state management unit (6), and the management strategy specifically comprises the following steps:
if the matching times of the tracked parking spaces are 1 and the matching success times are zero, judging that the visual parking spaces are unknown parking spaces;
if the matching times of the tracking parking spaces are larger than 1 and the matching success times are 1, judging that the visual parking spaces are successfully matched once and are in tracking confirmation;
if the matching times of the tracked parking spaces are more than 1 and the matching success times are more than 1, judging that the visual parking spaces are in a stable tracking state and the parking spaces are reliable;
otherwise, the parking space is judged to be an unreliable parking space.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110231087.3A CN113034963B (en) | 2021-03-02 | 2021-03-02 | Vision parking stall tracking system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110231087.3A CN113034963B (en) | 2021-03-02 | 2021-03-02 | Vision parking stall tracking system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113034963A CN113034963A (en) | 2021-06-25 |
| CN113034963B true CN113034963B (en) | 2022-08-02 |
Family
ID=76465560
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110231087.3A Active CN113034963B (en) | 2021-03-02 | 2021-03-02 | Vision parking stall tracking system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113034963B (en) |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| GB201305903D0 (en) * | 2012-04-06 | 2013-05-15 | Xerox Corp | Video-based detector and notifier for short-term parking violation enforcement |
| CN106548486A (en) * | 2016-11-01 | 2017-03-29 | 浙江大学 | A kind of unmanned vehicle location tracking method based on sparse visual signature map |
| CN109697860A (en) * | 2017-10-20 | 2019-04-30 | 上海欧菲智能车联科技有限公司 | Parking stall measure and tracking system and method and vehicle |
Family Cites Families (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US8289189B2 (en) * | 2009-05-11 | 2012-10-16 | Robert Bosch Gmbh | Camera system for use in vehicle parking |
| EP3283972A4 (en) * | 2015-04-17 | 2018-08-29 | General Electric Company | Identifying and tracking vehicles in motion |
| KR101795151B1 (en) * | 2015-10-05 | 2017-12-01 | 현대자동차주식회사 | Parking guidance apparatus and method for vehicle |
| CN106671974B (en) * | 2015-11-10 | 2019-09-20 | 新乡航空工业(集团)有限公司 | A kind of method for detecting parking stalls for Intelligent parking system |
| CN107093328B (en) * | 2017-06-06 | 2020-08-04 | 重庆邮电大学 | Parking lot navigation system and method based on machine vision |
| CN109766757B (en) * | 2018-12-11 | 2023-09-01 | 惠州市德赛西威汽车电子股份有限公司 | Parking space high-precision positioning method and system integrating vehicle and visual information |
| CN110942449B (en) * | 2019-10-30 | 2023-05-23 | 华南理工大学 | Vehicle detection method based on laser and vision fusion |
| CN112092825B (en) * | 2020-08-31 | 2022-04-29 | 英博超算(南京)科技有限公司 | Lane keeping method based on machine learning |
| CN112068475B (en) * | 2020-09-21 | 2021-07-13 | 英博超算(南京)科技有限公司 | Vehicle-mounted multi-domain controller |
-
2021
- 2021-03-02 CN CN202110231087.3A patent/CN113034963B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| GB201305903D0 (en) * | 2012-04-06 | 2013-05-15 | Xerox Corp | Video-based detector and notifier for short-term parking violation enforcement |
| CN106548486A (en) * | 2016-11-01 | 2017-03-29 | 浙江大学 | A kind of unmanned vehicle location tracking method based on sparse visual signature map |
| CN109697860A (en) * | 2017-10-20 | 2019-04-30 | 上海欧菲智能车联科技有限公司 | Parking stall measure and tracking system and method and vehicle |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113034963A (en) | 2021-06-25 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110472496B (en) | Traffic video intelligent analysis method based on target detection and tracking | |
| CN104517111B (en) | Method for detecting lane lines, system, lane departure warning method and system | |
| CN110287905B (en) | Deep learning-based real-time traffic jam area detection method | |
| CN110443225B (en) | Virtual and real lane line identification method and device based on feature pixel statistics | |
| CN104657735B (en) | Method for detecting lane lines, system, lane departure warning method and system | |
| JP4914234B2 (en) | Leading vehicle detection device | |
| CN109871732B (en) | Parking grid identification system and method thereof | |
| US7936283B2 (en) | Method for detecting parking area by using range sensor | |
| CN112349144B (en) | Monocular vision-based vehicle collision early warning method and system | |
| CN110458050B (en) | Vehicle cut-in detection method and device based on vehicle-mounted video | |
| CN106023595A (en) | Abnormal parking determination method, abnormal parking determination device and dome camera | |
| US20210174113A1 (en) | Method for limiting object detection area in a mobile system equipped with a rotation sensor or a position sensor with an image sensor, and apparatus for performing the same | |
| EP2813973B1 (en) | Method and system for processing video image | |
| JP4940177B2 (en) | Traffic flow measuring device | |
| CN102975659A (en) | Automobile night driving pre-warning system and method based on far infrared image | |
| CN106846369A (en) | Vehicular turn condition discrimination method and device based on binocular vision | |
| CN114998452B (en) | Vehicle-mounted camera online calibration method and system | |
| CN113281782A (en) | Laser radar snow point filtering method based on unmanned vehicle | |
| JP4296287B2 (en) | Vehicle recognition device | |
| CN113029185B (en) | Road marking change detection method and system in crowdsourcing type high-precision map updating | |
| CN113034963B (en) | Vision parking stall tracking system | |
| CN114495066A (en) | Method for assisting backing | |
| CN112666573B (en) | Detection method for retaining wall and barrier behind mine unloading area vehicle | |
| WO2022048053A1 (en) | Target tracking method, apparatus, and device, and computer-readable storage medium | |
| CN112991734B (en) | Parking space state detection system of visual parking space |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |