CN108389421A - The accurate inducible system in parking lot and method identified again based on image - Google Patents
The accurate inducible system in parking lot and method identified again based on image Download PDFInfo
- Publication number
- CN108389421A CN108389421A CN201810168950.3A CN201810168950A CN108389421A CN 108389421 A CN108389421 A CN 108389421A CN 201810168950 A CN201810168950 A CN 201810168950A CN 108389421 A CN108389421 A CN 108389421A
- Authority
- CN
- China
- Prior art keywords
- vehicle
- parking stall
- image
- sent
- license plate
- 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.)
- Granted
Links
- 230000001939 inductive effect Effects 0.000 title claims abstract description 11
- 238000000034 method Methods 0.000 title claims abstract description 10
- 238000001514 detection method Methods 0.000 claims abstract description 34
- 238000012544 monitoring process Methods 0.000 claims abstract description 22
- 238000012423 maintenance Methods 0.000 claims abstract description 6
- 238000005516 engineering process Methods 0.000 claims description 17
- 238000012549 training Methods 0.000 claims description 17
- 238000013480 data collection Methods 0.000 claims description 14
- 238000012545 processing Methods 0.000 claims description 7
- 238000013459 approach Methods 0.000 claims description 4
- 230000005540 biological transmission Effects 0.000 claims description 4
- 239000003607 modifier Substances 0.000 claims description 4
- 239000000284 extract Substances 0.000 claims description 3
- 238000009434 installation Methods 0.000 claims 1
- 230000004913 activation Effects 0.000 description 3
- 239000000686 essence Substances 0.000 description 2
- 238000013507 mapping Methods 0.000 description 2
- 238000001228 spectrum Methods 0.000 description 2
- 241000626238 Cepora Species 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 238000013527 convolutional neural network Methods 0.000 description 1
- 230000002349 favourable effect Effects 0.000 description 1
- 238000011478 gradient descent method Methods 0.000 description 1
- 238000007689 inspection Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000010606 normalization Methods 0.000 description 1
- 230000009466 transformation Effects 0.000 description 1
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
- G08G1/141—Traffic control systems for road vehicles indicating individual free spaces in parking areas with means giving the indication of available parking spaces
- G08G1/142—Traffic control systems for road vehicles indicating individual free spaces in parking areas with means giving the indication of available parking spaces external to the vehicles
-
- 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
- G08G1/017—Detecting movement of traffic to be counted or controlled identifying vehicles
- G08G1/0175—Detecting movement of traffic to be counted or controlled identifying vehicles by photographing vehicles, e.g. when violating traffic rules
Abstract
The present invention provides a kind of accurate inducible system in the parking lot identified again based on image and method, system include:Camera module, car plate detection module, path planning module, vehicle tracking module, parking stall monitoring mould, high in the clouds processing center, level-one display screen and multiple two level navigational panels.Parking lot may be implemented the present invention is based on the accurate inducible system in parking lot that image identifies again, and accurately and effectively parking stall is distributed, and is unfavorable for parking lot later maintenance, debugging and upgrading, is saved cost.
Description
Technical field
The present invention relates to instruction parking location technical field more particularly to a kind of parking lot essences identified again based on image
True inducible system and method.
Background technology
Current truck space guiding system commonly use ultrasonic sensor, the sensors such as magnetic detector and according to its from
The operation principle of body monitors the parking space state on each parking stall in parking lot in real time, and statistics parking position uses
Condition information.In recent years, the method for also occurring carrying out single-point detection to parking stall using camera, single-point detection mode have inspection
The advantages that measured data is accurate, real-time is high, domestic most of parking lots use this data acquisition modes.However this method
Have that cost is higher, and construction period is longer, it is larger to existing parking lot transformation difficulty, while be unfavorable for later maintenance, debugging and
The shortcomings of upgrading.
Prior art can not carry out accurately and effectively parking stall distribution, while being unfavorable for later maintenance, debugging and upgrading etc. and lacking
Point.
Invention content
The present invention provides a kind of accurate inducible system in the parking lot identified again based on image, to overcome above-mentioned technical problem.
The present invention is based on the accurate inducible systems in parking lot that image identifies again, including:
Camera module, car plate detection module, path planning module, vehicle tracking module, parking stall measure mould, high in the clouds processing
Center, level-one display screen and multiple two level navigational panels;
The camera module includes:Entrance camera, two level navigational panel camera and parking stall camera;
The Entrance camera be used for acquire enter parking lot vehicle portal image and be sent to the car plate detection module,
Vehicle tracking module, the two level navigational panel camera is for being installed on two level navigational panel bottom, for acquiring apart from institute
It states the real time picture of vehicle in two level navigational panel threshold range and is sent to the vehicle tracking module, the parking stall camera
It is installed on the front on parking stall, for acquiring the parking stall image in the parking stall camera shooting area and being sent to described
Parking stall monitoring modular;
The car plate detection module, for receiving portal image that the Entrance camera is sent and identifying vehicle
License plate number, the license plate number and the portal image are sent to high in the clouds processing center;
The path planning module, for receiving the license plate number and portal image and sky that the high in the clouds processing center is sent
The position of not busy parking stall generates the traveling road of vehicle according to the position of Entrance and the idle parking stall using shortest path
Line, and the travel route of the license plate number, the portal image and the vehicle is sent to the vehicle tracking module, it will
The license plate number and the travel route are sent to the vehicle tracking module, the high in the clouds processing center, the travel route
Including:The license plate number of vehicle, each sections of road direction, the corresponding two level navigational panel of each sections of road distance and each section are compiled
Number;
The vehicle tracking module, for receiving vehicle realtime graphic, the path rule that the two level navigational panel is sent
The license plate number of module transmission, the travel route of the portal image and the vehicle are drawn, is schemed in real time according to the vehicle
Picture and the portal image judge to exercise to whether the vehicle in two level navigational panel region is same vehicle, if so, to the cloud
End processing center sends the license plate number, the number of the two level navigational panel and parking stall position;
The parking stall monitoring modular, the parking stall image sent for receiving the parking stall camera, reads the high in the clouds
The parking space information table of processing center, the parking space information table include:Garage map, parking stall numbering, parking space, parking stall shape
State, the license plate number of corresponding parking stall and the entrance picture, the parking space state include:Free time has occupied and has distributed, root
Current parking space state is determined according to the parking stall image and changes the parking space information table, and modified parking space information table is sent
To high in the clouds processing center;
The high in the clouds processing center, for receiving the license plate number and portal image that the parking stall monitoring modular is sent, to
The path planning module sends the position of idle parking stall, receives the modified parking space information that the parking stall monitoring modular is sent
Table, and parking space information record sheet is changed according to the parking space information table, and by the idle parking stall in the parking space information record sheet
Quantity is sent to the level-one LED display;
The license plate number that the path planning module is sent is received, and idle parking stall is sent to the path planning module, is connect
Receive the corresponding travel route of the license plate number that the path planning module is sent;
The license plate number, two level navigational panel number and parking stall position that the vehicle tracking module is sent are received,
And the idle parking stall nearest apart from two level navigational panel position is searched according to two level navigational panel number, judge described nearest
Whether idle parking stall idle parking stall corresponding with travel route is consistent, if so, judging that next two level navigational panel number corresponds to
Parking stall position and the nearest idle parking stall it is whether consistent, if it is not, the nearest free time parking stall is corresponded to the license plate number, shape
State is revised as having occupied, and the nearest idle parking stall is sent and the license plate number is sent to path planning module;
The level-one display screen is set to Entrance, receives the idle parking stall quantity that the high in the clouds processing center is sent
And show the idle parking stall quantity;
The two level navigational panel is set at the crossing of parking lot, for showing license plate number and the corresponding institute of the license plate number
State the travel direction of travel route.
Further, the vehicle tracking module is specifically used for:
Vehicle data collection is obtained in the picture for from database including vehicle;
It builds image and identifies network again;
Identify that network training image identifies network model again again using image according to the vehicle data collection;
The vehicle pictures in the portal image are extracted according to vehicle characteristics point;
Identify network model by the vehicle in the vehicle and database in portal image again according to the image after the training
Image compares to obtain the probability value of the vehicle pictures;
Judge whether the probability value is more than threshold value, if so, determining that vehicle is same vehicle, if not, it is determined that vehicle
For two vehicles.
Further, the parking stall monitoring modular is specifically used for:
Detect the position of vehicle under camera;
Calculate the overlapping area of vehicle and parking stall position;
Using image recognition technology, judge whether current vehicle vehicle corresponding with parking space information table is same vehicle,
If so, by the license plate number, parking space information and parking space state are sent to high in the clouds processing center, if it is not, then image is used to know
Other technology determines the corresponding license plate number of current vehicle, and the license plate number, parking space information and parking space state is sent to described
High in the clouds processing center.
Whether the vehicle and distribution to the vehicle of the current parking stall for judging current parking stall are same vehicle, if so, root
The status modifier of current parking stall described in parking space state table has been occupied according to the corresponding license plate number of the vehicle, parking space information,
If it is not, then changing parking space information and parking space state according to license plate number.
The present invention also provides a kind of accurate abductive approach in the parking lot identified again based on image, including:
The portal image of Entrance camera collection vehicle, and the portal image is sent to vehicle detection mould
Block;
The vehicle detection module receives the portal image, identifies the license plate number of the vehicle, and by the license plate number
It is sent to high in the clouds processing center with the portal image;
The high in the clouds processing center receives the license plate number and the portal image, searches the idle vehicle in parking space state table
Position sends the position of the license plate number, the portal image and idle parking stall to path planning module;
The path planning module receives the position of the license plate number, the portal image and idle parking stall, according to stopping
The position of parking lot entrance and the idle parking stall generates the travel route of the vehicle using shortest path, and by the car plate
Number, the portal image and the travel route be sent to vehicle tracking module and the high in the clouds processing center;
The realtime graphic of two level navigational panel camera acquisition vehicle in threshold range, and the realtime graphic is sent out
It send to the vehicle tracking module;
The vehicle tracking module receive the license plate number that the high in the clouds processing center sends, the portal image and
The travel route receives the realtime graphic that the two level navigational panel camera is sent, according to the portal image and institute
It states realtime graphic and license plate number judges whether the vehicle is same vehicle, if so, to described in the transmission of high in the clouds processing center
License plate number, two level navigational panel number and idle parking stall position;
The high in the clouds processing center is numbered according to the two level navigational panel searches current time apart from the two level navigational panel
The nearest idle parking stall in position, judge nearest idle parking stall free time parking stall corresponding with the travel route whether one
It causes, if so, being to have occupied by the status modifier of idle parking stall described in the parking space state table, if it is not, then will be described nearest
The position of idle parking stall, license plate number be sent to the path planning module;
The path planning module regenerates vehicle according to the position of the nearest idle parking stall using shortest path
Travel route.
Further, the vehicle tracking module judges according to the portal image and the realtime graphic and license plate number
Whether the vehicle is same vehicle, including:
Vehicle tracking module obtains vehicle data collection from the picture that database includes vehicle;
The vehicle tracking module builds image and identifies network again;
The vehicle tracking module identifies that network training image identifies net again again according to the vehicle data collection using image
Network model;
The vehicle tracking module extracts the vehicle in portal image according to vehicle characteristics point;
The vehicle tracking module identifies network model by the vehicle in portal image again according to the image after the training
It compares to obtain the probability value of the vehicle pictures with the vehicle image in database;
Judge whether the probability value is more than as if so, determining that vehicle is same vehicle, if not, it is determined that vehicle
For two vehicles.
Further, the vehicle detection module receives the portal image, identifies the license plate number of the vehicle, including:
Vehicle detection module monitors the position of vehicle under camera;
The vehicle detection module calculates the overlapping area of vehicle and parking stall position;
The vehicle detection module judges current vehicle vehicle corresponding with parking space information table using image recognition technology
Whether it is same vehicle, if so, the corresponding license plate number of the vehicle, parking space information and parking space state are sent at high in the clouds
Reason center if it is not, then determining the corresponding license plate number of current vehicle using image recognition technology, and the license plate number, parking stall is believed
Breath and parking space state are sent to the high in the clouds processing center.
Further, the high in the clouds processing center is numbered according to the two level navigational panel searches current time apart from described two
The nearest idle parking stall in grade navigational panel position judges the nearest idle parking stall idle parking stall corresponding with the travel route
Whether it is consistent after, further include:
The parking stall monitoring modular, the parking stall image sent for receiving the parking stall camera, reads the high in the clouds
The parking space information table of processing center, the parking space information table include:Garage map, parking stall numbering, parking space, parking stall shape
State, the license plate number of corresponding parking stall and the portal image, the parking space state include:Free time has occupied and has distributed, root
Current parking space state is determined according to the parking stall image and changes the parking space information table, and modified parking space information table is sent
To high in the clouds processing center.
The present invention is based on the accurate inducible systems in parking lot that image identifies again, and parking lot accurately and effectively parking stall may be implemented
Distribution, is unfavorable for parking lot later maintenance, debugging and upgrading, saves cost.
Description of the drawings
In order to more clearly explain the embodiment of the invention or the technical proposal in the existing technology, to embodiment or will show below
There is attached drawing needed in technology description to be briefly described, it should be apparent that, the accompanying drawings in the following description is this hair
Some bright embodiments for those of ordinary skill in the art without having to pay creative labor, can be with
Obtain other attached drawings according to these attached drawings.
Fig. 1 is that the present invention is based on the accurate inducible systems in parking lot that image identifies again;
Fig. 2 is that the present invention is based on the accurate abductive approach flow charts in parking lot that image identifies again.
Specific implementation mode
In order to make the object, technical scheme and advantages of the embodiment of the invention clearer, below in conjunction with the embodiment of the present invention
In attached drawing, technical scheme in the embodiment of the invention is clearly and completely described, it is clear that described embodiment is
A part of the embodiment of the present invention, instead of all the embodiments.Based on the embodiments of the present invention, those of ordinary skill in the art
The every other embodiment obtained without creative efforts, shall fall within the protection scope of the present invention.
Fig. 1 be the present invention is based on the accurate inducible system in parking lot that image identifies again, as shown in Figure 1, the present embodiment system,
Including:
Camera module 101, car plate detection module 102, path planning module 103, vehicle tracking module 104, parking stall prison
Survey module 105, high in the clouds processing center 106, level-one display screen 107 and multiple two level navigational panels 108;
The camera module includes:Entrance camera, two level navigational panel camera and parking stall camera;
The Entrance camera be used for acquire enter parking lot vehicle portal image and be sent to the car plate detection module,
Vehicle tracking module, the two level navigational panel camera is for being installed on two level navigational panel bottom, for acquiring apart from institute
It states the real time picture of vehicle in two level navigational panel threshold range and is sent to the vehicle tracking module, the parking stall camera
It is installed on the front on parking stall, for acquiring the parking stall image in the parking stall camera shooting area and being sent to described
Parking stall monitoring modular;
The car plate detection module, for receiving portal image that the Entrance camera is sent and identifying vehicle
License plate number, the license plate number and the portal image are sent to high in the clouds processing center;
The path planning module, for receiving the license plate number and portal image and sky that the high in the clouds processing center is sent
The position of not busy parking stall generates the traveling road of vehicle according to the position of Entrance and the idle parking stall using shortest path
Line, and the travel route of the license plate number, the portal image and the vehicle is sent to the vehicle tracking module, it will
The license plate number and the travel route are sent to the vehicle tracking module, the high in the clouds processing center, the travel route
Including:The license plate number of vehicle, each sections of road direction, the corresponding two level navigational panel of each sections of road distance and each section are compiled
Number;
The vehicle tracking module, for receiving vehicle realtime graphic, the path rule that the two level navigational panel is sent
The license plate number of module transmission, the travel route of the portal image and the vehicle are drawn, is schemed in real time according to the vehicle
Picture and the portal image judge to exercise to whether the vehicle in two level navigational panel region is same vehicle, if so, to the cloud
End processing center sends the license plate number, the number of the two level navigational panel and parking stall position;
The parking stall monitoring modular, the parking stall image sent for receiving the parking stall camera, reads the high in the clouds
The parking space information table of processing center, the parking space information table include:Garage map, parking stall numbering, parking space, parking stall shape
State, the license plate number of corresponding parking stall and the entrance picture, the parking space state include:Free time has occupied and has distributed, root
Current parking space state is determined according to the parking stall image and changes the parking space information table, and modified parking space information table is sent
To high in the clouds processing center;
The high in the clouds processing center, for receiving the license plate number and portal image that the parking stall monitoring modular is sent, to
The path planning module sends the position of idle parking stall, receives the modified parking space information that the parking stall monitoring modular is sent
Table, and parking space information record sheet is changed according to the parking space information table, and by the idle parking stall in the parking space information record sheet
Quantity is sent to the level-one LED display;
The license plate number that the path planning module is sent is received, and idle parking stall is sent to the path planning module, is connect
Receive the corresponding travel route of the license plate number that the path planning module is sent;
The license plate number, two level navigational panel number and parking stall position that the vehicle tracking module is sent are received,
And the idle parking stall nearest apart from two level navigational panel position is searched according to two level navigational panel number, judge described nearest
Whether idle parking stall idle parking stall corresponding with travel route is consistent, if so, judging that next two level navigational panel number corresponds to
Parking stall position and the nearest idle parking stall it is whether consistent, if it is not, the nearest free time parking stall is corresponded to the license plate number, shape
State is revised as having occupied, and the nearest idle parking stall is sent and the license plate number is sent to path planning module;
The level-one display screen is set to Entrance, receives the idle parking stall quantity that the high in the clouds processing center is sent
And show the idle parking stall quantity;The two level navigational panel is set at the crossing of parking lot, receives the high in the clouds processing center
The license plate number, the travel direction for showing license plate number and the corresponding travel route of the license plate number sent.
Specifically, after vehicle enters parking lot, the portal image of the vehicle, two level are acquired by Entrance camera
The real time picture of navigational panel camera acquisition vehicle in the two level navigational panel threshold range, Entrance camera should
Portal image is sent to vehicle detection module, which is sent to vehicle tracking module, vehicle by two level navigational panel camera
Detection module identifies the license plate number of the vehicle, label of the license plate number as the vehicle according to the portal image.Vehicle detection
The license plate number of the vehicle and the portal image are sent to high in the clouds processing center by module.High in the clouds processing center receive license plate number with
And after portal image, the idle parking stall in parking space state table is searched, license plate number, portal image and sky are sent to path planning module
The position of not busy parking stall.After path planning module receives the position of the license plate number, portal image and idle parking stall, according to parking lot
The position of entrance and the free time parking stall generates the travel route of vehicle using shortest path, and the travel route is sent to vehicle
Tracking module, high in the clouds processing center.Vehicle tracking module receives portal image, realtime graphic, license plate number and the traveling of vehicle
Route judges whether the vehicle is same vehicle according to portal image and realtime graphic and license plate number, if so, at high in the clouds
The license plate number, the number of two level navigational panel and idle parking stall position are sent in reason, are taken the photograph if it is not, current vehicle does not appear in this
As under head, continuing to detect the vehicle.High in the clouds processing center is numbered according to the two level navigational panel searches current time apart from the two level
The nearest idle parking stall in navigational panel position judges that the idle parking stall of current time recently idle parking stall corresponding with travel route is
It is no consistent, if so, idle parking space state in parking space state table is revised as having occupied, the corresponding license plate number.And by the free time
Parking stall, license plate number are sent to path planning module, and path planning module will be according to current idle parking stall position, using shortest path
The travel route of vehicle is generated again.Wherein, two level navigational panel auxiliary camera be ranked up according to the distance apart from entrance and
Parking stall camera is ranked up according to distance outlet distance.Table 1 is parking space state table.
Table 1
Wherein, the position of parking stall can be found in garage map according to parking space and parking stall numbering again.
When each two level navigational panel that the vehicle is passed through before reaching idle parking stall, can all it repeat the above, it is right
Travel route is modified, to improve the efficiency that vehicle enters parking stall.
Further, the vehicle tracking module is specifically used for:
Vehicle data collection is obtained in the picture for from database including vehicle;
It builds image and identifies network again;
Identify that network training image identifies network model again again using image according to the vehicle data collection;
The vehicle pictures in the portal image are extracted according to vehicle characteristics point;
Identify network model by the vehicle in the vehicle and database in portal image again according to the image after the training
Image compares to obtain the probability value of the vehicle pictures;
Judge whether the probability value is more than threshold value, if so, determining that vehicle is same vehicle, if not, it is determined that vehicle
For two vehicles.
Further, the parking stall monitoring modular is specifically used for:
Detect the position of vehicle under camera;
Calculate the overlapping area of vehicle and parking stall position;
Using image recognition technology, judge whether current vehicle vehicle corresponding with parking space information table is same vehicle,
If so, by the license plate number, parking space information and parking space state are sent to high in the clouds processing center, if it is not, then image is used to know
Other technology determines the corresponding license plate number of current vehicle, and the license plate number, parking space information and parking space state is sent to described
High in the clouds processing center.
Specifically, detect vehicle using the realtime graphic under the camera of parking stall, frame and parking stall position are drawn to vehicle
Frame compares overlapping region, and overlapping region area is more than threshold value, it is determined that parking stall is occupied.
Above-mentioned vehicle tracking module identifies network model by the vehicle in portal image again according to the image after the training
It compares to obtain the probability value of the vehicle pictures with the vehicle image in database and parking stall monitoring modular uses image recognition
Technology, judge current vehicle vehicle corresponding with parking space information table whether be same vehicle the specific steps are:
Vehicle data collection is divided into three parts:Positive sample, negative sample, target sample;It is created from Stanford University
Cars data sets, 196 class vehicle pictures, totally 16185 pictures.The CompCars numbers that Tang Xiao gulls team of Hong Kong Chinese University establishes
According to collection, including the picture of 136727 vehicle vehicle pictures and 27618 vehicle parts.Several data are chosen as target sample
This is not used as negative sample with the conduct positive sample that target sample is same vehicle for same vehicle;Then, image is built again
Identify network.
The image identify again network include 9 convolutional layers, 3 pond layers, 10 ReLU layer, criticize standardize (batch
Normalization) layer and 3 Dropout layers;Every layer of convolutional layer and batch processing layer alternate links, and ReLU layers are connected
In convolutional layer top, full articulamentum is connected after the last one pond layer, and its training is prevented using Dropout layers and ReLU layers
Over-fitting;In convolutional layer:The Feature Mapping figure of each convolutional layer can use multiple convolution kernels, pass through what is obtained to preceding layer
Feature Mapping figure carries out convolution operation, and then combination obtains;In the layer of pond:The N number of characteristic spectrum exported in the layer of pond is defeated with it
The N number of characteristic spectrum entered is corresponding.
First layer convolution z1=W1*I+B1, wherein I is input picture, and size is 256 × 256, W1Represent f1×f1×n1Filter
Wave device, convolution kernel size f1=3, step-length 1, convolution kernel number is n1=64, that is, export n1A characteristic pattern, * indicate convolution behaviour
Make, B1Represent n1Tie up deviation.z1Indicate the characteristic pattern exported after this layer.256 × 256 image passes through this convolution, generates 64
The characteristic pattern that size is 254 × 254.
Second layer convolution z2=W2*F1+B2, wherein W2It representsFilter, convolution kernel size f2=3, volume
Product core number is n2=128, step-length 1 exports n2A characteristic pattern, * indicate convolution operation, B2Represent n2Tie up deviation.F1Table
Show this layer of input feature vector figure, z2Indicate the characteristic pattern exported after this layer.254 × 254 characteristic patterns pass through this convolution, generate
The characteristic pattern that 128 sizes are 252 × 252.
Third layer convolution z3=W3*F2+B3, wherein W3It representsFilter, convolution kernel size f3=3,
Convolution kernel number is n3=256, step-length 1 exports n3A characteristic pattern, * indicate convolution operation, B3Represent n3Tie up deviation.F2
Indicate this layer of input feature vector figure, z3Indicate the characteristic pattern exported after this layer.252 × 252 characteristic patterns pass through this convolution, generate
The characteristic pattern that 256 sizes are 250 × 250.
4th layer of convolution z4=W4*F3+B4, wherein W4It representsFilter, convolution kernel size f4=3,
Convolution kernel number is n4=512, step-length 1 exports n4A characteristic pattern, * indicate convolution operation, B4Represent n4Tie up deviation.F3
Indicate this layer of input feature vector figure, z4Indicate the characteristic pattern exported after this layer.250 × 250 characteristic patterns pass through this convolution, generate
The characteristic pattern that 512 sizes are 248 × 248.
There is one layer of pond behind the 4th layer of convolutionz4It indicates the characteristic pattern inputted after this layer, rolls up
Product core size fp=3, stride indicate step-length, step-length 2.There is one layer of Maxout activation behind pondI is
This node layer, j are the implicit node corresponding to this layer of each node.The value range of j is [1, k], required for k is maxout layers
Parameter.P4Indicate the characteristic pattern of this layer of input, F4Indicate that the characteristic pattern of this layer of output, final output characteristic pattern number areImage pass through this pond, generate 512 sizes as 124 × 124 characteristic pattern.
Layer 5 convolution z5=W5*F4+B5, wherein W5It representsFilter, convolution kernel size f5=3,
Convolution kernel number is n5=512, step-length 1 exports n5A characteristic pattern, * indicate convolution operation, B5Represent n5Tie up deviation.F4
Indicate this layer of input feature vector figure, z5Indicate the characteristic pattern exported after this layer.124 × 124 characteristic patterns pass through this convolution, generate
The characteristic pattern that 512 sizes are 122 × 122.
Layer 6 convolution z6=W6*F5+B6, wherein W6It representsFilter, convolution kernel size f6=3,
Convolution kernel number is n6=512, step-length 1 exports n6A characteristic pattern, * indicate convolution operation, B6Represent n6Tie up deviation.F5
Indicate this layer of input feature vector figure, z6Indicate the characteristic pattern exported after this layer.122 × 122 characteristic patterns pass through this convolution, generate
The characteristic pattern that 512 sizes are 120 × 120.
Layer 7 convolution z7=W7*F6+B7, wherein W7It representsFilter, convolution kernel size f7=3,
Convolution kernel number is n7=512, step-length 1 exports n7A characteristic pattern, * indicate convolution operation, B7Represent n7Tie up deviation.F6
Indicate this layer of input feature vector figure, z7Indicate the characteristic pattern exported after this layer.120 × 120 characteristic patterns pass through this convolution, generate
The characteristic pattern that 512 sizes are 118 × 118.
There is one layer of pond behind layer 7 convolutionz7It indicates the characteristic pattern inputted after this layer, rolls up
Product core size fp=3, stride indicate step-length, step-length 2.There is one layer of Maxout activation behind pondI is
This node layer, j are the implicit node corresponding to this layer of each node.The value range of j is [1, k], required for k is maxout layers
Parameter.p7Indicate the characteristic pattern of this layer of input, F7Indicate that the characteristic pattern of this layer of output, final output characteristic pattern number areA characteristic pattern passes through this convolution, generates the characteristic pattern that 512 sizes are 59 × 59.
8th layer of convolution z8=W8*F7+B8, wherein W8It representsFilter, convolution kernel size f8=3,
Convolution kernel number is n8=512, step-length 1 exports n8A characteristic pattern, * indicate convolution operation, B8Represent n8Tie up deviation.F7
Indicate this layer of input feature vector figure, z8Indicate the characteristic pattern exported after this layer.59 × 59 characteristic patterns pass through this convolution, generate
The characteristic pattern that 512 sizes are 57 × 57.
9th layer of convolution z9=W9*F8+B9, wherein W9It representsFilter, convolution kernel size f9=3,
Convolution kernel number is n9=512, step-length 1 exports n9A characteristic pattern, * indicate convolution operation, B9Represent n9Tie up deviation.F8
Indicate this layer of input feature vector figure, z9Indicate the characteristic pattern exported after this layer.57 × 57 characteristic patterns pass through this convolution, generate
The characteristic pattern that 512 sizes are 55 × 55.
There is one layer of pond behind the 9th layer of convolutionz9It indicates the characteristic pattern inputted after this layer, rolls up
Product core size fp=3, stride indicate step-length, step-length 2.There is one layer of Maxout activation behind pondI is
This node layer, j are the implicit node corresponding to this layer of each node.The value range of j is [1, k], required for k is maxout layers
Parameter.p9Indicate the characteristic pattern of this layer of input, F9Indicate that the characteristic pattern of this layer of output, final output characteristic pattern number areA characteristic pattern passes through this convolution, generates the characteristic pattern that 512 sizes are 27 × 27.
After the completion of last layer of pond layer processing, by positive sample, the characteristic pattern that negative sample and target sample obtain connects respectively
Enter to two full articulamentums, generate respectively the feature vector that 1 size is 4096 × 1 and the feature that 1 size is 512 × 1 to
Amount, the full articulamentum of first layer are followed by batch processing layer, ReLU layer with Dropout layers.Characteristic pattern that the full articulamentum of the second layer obtains and
The formula of sample characteristics counting loss, counting loss is:
Wherein,Indicate target sample,Indicate positive sample,Indicate that negative sample, threshold indicate specific threshold.
Identify that network training image identifies network model again again using image according to the vehicle data collection, specifically:
After initialization, network model is iterated training to be identified again to the image of above-mentioned structure using stochastic gradient descent method, per iteration
The value of one subgradient of one-time detection and loss function, to obtain the optimal of each weighted value W and bias b in network architecture
Solution, the optimum image that this training is obtained after iteration is multiple identify network model again.It is examined using optimal convolutional neural networks model
It surveys and compares the vehicle pictures preserved in the vehicle and database server in realtime graphic that LED navigational panel cameras return one by one
It is right, export the probability that the vehicle pictures preserved in vehicle and database server in realtime graphic are same vehicles.If probability
More than the probability value of setting, then the maximum vehicle of probability value is determined as same vehicle, then makees the license plate number of secondary vehicle
It is stored in database for the label of the vehicle.
Present system by vehicle tracking module combination two level navigational panel to the traveling real-time condition of vehicle into line trace,
To be adjusted in real time to idle parking stall, the working efficiency of distribution parking stall is improved.Improve the utilization rate of parking stall.It supervises parking stall
Real-time monitoring of the module for parking stall is surveyed, the accuracy rate of parking stall distribution is improved.Accurate, the efficient distribution of the parking stall of realization, together
When be additionally favorable for the maintenance in parking lot.
Fig. 2 is that the present invention is based on the accurate abductive approach flow charts in parking lot that image identifies again, as shown in Fig. 2, including:
The portal image of step 101, Entrance camera collection vehicle, and the portal image is sent to vehicle
Detection module;
Step 102, the vehicle detection module receive the portal image, identify the license plate number of the vehicle, and by institute
It states license plate number and the portal image is sent to high in the clouds processing center;
Step 103, the high in the clouds processing center receive the license plate number and the portal image, search in parking space state table
Idle parking stall, the position of the license plate number, the portal image and idle parking stall is sent to path planning module;
Step 104, the path planning module receive the position of the license plate number, the portal image and idle parking stall
It sets, generates the travel route of the vehicle using shortest path according to the position of Entrance and the idle parking stall, and will
The license plate number, the portal image and the travel route are sent to vehicle tracking module and the high in the clouds processing center;
The realtime graphic of step 105, two level navigational panel camera acquisition vehicle in threshold range, and by the reality
When image be sent to the vehicle tracking module;
Step 106, the vehicle tracking module receive the license plate number, the entrance that the high in the clouds processing center is sent
Image and the travel route receive the realtime graphic that the two level navigational panel camera is sent, according to the entrance
Image and the realtime graphic and license plate number judge whether the vehicle is same vehicle, if so, to high in the clouds processing center
Send the license plate number, two level navigational panel number and idle parking stall position;
Step 107, the high in the clouds processing center are numbered according to the two level navigational panel searches current time apart from described two
The nearest idle parking stall in grade navigational panel position judges the nearest idle parking stall idle parking stall corresponding with the travel route
It is whether consistent, if so, being to have occupied by the status modifier of idle parking stall described in the parking space state table, if it is not, then by institute
State the position of nearest idle parking stall, license plate number is sent to the path planning module;
Step 108, the path planning module use shortest path again according to the position of the nearest idle parking stall
Generate the travel route of vehicle.
Further, the vehicle tracking module judges according to the portal image and the realtime graphic and license plate number
Whether the vehicle is same vehicle, including:
Vehicle tracking module obtains vehicle data collection from the picture that database includes vehicle;
The vehicle tracking module builds image and identifies network again;
The vehicle tracking module identifies that network training image identifies net again again according to the vehicle data collection using image
Network model;
The vehicle tracking module extracts the vehicle in portal image according to vehicle characteristics point;
The vehicle tracking module identifies network model by the vehicle in portal image again according to the image after the training
It compares to obtain the probability value of the vehicle pictures with the vehicle image in database;
Judge whether the probability value is more than as if so, determining that vehicle is same vehicle, if not, it is determined that vehicle
For two vehicles.
Further, the vehicle detection module receives the portal image, identifies the license plate number of the vehicle, including:
Vehicle detection module monitors the position of vehicle under camera;
The vehicle detection module calculates the overlapping area of vehicle and parking stall position;
The vehicle detection module judges current vehicle vehicle corresponding with parking space information table using image recognition technology
Whether it is same vehicle, if so, the corresponding license plate number of the vehicle, parking space information and parking space state are sent at high in the clouds
Reason center if it is not, then determining the corresponding license plate number of current vehicle using image recognition technology, and the license plate number, parking stall is believed
Breath and parking space state are sent to the high in the clouds processing center.
Further, the high in the clouds processing center is numbered according to the two level navigational panel searches current time apart from described two
The nearest idle parking stall in grade navigational panel position judges the nearest idle parking stall idle parking stall corresponding with the travel route
Whether it is consistent after, further include:
The parking stall monitoring modular, the parking stall image sent for receiving the parking stall camera, reads the high in the clouds
The parking space information table of processing center, the parking space information table include:Garage map, parking stall numbering, parking space, parking stall shape
State, the license plate number of corresponding parking stall and the portal image, the parking space state include:Free time has occupied and has distributed, root
Current parking space state is determined according to the parking stall image and changes the parking space information table, and modified parking space information table is sent
To high in the clouds processing center.
Method shown in Fig. 2 is identical as System Working Principle shown in FIG. 1, and details are not described herein again.
Finally it should be noted that:The above embodiments are only used to illustrate the technical solution of the present invention., rather than its limitations;To the greatest extent
Present invention has been described in detail with reference to the aforementioned embodiments for pipe, it will be understood by those of ordinary skill in the art that:Its according to
So can with technical scheme described in the above embodiments is modified, either to which part or all technical features into
Row equivalent replacement;And these modifications or replacements, various embodiments of the present invention technology that it does not separate the essence of the corresponding technical solution
The range of scheme.
Claims (7)
1. a kind of accurate inducible system in the parking lot identified again based on image, which is characterized in that including:
In camera module, car plate detection module, path planning module, vehicle tracking module, parking stall monitoring mould, high in the clouds processing
The heart, level-one display screen and multiple two level navigational panels;
The camera module includes:Entrance camera, two level navigational panel camera and parking stall camera;It is described
Entrance camera is used to acquire the portal image for entering parking lot vehicle and is sent to the car plate detection module, vehicle
Tracking module, the two level navigational panel camera is for being installed on two level navigational panel bottom, for acquiring apart from described two
The real time picture of vehicle and it is sent to the vehicle tracking module in grade navigational panel threshold range, parking stall camera installation
Front in parking stall, for acquiring the parking stall image in the parking stall camera shooting area and being sent to the parking stall
Monitoring modular;
The car plate detection module, for receiving the portal image and the vehicle for identifying vehicle that the Entrance camera is sent
The license plate number and the portal image are sent to high in the clouds processing center by the trade mark;
The path planning module, for receiving the license plate number and portal image and idle vehicle that the high in the clouds processing center is sent
The position of position generates the travel route of vehicle according to the position of Entrance and the idle parking stall using shortest path, and
The travel route of the license plate number, the portal image and the vehicle is sent to the vehicle tracking module, it will be described
License plate number and the travel route are sent to the vehicle tracking module, the high in the clouds processing center, and the travel route includes:
The license plate number of vehicle, each sections of road direction, the corresponding two level navigational panel number of each sections of road distance and each section;
The vehicle tracking module, for receiving vehicle realtime graphic, the path planning mould that the two level navigational panel is sent
The travel route of the license plate number, the portal image and the vehicle that block is sent, according to the vehicle realtime graphic and
The portal image judges to exercise to whether the vehicle in two level navigational panel region is same vehicle, if so, at the high in the clouds
Reason center sends the license plate number, the number of the two level navigational panel and parking stall position;
The parking stall monitoring modular, the parking stall image sent for receiving the parking stall camera read the high in the clouds processing
The parking space information table at center, the parking space information table include:Garage map, parking stall numbering, parking space, parking space state,
The license plate number and the portal image, the parking space state of corresponding parking stall include:Free time has occupied and has distributed, according to
The parking stall image determines current parking space state and changes the parking space information table, and modified parking space information table is sent to
High in the clouds processing center;
The high in the clouds processing center, for receiving the license plate number and portal image that the parking stall monitoring modular is sent, to described
Path planning module sends the position of idle parking stall, receives the modified parking space information table that the parking stall monitoring modular is sent,
And parking space information record sheet is changed according to the parking space information table, and by the idle parking stall quantity in the parking space information record sheet
It is sent to the level-one LED display;
The license plate number that the path planning module is sent is received, and idle parking stall is sent to the path planning module, receives institute
State the corresponding travel route of the license plate number of path planning module transmission;
Receive the license plate number, two level navigational panel number and parking stall position that the vehicle tracking module is sent, and root
The idle parking stall nearest apart from two level navigational panel position is searched according to two level navigational panel number, judges the nearest free time
Whether parking stall idle parking stall corresponding with travel route is consistent, if so, judging that next two level navigational panel numbers corresponding vehicle
Whether position position and the nearest idle parking stall are consistent, if it is not, the nearest idle parking stall is corresponded to the license plate number, status maintenance
It is changed to occupy, the nearest idle parking stall is sent and the license plate number is sent to path planning module;
The level-one display screen is set to Entrance, receives the idle parking stall quantity that the high in the clouds processing center is sent and shows
Show the idle parking stall quantity;
The two level navigational panel is set at the crossing of parking lot, for showing license plate number and the corresponding row of the license plate number
Sail the travel direction of route.
2. system according to claim 1, which is characterized in that the vehicle tracking module is specifically used for:
Vehicle data collection is obtained in the picture for from database including vehicle;
It builds image and identifies network again;
Identify that network training image identifies network model again again using image according to the vehicle data collection;
The vehicle pictures in the portal image are extracted according to vehicle characteristics point;
Identify network model by the vehicle image in the vehicle and database in portal image again according to the image after the training
Comparison obtains the probability value of the vehicle pictures;
Judge whether the probability value is more than threshold value, if so, determining that vehicle is same vehicle, if not, it is determined that vehicle two
Vehicle.
3. system according to claim 1, which is characterized in that the parking stall monitoring modular is specifically used for:
Detect the position of vehicle under camera;
Calculate the overlapping area of vehicle and parking stall position;
Using image recognition technology, judge whether current vehicle vehicle corresponding with parking space information table is same vehicle, if so,
The license plate number, parking space information and parking space state are then sent to high in the clouds processing center, if it is not, then using image recognition technology
It determines the corresponding license plate number of current vehicle, and the license plate number, parking space information and parking space state is sent at the high in the clouds
Reason center.
4. a kind of accurate abductive approach in the parking lot identified again based on image, which is characterized in that including:
The portal image of Entrance camera collection vehicle, and the portal image is sent to vehicle detection module;
The vehicle detection module receives the portal image, identifies the license plate number of the vehicle, and by the license plate number and institute
It states portal image and is sent to high in the clouds processing center;
The high in the clouds processing center receives the license plate number and the portal image, searches the idle parking stall in parking space state table,
The position of the license plate number, the portal image and idle parking stall is sent to path planning module;
The path planning module receives the position of the license plate number, the portal image and idle parking stall, according to parking lot
The position of entrance and the idle parking stall generates the travel route of the vehicle using shortest path, and by the license plate number, institute
It states portal image and the travel route is sent to vehicle tracking module and the high in the clouds processing center;
The realtime graphic of two level navigational panel camera acquisition vehicle in threshold range, and the realtime graphic is sent to
The vehicle tracking module;
The vehicle tracking module receives the license plate number that the high in the clouds processing center sends, the portal image and described
Travel route receives the realtime graphic that the two level navigational panel camera is sent, according to the portal image and the reality
When image and license plate number judge whether the vehicle is same vehicle, if so, sending the car plate to high in the clouds processing center
Number, two level navigational panel number and idle parking stall position;
The high in the clouds processing center is numbered according to the two level navigational panel searches current time apart from two level navigational panel position
Nearest idle parking stall judges whether the nearest idle parking stall idle parking stall corresponding with the travel route is consistent, if
It is then to have occupied by the status modifier of idle parking stall described in the parking space state table, if it is not, then by the nearest free time to be
The position of parking stall, license plate number are sent to the path planning module;
The path planning module regenerates the row of vehicle according to the position of the nearest idle parking stall using shortest path
Sail route.
5. according to the method described in claim 4, it is characterized in that, the vehicle tracking module is according to the portal image and institute
It states realtime graphic and license plate number judges whether the vehicle is same vehicle, including:
Vehicle tracking module obtains vehicle data collection from the picture that database includes vehicle;
The vehicle tracking module builds image and identifies network again;
The vehicle tracking module identifies that network training image identifies network mould again again according to the vehicle data collection using image
Type;
The vehicle tracking module extracts the vehicle in portal image according to vehicle characteristics point;
The vehicle tracking module identifies network model by the vehicle sum number in portal image again according to the image after the training
It compares to obtain the probability value of the vehicle pictures according to the vehicle image in library;
Judge whether the probability value is more than as if so, determining that vehicle is same vehicle, if not, it is determined that vehicle two
Vehicle.
6. according to the method described in claim 5, it is characterized in that, the vehicle detection module receives the portal image, knowledge
The license plate number of the not described vehicle, including:
Vehicle detection module monitors the position of vehicle under camera;
The vehicle detection module calculates the overlapping area of vehicle and parking stall position;
Whether the vehicle detection module judges current vehicle vehicle corresponding with parking space information table using image recognition technology
For same vehicle, if so, the corresponding license plate number of the vehicle, parking space information and parking space state are sent in the processing of high in the clouds
The heart, if it is not, then determine the corresponding license plate number of current vehicle using image recognition technology, and by the license plate number, parking space information with
And parking space state is sent to the high in the clouds processing center.
7. according to the method described in claim 4, it is characterized in that, the high in the clouds processing center is compiled according to the two level navigational panel
The current time idle parking stall nearest apart from two level navigational panel position number is searched, judges the idle parking stall and institute recently
State the corresponding idle parking stall of travel route it is whether consistent after, further include:
The parking stall monitoring modular, the parking stall image sent for receiving the parking stall camera read the high in the clouds processing
The parking space information table at center, the parking space information table include:Garage map, parking stall numbering, parking space, parking space state,
The license plate number and the portal image, the parking space state of corresponding parking stall include:Free time has occupied and has distributed, according to
The parking stall image determines current parking space state and changes the parking space information table, and modified parking space information table is sent to
High in the clouds processing center.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810168950.3A CN108389421B (en) | 2018-02-28 | 2018-02-28 | Parking lot accurate induction system and method based on image re-identification |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810168950.3A CN108389421B (en) | 2018-02-28 | 2018-02-28 | Parking lot accurate induction system and method based on image re-identification |
Publications (2)
Publication Number | Publication Date |
---|---|
CN108389421A true CN108389421A (en) | 2018-08-10 |
CN108389421B CN108389421B (en) | 2020-12-22 |
Family
ID=63069158
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201810168950.3A Active CN108389421B (en) | 2018-02-28 | 2018-02-28 | Parking lot accurate induction system and method based on image re-identification |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN108389421B (en) |
Cited By (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109345866A (en) * | 2018-11-26 | 2019-02-15 | 武汉纵横智慧城市股份有限公司 | Methods of exhibiting, device, equipment and the storage medium in Intelligentized parking lot vacancy |
CN109741617A (en) * | 2018-11-12 | 2019-05-10 | 浙江吉利汽车研究院有限公司 | A kind of parking lot is parked air navigation aid and device |
CN110264768A (en) * | 2019-03-18 | 2019-09-20 | 奥菲(泰州)光电传感技术有限公司 | Intelligent Platform of Image Recognition |
CN111275982A (en) * | 2020-01-22 | 2020-06-12 | 于志国 | Illegal vehicle big data reporting system based on lane detection |
CN111785076A (en) * | 2020-08-05 | 2020-10-16 | 杭州华云科技有限公司 | Method for inducing reserved parking spaces of ground parking lot |
CN111833637A (en) * | 2019-04-17 | 2020-10-27 | 阿里巴巴集团控股有限公司 | Vehicle monitoring method, device, equipment and system |
CN112382104A (en) * | 2020-11-13 | 2021-02-19 | 重庆盘古美天物联网科技有限公司 | Roadside parking management method based on vehicle track analysis |
CN112885136A (en) * | 2021-01-27 | 2021-06-01 | 安徽蓝鸟智能停车科技产业化有限公司 | Parking space position dynamic right confirming method and system based on block chain |
CN115641747A (en) * | 2022-12-23 | 2023-01-24 | 中电信数字城市科技有限公司 | Parking navigation method and system and electronic equipment |
Citations (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102944237A (en) * | 2012-11-27 | 2013-02-27 | 北京邮电大学 | Parking guidance method based on Internet of Things |
CN103337196A (en) * | 2013-06-21 | 2013-10-02 | 苏州大方特种车股份有限公司 | Parking stall guiding and reversed car searching method and systems thereof |
KR20130134265A (en) * | 2012-05-30 | 2013-12-10 | (주)에스디시스템 | Light guiding parking system |
CN103985270A (en) * | 2014-04-17 | 2014-08-13 | 北京汉王智通科技有限公司 | Parking lot control system |
US20150009047A1 (en) * | 2013-07-04 | 2015-01-08 | Mordechai ASHKENAZI | Method and apparatus for vehicle parking spaces management using image processing |
CN105096653A (en) * | 2015-08-10 | 2015-11-25 | 电子科技大学 | Parking guidance method and parking guidance system |
CN105913685A (en) * | 2016-06-25 | 2016-08-31 | 上海大学 | Video surveillance-based carport recognition and intelligent guide method |
CN205722455U (en) * | 2016-04-19 | 2016-11-23 | 西南交通大学 | Stopping guide system based on video identification |
CN107730993A (en) * | 2017-11-17 | 2018-02-23 | 大连海事大学 | The parking lot intelligent vehicle-tracing system and method identified again based on image |
-
2018
- 2018-02-28 CN CN201810168950.3A patent/CN108389421B/en active Active
Patent Citations (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
KR20130134265A (en) * | 2012-05-30 | 2013-12-10 | (주)에스디시스템 | Light guiding parking system |
CN102944237A (en) * | 2012-11-27 | 2013-02-27 | 北京邮电大学 | Parking guidance method based on Internet of Things |
CN103337196A (en) * | 2013-06-21 | 2013-10-02 | 苏州大方特种车股份有限公司 | Parking stall guiding and reversed car searching method and systems thereof |
US20150009047A1 (en) * | 2013-07-04 | 2015-01-08 | Mordechai ASHKENAZI | Method and apparatus for vehicle parking spaces management using image processing |
CN103985270A (en) * | 2014-04-17 | 2014-08-13 | 北京汉王智通科技有限公司 | Parking lot control system |
CN105096653A (en) * | 2015-08-10 | 2015-11-25 | 电子科技大学 | Parking guidance method and parking guidance system |
CN205722455U (en) * | 2016-04-19 | 2016-11-23 | 西南交通大学 | Stopping guide system based on video identification |
CN105913685A (en) * | 2016-06-25 | 2016-08-31 | 上海大学 | Video surveillance-based carport recognition and intelligent guide method |
CN107730993A (en) * | 2017-11-17 | 2018-02-23 | 大连海事大学 | The parking lot intelligent vehicle-tracing system and method identified again based on image |
Non-Patent Citations (3)
Title |
---|
XIANPING FU: "A New Fuzzy based Fast and High Definition Template Matching Algorithm", 《2015 IEEE 10TH CONFERENCE ON INDUSTRIAL ELECTRONICS AND APPLICATIONS (ICIEA)》 * |
周思浩: "基于无线传感网络的停车场内智能引导系统", 《中国优秀硕士学位论文全文数据库信息科技辑》 * |
彭锦佳: "摄像头网络中车辆检测和识别方法的研究", 《中国优秀硕士学位论文全文数据库信息科技辑》 * |
Cited By (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109741617A (en) * | 2018-11-12 | 2019-05-10 | 浙江吉利汽车研究院有限公司 | A kind of parking lot is parked air navigation aid and device |
CN109345866A (en) * | 2018-11-26 | 2019-02-15 | 武汉纵横智慧城市股份有限公司 | Methods of exhibiting, device, equipment and the storage medium in Intelligentized parking lot vacancy |
CN110264768A (en) * | 2019-03-18 | 2019-09-20 | 奥菲(泰州)光电传感技术有限公司 | Intelligent Platform of Image Recognition |
CN111833637A (en) * | 2019-04-17 | 2020-10-27 | 阿里巴巴集团控股有限公司 | Vehicle monitoring method, device, equipment and system |
CN111275982A (en) * | 2020-01-22 | 2020-06-12 | 于志国 | Illegal vehicle big data reporting system based on lane detection |
CN111785076A (en) * | 2020-08-05 | 2020-10-16 | 杭州华云科技有限公司 | Method for inducing reserved parking spaces of ground parking lot |
CN112382104A (en) * | 2020-11-13 | 2021-02-19 | 重庆盘古美天物联网科技有限公司 | Roadside parking management method based on vehicle track analysis |
CN112885136A (en) * | 2021-01-27 | 2021-06-01 | 安徽蓝鸟智能停车科技产业化有限公司 | Parking space position dynamic right confirming method and system based on block chain |
CN115641747A (en) * | 2022-12-23 | 2023-01-24 | 中电信数字城市科技有限公司 | Parking navigation method and system and electronic equipment |
Also Published As
Publication number | Publication date |
---|---|
CN108389421B (en) | 2020-12-22 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN108389421A (en) | The accurate inducible system in parking lot and method identified again based on image | |
CN106023605B (en) | A kind of method for controlling traffic signal lights based on depth convolutional neural networks | |
CN106652483B (en) | The method for laying traffic information test point in regional highway network using detection device | |
CN103258427B (en) | Urban expressway traffic real-time monitoring system and method based on information physical network | |
CN109544932A (en) | A kind of city road network flow estimation method based on GPS data from taxi Yu bayonet data fusion | |
CN103235933B (en) | A kind of vehicle abnormality behavioral value method based on HMM | |
CN103985250B (en) | The holographic road traffic state vision inspection apparatus of lightweight | |
CN104778834B (en) | Urban road traffic jam judging method based on vehicle GPS data | |
CN108492555A (en) | A kind of city road net traffic state evaluation method and device | |
CN111325978B (en) | Whole-process monitoring and warning system and method for abnormal behaviors of vehicles on expressway | |
CN111079602A (en) | Vehicle fine granularity identification method and device based on multi-scale regional feature constraint | |
CN101188064B (en) | 3D integrated freeway traffic event automatic detection method | |
CN106097755A (en) | For identifying the method parked place and/or vacate place | |
CN107730993A (en) | The parking lot intelligent vehicle-tracing system and method identified again based on image | |
CN101882374A (en) | System and method for analysis of comprehensive information on vehicle trip distribution | |
CN109190444A (en) | A kind of implementation method of the lane in which the drivers should pay fees vehicle feature recognition system based on video | |
CN114023062B (en) | Traffic flow information monitoring method based on deep learning and edge calculation | |
CN107437339A (en) | Variable information advices plate control method for coordinating and system under a kind of information guidance | |
CN106327880B (en) | A kind of speed recognition methods and its system based on monitor video | |
CN109935080A (en) | The monitoring system and method that a kind of vehicle flowrate on traffic route calculates in real time | |
CN110083099A (en) | One kind meeting automobile function safety standard automatic Pilot architecture system and working method | |
CN107516423A (en) | A kind of vehicle heading detection method based on video | |
CN110119790A (en) | The method of shared bicycle quantity statistics and density estimation based on computer vision | |
CN106384507A (en) | Travel time real-time estimation method based on sparse detector | |
CN110021161A (en) | A kind of prediction technique and system of traffic direction |
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 | ||
EE01 | Entry into force of recordation of patent licensing contract |
Application publication date: 20180810 Assignee: DALIAN HAIDA YINGHAI TECHNOLOGY Co.,Ltd. Assignor: Dalian Maritime University Contract record no.: X2023980053274 Denomination of invention: A parking lot precise guidance system and method based on image re recognition Granted publication date: 20201222 License type: Exclusive License Record date: 20231220 |
|
EE01 | Entry into force of recordation of patent licensing contract |