CN106846870A - The intelligent parking system and method for the parking lot vehicle collaboration based on centralized vision - Google Patents

Abstract

The present invention claims a centralized vision-based parking lot-vehicle coordinated intelligent parking system and method. The system consists of a parking lot service center, intelligent vehicle on-board equipment and user mobile terminals. The parking lot service center includes vision sensor group, vehicle tracking, path planning, data management and communication modules. The observation range of the visual sensor group covers all the passing areas in the parking lot; the vehicle tracking module fuses the images collected by the visual sensor group to detect the position and attitude of the vehicle target in the passing area; the path planning module based on the current position, attitude and parking space of the vehicle Plan out the driving route. The on-board equipment includes a communication module and a vehicle control module. The vehicle control module receives the route planned by the service center, and controls the vehicle to automatically drive to the destination according to the route. The user's mobile terminal is used to send request information for entering/exiting the parking lot to the parking lot service center. The invention realizes the automation and intelligentization of vehicles entering and exiting the parking lot.

Description

Parking lot-vehicle collaborative intelligent parking system and method based on centralized vision

technical field

The invention belongs to the field of intelligent transportation and automatic control, in particular to an intelligent parking system and method based on centralized vision-based parking lot-vehicle coordination.

Background technique

Traditional parking lots only realize basic service functions such as hourly charging, parking space inquiry, and license plate recognition, and the level of automation and intelligence is not high. Compared with outdoor parking lots, indoor parking lots often have the characteristics of complex environment, narrow roads, and limited vision, which make it more troublesome for vehicles to find vacant parking spaces when driving in and to find vehicles when driving out. In addition, the current solutions to the parking problem of intelligent vehicles usually only focus on a single intelligent vehicle itself, which makes the intelligent vehicle's ability to perceive the overall situation of the parking lot limited, lacks an effective collaborative relationship between the parking lot and the vehicle, and cannot provide higher quality parking. Serve.

Chinese patent application: Visual Assistance System and Method for Guiding Vehicles (Application No.: 200580045956.2) discloses a visual assistance system and method for guiding vehicles, the system includes a location confirmation receiver for collecting location data for vehicles, and collecting visual data for vehicles vision module. This patent proposes the idea of using the visual sensor to assist positioning, and the position data obtained by GPS and the visual position data obtained by the visual sensor are weighted to obtain the vehicle position, thereby assisting vehicle control. This patent is mainly based on GPS and assisted by machine vision. It is only applicable to outdoor parking lots that can obtain GPS data, and is not suitable for automatic parking requirements in indoor parking lots. Chinese patent application: A method of assisted parking based on a bird's-eye view system (application number: 201310109884.X) discloses a method of assisted parking based on machine vision. The environment carries out image acquisition, and through image processing technology, the environmental information around the vehicle is returned to the driver in the form of images, so that the output results are intuitive. The system relies only on on-board sensors and does not achieve parking through coordination between the vehicle and the parking lot. Chinese patent application: A fully automatic parking system (application number: CN201410215303.5) discloses a fully automatic parking system, which uses the ultrasonic sensor and camera of the environmental information collection system to obtain parking information and plans feasible parking Make relevant instructions, send the instructions to each actuator, and obtain actual feedback signals to constantly correct the planned parking path to achieve closed-loop control of the vehicle. The system also relies only on on-board sensors and does not achieve parking through coordination between the vehicle and the parking lot. Chinese patent application: An intelligent vehicle automatic driving system and method coordinated with an intelligent parking lot (application number: 201610107483.4) discloses an intelligent vehicle automatic driving system and method coordinated with an intelligent parking lot, which realizes the vehicle and parking Collaboration between fields, but the background service system only performs location allocation and map transmission, while environment detection and path planning are performed by vehicles, relying on the support of on-board sensors. Chinese patent application: Parking lot-vehicle collaborative intelligent parking system and method based on distributed vision (application number: 201610228232.1) discloses a parking lot-vehicle collaborative intelligent parking system and method based on distributed vision, by the parking lot service center Cover the parking lot area with distributed vision sensors to detect target motion status and obstacles, allocate parking positions and plan vehicle driving paths. This method adopts a distributed structure, and each visual sensor requires a local tracker module, and the cost will be relatively high when the number of visual sensors is large.

Aiming at the requirement of automatic entry/exit of vehicles in the indoor parking lot, the present invention proposes a centralized vision-based intelligent parking system and method for synergy between the parking lot and the vehicle, which reduces the difficulty of parking and finding a car in a complex parking lot environment. It realizes the automation and intelligence of vehicles entering and exiting the parking lot, and reduces the requirements for system configuration.

Contents of the invention

The present invention aims to solve the above problems of the prior art. A parking lot-vehicle based on centralized vision is proposed to reduce the difficulty of parking and finding cars in a complex parking lot environment, realize the automation and intelligence of vehicles entering and leaving the parking lot, and reduce the requirements for system configuration and cost. Coordinated intelligent parking system and method. Technical scheme of the present invention is as follows:

A parking lot-vehicle collaborative smart parking system based on centralized vision, including a parking lot service center, smart vehicle on-board equipment and user mobile terminals,

The parking lot service center includes a visual sensor group, a vehicle tracking module, a path planning module, a data management module and a parking lot communication module, and the visual sensor group is used to collect image sequences of the parking lot passing area and transmit them to the vehicle tracking module in real time module; the vehicle tracking module receives the image sequence from the visual sensor group, obtains the global image in the parking lot through back projection transformation, registration and splicing, and detects the vehicle target position and obstacles in the parking lot passing area on the global image Information to obtain a real-time map of the parking lot; the path planning module is used to plan the driving path of the vehicle in real time according to the current position, the terminal point and the real-time map of the parking lot of the smart vehicle; the data management module is used to query, allocate and update the parking space data, And store the real-time map of the parking lot, the target position of the moving vehicle and its driving path; the parking lot communication module is used to receive the entry/exit request from the user's mobile terminal, and send the driving path planned by the path planning module to the smart vehicle ;

The on-board equipment of the intelligent vehicle includes two parts, a vehicle communication module and a vehicle control module, the vehicle communication module is used to receive the driving path planned by the path planning module from the parking lot service center; the vehicle control module is based on the received parking lot service The path planned by the center controls the intelligent vehicle to automatically drive to the end point according to the path;

The user mobile terminal has a user terminal communication module, which realizes sending request information for entering/exiting the parking lot to the parking lot service center, and can receive processing results sent by the parking lot service center.

Further, the vehicle tracking module receives and processes the image sequence from the visual sensor group, specifically including the following steps:

2.1 Receive a real-time image sequence from the visual sensor group;

2.2 Perform distortion correction on the images obtained by different visual sensors received by using the internal parameters obtained by the corresponding visual sensor through parameter calibration, and then perform inverse projection transformation to uniformly transform the viewing angles of each image to the overlooking angle;

2.3 Registration and splicing are performed on the images after step 2.2 correction processing and back projection transformation to obtain the global image in the parking lot;

2.4 Detect the target vehicle from the global image and calculate its position and attitude angle, perform tracking filtering to obtain the optimal estimation of the target vehicle's motion state; detect obstacles from the global image, and generate a real-time map of the parking lot according to the target vehicle and obstacles.

Further, when the intelligent vehicle enters the parking lot, the end point planned by the path planning module is the parking space allocated to the vehicle by the parking lot service center; when the intelligent vehicle exits the parking lot, the end point planned by the path planning module is the exit location of the parking lot requested by the user.

Furthermore, during the process of the vehicle entering the parking lot, the data management module searches for the free parking space and its coordinates closest to the vehicle to be parked according to the given parking space allocation method (such as the closest to the destination), and completes Then update the free parking space information; in the process of driving out of the parking lot, the data management module searches for the corresponding vehicle parking position according to the target vehicle information, finds the coordinates of the exit on the real-time map according to the exit position, and updates the free parking space information after the exit task is completed .

A parking lot-vehicle coordinated intelligent parking method based on centralized vision of the system, comprising the following steps when the intelligent vehicle enters the parking lot:

5.1 When the smart vehicle drives to the entrance of the parking lot, the owner uses the user's mobile terminal to send the vehicle's current location and entry request to the parking lot service center;

5.2 The parking lot service center receives the entry request sent by the user's mobile terminal in step 5.1 through the parking lot communication module, and after identity verification, forwards the vehicle's current location information and the entry request to the data management module, which allocates free parking spaces , to determine the location of the free parking space;

5.3 The visual sensor group of the parking lot service center collects the image sequence of the parking lot passing area, transmits it to the vehicle tracking module for processing in real time, detects the position, posture and obstacle position of the vehicle target in the parking lot passing area, and constructs a real-time map of the parking lot ;

5.4 The path planning module of the parking lot service center takes the current location of the smart vehicle as the starting point and the allocated parking space as the end point, and plans the driving path of the smart vehicle according to the real-time map of the parking lot in step 5.3;

5.5 The communication module of the parking lot service center sends the current position, posture and planned vehicle driving path of the smart vehicle to the smart vehicle, and sends the current position of the smart vehicle and the allocated parking space to the user's mobile terminal;

5.6 The communication module of the on-board equipment of the intelligent vehicle receives the current position, attitude and the planned vehicle driving path of the intelligent vehicle in step 5.5;

5.7 The vehicle control module of the on-board equipment of the intelligent vehicle controls the automatic driving of the intelligent vehicle according to the information received by the communication module in step 5.6;

5.8 Repeat steps 5.3 to 5.7 until the smart vehicle enters the designated parking space, and the entering task is over.

Further, when the intelligent vehicle drives out, the method includes the following steps:

6.1 The owner uses the user's mobile terminal to send the vehicle identity, exit request and exit location to the parking lot service center;

6.2 The communication module of the parking lot service center receives the exit request, and after identity verification, forwards the vehicle identity, exit request and exit location information to the data management module to find the parking location of the vehicle;

6.3 The visual sensor group of the parking lot service center collects the image sequence of the parking lot passing area, transmits it to the vehicle tracking module for processing in real time, detects the position, posture and obstacle position of the vehicle target in the parking lot passing area, and constructs a real-time map of the parking lot ;

6.4 The route planning module of the parking lot service center takes the current position of the smart vehicle as the starting point and the exit position as the end point, and plans the driving route of the smart vehicle according to the real-time map of the parking lot in step 6.3;

6.5 The communication module of the parking lot service center sends the current position, posture and planned vehicle driving path of the smart vehicle to the smart vehicle, and sends the current position of the smart vehicle to the user's mobile terminal;

6.6 The communication module of the on-board equipment of the intelligent vehicle receives the current position, attitude and the planned vehicle driving path of the intelligent vehicle in step 6.5;

6.7 The vehicle control module of the on-board equipment of the intelligent vehicle controls the automatic driving of the intelligent vehicle according to the information received by the communication module in step 6.6;

6.8 Repeat steps 4.3 to 6.7 until the intelligent vehicle reaches the exit, and the exit task is over.

Further, the step 2.2 inverse projection transforms uniformly transforms the angles of view of each image to the angle of view; it includes the step of: using the external parameters of the camera to complete the geometric transformation of the image coordinates to convert it into an overhead view. The external parameters include: the vertical angle of view of the camera 2α, camera horizontal angle of view 2β, camera pitch angle θ, camera installation height h, let the resolution of the camera be m*n, then the road surface coordinates in the world coordinate system and the road surface coordinates in the image coordinate system (x, y, 0) and The relationship between the image coordinates (u, v) in the image coordinate system is:

Taking the center of the image as the coordinate origin to invert the above formula, there are:

The inverse projection transformed image can be obtained by geometrically transforming the source image using the inversion formula.

Further, the vehicle tracking module of the parking lot service center has a vehicle positioning and multi-target tracking module, comprising the following steps:

(1) Establish a background model using global images of continuous multiple frames;

(2) Use the background difference method to detect the vehicle target, and initialize the search window according to the detection result;

(3) Initialization of the Kalman filter;

(4) Utilize the initialization state information or the state information of the previous moment to predict the target position of the vehicle at the current moment;

(5) Use the adaptive mean shift algorithm to iteratively search for the vehicle target within a certain range near the predicted value until the best matching target is found;

(6) Use the evaluation parameters to evaluate the search results to determine whether there is occlusion or color interference in the tracking target. If the search results are wrong, use the predicted value to replace the search results; otherwise, use the search results as the tracking results;

(7) If the pixel difference between the tracking target and the predicted target is greater than T pixels in consecutive M frames, it will be marked as a new target, and a new search box and Kalman filter will be assigned to the new target, so as to achieve multi-target tracking;

(8) Convert the above-mentioned image coordinates to world coordinates, save to the data management module, and send to the path planning module;

(9) Use the tracking result to update the state of the Kalman filter, repeat steps (4), (5), (6), (7), and (8), and continue to track the vehicle target.

Advantage of the present invention and beneficial effect are as follows:

The invention proposes a parking lot-vehicle coordinated intelligent parking system and method based on centralized vision. This system uses the mobile communication network to realize the interconnection between the intelligent vehicle and the background service center, the user's mobile terminal and the background service center, uses the background service center of the parking lot to detect the movement status of each vehicle, plans the vehicle driving path, and combines the automatic driving function of the intelligent vehicle. It realizes the automation and intelligence of vehicles entering and exiting the parking lot, and reduces the difficulty of parking and finding cars in complex parking lot environments. Reduce parking system configuration requirements by centralizing images from multiple vision sensors.

Description of drawings

Fig. 1. the overall structure of the intelligent parking system of parking lot-vehicle coordination of the present invention based on centralized vision;

Fig. 2. the present invention realizes the flow process of intelligent vehicle driving into the parking lot automatically;

Fig. 3. the present invention realizes the flow process that the intelligent vehicle drives out of the parking lot automatically;

Fig. 4. the vision sensor group setting effect of parking lot service center of the present invention;

Fig. 5. installation diagram of a single visual sensor in the visual sensor group of the parking lot service center of the present invention;

Figure 6. The global image generation process of the vehicle tracking module of the parking lot service center of the present invention;

Figure 7. The vehicle detection and tracking process of the vehicle tracking module of the parking lot service center of the present invention.

detailed description

Below in conjunction with accompanying drawing, the present invention will be further described:

The present invention proposes a parking lot-vehicle collaborative intelligent parking system and method based on centralized vision. The mobile communication network is used to realize the interconnection between the intelligent vehicle-parking service center and the user mobile terminal-parking service center. The advanced visual sensor group and vehicle tracking module detect the movement status of each vehicle, plan the driving path of each intelligent vehicle, combine the automatic driving function of the intelligent vehicle, guide the automatic driving of the intelligent vehicle, and finally realize the automatic driving of the intelligent vehicle into/out of the parking lot.

The specific implementation manners of the present invention will be described below in conjunction with the accompanying drawings and specific examples.

As shown in Fig. 1, it is an overall architecture of a parking lot-vehicle cooperative intelligent parking system based on centralized vision proposed by the present invention. The system consists of three parts: the parking lot service center, the on-board equipment of the intelligent vehicle and the user's mobile terminal.

The parking lot service center includes a visual sensor group, a vehicle tracking module, a path planning module, a data management module and a communication module. The visual sensor group is used to collect the image sequence of the parking lot passing area and transmit it to the vehicle tracking module in real time. The vehicle tracking module receives and processes the image sequence from the visual sensor group, detects the position, posture and obstacle position of the vehicle target in the parking area in real time, and constructs a real-time map of the parking lot. The path planning module plans the driving path according to the real-time map of the parking lot, the current location of the smart vehicle and the destination. The destination of the smart vehicle when it enters the parking lot is the location of the parking space allocated for the vehicle by the parking lot service center, and the smart vehicle drives out of the parking lot. The end point of time is the exit position of the parking lot requested by the user. The data management module is used to query, allocate and update the parking space data, and store the real-time map of the parking lot, the position of the moving vehicle target and its driving path. When the vehicle enters the parking lot, the data management module finds the appropriate free parking space and its coordinates according to the given parking space allocation method, and updates the information of the free parking space; when the vehicle leaves the parking lot, the data management module searches for the parking position according to the vehicle information , find the coordinates of the exit on the real-time map according to the exit position, and update the free parking space information after the exit task is completed. The communication module is used to receive the entry/exit request from the user's mobile terminal, send the planned driving route to the smart vehicle, and send the current position and end position of the smart vehicle to the user's mobile terminal.

The on-board equipment of the intelligent vehicle includes a communication module and a vehicle control module. The communication module is used to receive the planned driving path of the vehicle from the parking lot service center; the vehicle control module controls the intelligent vehicle to automatically drive to the destination according to the path planned by the parking lot service center.

The user mobile terminal is used to send request information for entering/exiting the parking lot to the parking lot service center, and receive the processing result sent by the parking lot service center: when driving in, it is the current position of the intelligent vehicle and the allocated parking space position, When driving out, it is the current position and exit position information of the smart vehicle.

As shown in Fig. 2, it is the flow process that the present invention realizes that the intelligent vehicle drives into the parking lot automatically, including the following steps:

(1) When the smart vehicle drives to the entrance of the parking lot, the owner uses the user's mobile terminal to send the vehicle's current location and entry request to the parking lot service center;

(2) The parking lot service center receives the entry request sent by the user's mobile terminal in step (1) through the communication module, and after identity verification, forwards the vehicle's current location information and the entry request to the data management module, which distributes Empty parking space, determine the location of the vacant parking space;

(3) The visual sensor group of the parking lot service center collects the image sequence of the parking lot passing area, transmits it to the vehicle tracking module for processing in real time, detects the position, posture and obstacle position of the vehicle target in the parking lot passing area, and constructs the parking lot real-time map;

(4) The path planning module of the parking lot service center takes the current position of the smart vehicle as the starting point, and takes the allocated parking space position as the end point, and plans the driving path of the smart vehicle according to the real-time map of the parking lot in step (3);

(5) The communication module of the parking lot service center sends the current position, posture and planned vehicle driving path of the smart vehicle to the smart vehicle, and sends the current position of the smart vehicle and the allocated parking space to the user's mobile terminal;

(6) the current position, attitude and the vehicle travel path of the intelligent vehicle in the intelligent vehicle vehicle equipment communication module receiving step (5);

(7) The vehicle control module of the vehicle-mounted equipment of the intelligent vehicle controls the automatic driving of the intelligent vehicle according to the information received by the communication module in step (6);

(8) Steps (3) to (7) are repeated until the intelligent vehicle enters the designated parking space, and the entering task ends.

As shown in Fig. 3, it is the flow process that the present invention realizes that the intelligent vehicle drives out of the parking lot automatically, including the following steps:

(1) The owner uses the user's mobile terminal to send the vehicle identity, exit request and exit location to the parking lot service center;

(2) The communication module of the parking lot service center receives the exit request, and after identity verification, forwards the vehicle identity, exit request and exit location information to the data management module to find the parking location of the vehicle;

(3) The visual sensor group of the parking lot service center collects the image sequence of the parking lot passing area, transmits it to the vehicle tracking module for processing in real time, detects the position, posture and obstacle position of the vehicle target in the parking lot passing area, and constructs the parking lot real-time map;

(4) The path planning module of the parking lot service center takes the current position of the smart vehicle as the starting point, and takes the exit position as the end point, and plans the driving path of the smart vehicle according to the real-time map of the parking lot in step (3);

(5) The communication module of the parking lot service center sends the current position, posture and planned vehicle driving path of the intelligent vehicle to the intelligent vehicle, and sends the current position and exit position of the intelligent vehicle to the user's mobile terminal;

(6) the current position, attitude and the vehicle travel path of the intelligent vehicle in the intelligent vehicle vehicle equipment communication module receiving step (5);

(7) The vehicle control module of the vehicle-mounted equipment of the intelligent vehicle controls the automatic driving of the intelligent vehicle according to the information received by the communication module in step (6);

(8) Steps (3)-(7) are repeated until the intelligent vehicle reaches the exit, and the exit task ends.

As shown in Fig. 4, it is an example of the setting effect of the visual sensor group in the parking lot service center of the present invention.

The visual sensor group includes a plurality of visual sensors installed at different locations in the parking lot, and the observation ranges of each visual sensor partially overlap, but the union of the observation ranges of multiple visual sensors covers all the entrance and exit areas and passing areas of the parking lot . This example is only used to describe the effect of the vision sensor group setting, and is not limited to the scene in the picture.

Fig. 5 is a schematic diagram showing the installation of a single visual sensor in the visual sensor group of the parking lot service center of the present invention. Assume that the origin of the global coordinate system of the parking lot is O(0,0,0), and the coordinates of the visual sensor C as an example are (x _c , y _c , h). The visual sensor C includes external parameters: 2α (vertical viewing angle), 2β (horizontal viewing angle), θ (pitch angle), h (height). Among them, the installation height h is fixed, and other parameters can be obtained through calibration.

Figure 6 shows the global image generation process of the vehicle tracking module of the parking lot service center of the present invention, including the following steps:

(1) Receive images collected by the visual sensor group, and perform distortion correction on images from each visual sensor;

(2) Perform inverse projection transformation on the images from each visual sensor to convert them into a top view. Assuming that the resolution of the visual sensor C is m*n, according to the imaging principle of the visual sensor, it corresponds to the difference between the coordinates (x, y, 0) in the world coordinate system of the visual sensor C and the coordinates (u, v) in the image coordinate system The relationship between is:

From the above formulas (1)-(2), the inverse projection transformation relationship can be solved:

(3) According to the homography matrix between the top view images obtained by the inverse projection transformation, the images are mapped to the global coordinate system of the parking lot one by one, and the fusion and boundary smoothing are performed to obtain the global image.

Figure 7 shows the vehicle detection and tracking process of the vehicle tracking module of the parking lot service center of the present invention, including the following steps:

(1) Initialization: use continuous multi-frame global images to establish a background model, set a total of [180/R] templates with vehicle attitude angles from -90° to 90°, where R is the resolution of the template group, and initialize the Kalman filter device;

(2) Use the background difference method to detect the vehicle target, and initialize the search window according to the detection result;

(3) Utilize the initialization state information or the state information of the last moment to predict the position of the vehicle target at the current moment according to the Kalman filter;

(4) Use Continuously Adaptive Mean-SHIFT (CamShift) to iteratively search for the vehicle target within a certain range near the predicted value until the best matching target is found;

(5) Use the evaluation parameters to evaluate the search results to determine whether there is serious occlusion or color interference in the tracking target. If the search results are wrong, use the predicted value to replace the search results; otherwise, use the search results as the tracking results;

(6) If the pixel difference between the tracking target and the predicted target is greater than T pixels in consecutive M frames, it will be marked as a new target, and a new search box and Kalman filter will be assigned to the new target, so as to achieve Tracking of multiple vehicle targets;

(7) The image of each detected vehicle target is matched with the attitude angle template, and the angle corresponding to the best matching template is the current attitude angle of the vehicle;

(8) Use the tracking result to update the state of the Kalman filter,

(9) Repeat steps (3)-(9) to continue tracking the vehicle target.

The above embodiments should be understood as only for illustrating the present invention but not for limiting the protection scope of the present invention. After reading the contents of the present invention, skilled persons can make various changes or modifications to the present invention, and these equivalent changes and modifications also fall within the scope defined by the claims of the present invention.

Claims (5)

1. the intelligent parking system that a kind of parking lot based on centralized vision-vehicle is cooperateed with, it is characterised in that：Including parking lot Service centre, intelligent vehicle mobile unit and the part of customer mobile terminal three, wherein

The parking lot service centre includes vision sensor group, vehicle tracking module, path planning module, data management module And communication module, vision sensor group is for gathering the image sequence of parking lot traffic areas and being real-time transmitted to vehicle tracking Module；The vehicle tracking module receives the image sequence from vision sensor group, by Inverse projection, registration and spelling Connect, obtain the global image in parking lot, position, the attitude of the vehicle target of parking lot traffic areas are detected on global image And the position of barrier, construct parking lot real-time map；Path planning module is current according to parking lot real-time map, intelligent vehicle Position and terminal cook up driving path；Data management module is used to that parking stall data to be inquired about, distributed and updated, and deposits Storage parking lot real-time map, the position of moving vehicle target and its driving path；Communication module is used to receive from user's movement Request is sailed/rolled away to terminal into, and the driving path of planning is sent into intelligent vehicle；

The intelligent vehicle mobile unit includes communication module and vehicle control module, and communication module is used to receive from parking lot This car driving path of the planning of service centre；Vehicle control module controls intelligent vehicle according to the path of the planning for receiving By the path automatic Pilot to terminal；

The customer mobile terminal is used to be sent to parking lot service centre the solicited message for sailing/rolling away from parking lot into, and receives The result that parking lot service centre sends.

2. the intelligent parking system that the parking lot based on centralized vision according to claim 1-vehicle is cooperateed with, its feature It is：Can realize that intelligent vehicle sails the specified location in parking lot into from parking automatically outside the venue, and be sailed automatically from parking lot Go out to specified outlet.

3. the intelligent parking system that the parking lot based on the centralized vision-vehicle of system according to claim 1,2 is cooperateed with, Characterized in that, the intelligent vehicle method of sailing into is comprised the following steps：

3.1 when intelligent vehicle drives to Entrance, and car owner is sent using customer mobile terminal to parking lot service centre Current vehicle position and sail request into；

3.2 parking lot service centres sail request into by what customer mobile terminal in communication module receiving step 3.1 sent, carry out After authentication, by current vehicle position information and sail request into and be transmitted to data management module, by data management module point With idle parking stall, it is determined that idle parking stall position；

3.3 parking lot service centre vision sensor groups gather the image sequence of parking lot traffic areas, are real-time transmitted to vehicle Tracking module is processed, and detects the position of the position, attitude and barrier of the vehicle target of parking lot traffic areas, and is constructed Parking lot real-time map；

With intelligent vehicle current location as starting point, the parking stall position with distribution is 3.4 parking lot service centre path planning modules Terminal, the parking lot real-time map planning intelligent vehicle driving path according to step 3.3；

3.5 parking lot service centre communication modules are by the vehicle running path of the current location of intelligent vehicle, attitude and planning Intelligent vehicle is sent to, and intelligent vehicle current location is sent to customer mobile terminal with the parking stall position of distribution；

The current location of the intelligent vehicle in 3.6 intelligent vehicle vehicle equipment communication module receiving steps 3.5, attitude and planning Vehicle running path；

The information that 3.7 intelligent vehicle mobile unit vehicle control modules are received according to communication module in step 3.6 controls intelligence Vehicular automatic driving；

3.8 repeat steps 3.3~3.7, until intelligent vehicle sails specified parking stall into, the task of sailing into terminates.

4. the intelligent parking system that the parking lot based on the centralized vision-vehicle of system according to claim 1,2 is cooperateed with, Characterized in that, intelligent vehicle rolls method away from comprising the following steps：

4.1 car owners send testing vehicle register, roll request and outlet port away from using customer mobile terminal to parking lot service centre；

4.2 parking lot service centre communication modules are received rolls request away from, and after carrying out authentication, by testing vehicle register, rolling away from please Ask and exits location information is transmitted to data management module, search storing cycle position；

4.3 parking lot service centre vision sensor groups gather the image sequence of parking lot traffic areas, are real-time transmitted to vehicle Tracking module is processed, and detects the position of the position, attitude and barrier of the vehicle target of parking lot traffic areas, and is constructed Parking lot real-time map；

4.4 parking lot service centre path planning modules with intelligent vehicle current location as starting point, with outlet port as terminal, root According to the parking lot real-time map planning intelligent vehicle driving path of step 4.3；

4.5 parking lot service centre communication modules are by the vehicle running path of the current location of intelligent vehicle, attitude and planning Intelligent vehicle is sent to, and intelligent vehicle current location and outlet port are sent to customer mobile terminal；

The current location of the intelligent vehicle in 4.6 intelligent vehicle vehicle equipment communication module receiving steps 4.5, attitude and planning Vehicle running path；

The information that 4.7 intelligent vehicle mobile unit vehicle control modules are received according to communication module in step 4.6 controls intelligence Vehicular automatic driving；

4.8 repeat steps 4.3~4.7, export until intelligent vehicle is reached, and the task of rolling away from terminates.

5. according to claim 1 and 2 the parking lot service centre vehicle tracking module of system method for tracking target, it is special Levy and be, comprised the following steps in each process cycle：

5.1 receive the real-time image sequences from vision sensor group；

Each image of 5.2 pairs of different vision sensors acquisitions for receiving is utilized respectively correspondence vision sensor and passes through parameter calibration The inner parameter of acquisition carries out distortion correction, then carries out Inverse projection, and the unification of each image aspects is transformed into depression angle；

5.3 pairs carry out registration and splicing by each image after step 5.2 correction process and Inverse projection, obtain in parking lot Global image；

5.4 detect target vehicle from global image, are tracked the optimal estimation that filtering obtains target vehicle motion state；From Barrier is detected in global image, the real-time map in parking lot is generated according to target vehicle and barrier.