CN110307809A - A kind of model recognizing method and device - Google Patents
A kind of model recognizing method and device Download PDFInfo
- Publication number
- CN110307809A CN110307809A CN201810228691.9A CN201810228691A CN110307809A CN 110307809 A CN110307809 A CN 110307809A CN 201810228691 A CN201810228691 A CN 201810228691A CN 110307809 A CN110307809 A CN 110307809A
- Authority
- CN
- China
- Prior art keywords
- vehicle
- coordinate
- template
- point set
- measured
- 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
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01B—MEASURING LENGTH, THICKNESS OR SIMILAR LINEAR DIMENSIONS; MEASURING ANGLES; MEASURING AREAS; MEASURING IRREGULARITIES OF SURFACES OR CONTOURS
- G01B17/00—Measuring arrangements characterised by the use of infrasonic, sonic or ultrasonic vibrations
- G01B17/02—Measuring arrangements characterised by the use of infrasonic, sonic or ultrasonic vibrations for measuring thickness
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/22—Matching criteria, e.g. proximity measures
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Theoretical Computer Science (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Artificial Intelligence (AREA)
- Bioinformatics & Computational Biology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- General Engineering & Computer Science (AREA)
- Life Sciences & Earth Sciences (AREA)
- Image Analysis (AREA)
- Traffic Control Systems (AREA)
Abstract
The embodiment of the present invention provides a kind of model recognizing method and device, to solve the technical problem that recognition accuracy existing for existing model recognizing method is low, enforcement difficulty is big.Wherein, which comprises determine the first coordinate point set for characterizing vehicle roof contour line to be measured under side view visual angle, each coordinate points which concentrates represent a coordinate position；Based on preset similarity distance algorithm, determine similarity distance respectively between the first coordinate point set and the multiple template coordinate point set for belonging to various, wherein, every kind of vehicle includes at least one template coordinate point set in the various, each template coordinate point set characterizes vehicle vehicle roof outline line under a kind of side view visual angle, and similarity distance is the distance that indicates the similarity between point set；According to the similarity distance respectively, the vehicle of the vehicle to be measured is determined.
Description
Technical field
The present invention relates to field field of intelligent transportation technology more particularly to a kind of model recognizing methods and device.
Background technique
Vehicle cab recognition is the important component in intelligent transportation system.For example, can be realized based on vehicle cab recognition to road
The vehicle flowrate of road different automobile types, the vehicle flowrate data for counting the different automobile types of acquisition are to carry out traffic guidance, highway dimension
The key basic data of shield.
Currently, the method for vehicle cab recognition mainly include Pressure identification method and image-recognizing method both.Wherein, it presses
Power recognition methods is the vehicle according to identification vehicle with the weight of vehicle by the weight of detection vehicle；Image-recognizing method is logical
It crosses shooting vehicle image, and the processing such as background elimination is carried out to vehicle image, by treated vehicle image and the vehicle that prestores
Image is compared, to identify the vehicle of vehicle.
However, when carrying out vehicle cab recognition based on Pressure identification method, since the weight of each vehicle is not strictly to distinguish,
Thus it can only roughly identify vehicle, the accuracy of recognition result is lower.
When carrying out vehicle cab recognition based on image-recognizing method, information storage higher to the resolution requirement of vehicle image
Greatly, the angle to vehicle in vehicle image, size require also higher, need to carry out angle calibration system and ruler to vehicle image
Very little adjusting, also, the calculation amount of vehicle image comparison procedure is larger.
As it can be seen that the existing model recognizing method technical problem that there are recognition accuracies is low, enforcement difficulty is big.
Summary of the invention
The embodiment of the present invention provides a kind of model recognizing method and device, exists to solve existing model recognizing method
Recognition accuracy is low, technical problem that enforcement difficulty is big.
In a first aspect, a kind of model recognizing method is provided, this method comprises: determining for characterizing under side view visual angle to measuring car
First coordinate point set of vehicle roof outline line, each coordinate points which concentrates represent a coordinate position；It is based on
Preset similarity distance algorithm determines and divides between the first coordinate point set and the multiple template coordinate point set for belonging to various
Other similarity distance, wherein every kind of vehicle includes at least one template coordinate point set, each template coordinate in the various
Point set characterizes vehicle vehicle roof outline line under a kind of side view visual angle, and similarity distance is the distance that indicates the similarity between point set；
According to the similarity distance respectively, the vehicle of the vehicle to be measured is determined.
The model recognizing method provided according to embodiments of the present invention, due to vehicle of the vehicle under side view visual angle of different automobile types
Top contour line differs greatly, thus the comparison between the coordinate point set based on vehicle vehicle roof outline line under characterization side view visual angle,
Vehicle can be accurately identified, also, the calculation amount of the similarity distance between coordinate point set is smaller, so that the vehicle cab recognition
The enforcement difficulty of method is low, and can be improved the efficiency of vehicle cab recognition, furthermore, memory space needed for storing coordinate point set is small,
It further reduced the enforcement difficulty of the model recognizing method.
In one possible implementation, first for characterizing vehicle roof contour line to be measured under side view visual angle is determined
Coordinate point set, comprising: multiple height detection is carried out to by the vehicle to be measured of height detecting device with prefixed time interval, with
Obtain multiple bodywork height data of the vehicle to be measured from headstock portion to tailstock portion；And it detects the vehicle to be measured and passes through the height
Spend the speed during detection device；According to multiple bodywork height data, determine that first coordinate points concentrate each coordinate points
Vertical coordinate parameter；According to the speed and the prefixed time interval, the horizontal coordinate parameter of each coordinate points is determined.
The model recognizing method provided according to embodiments of the present invention, in the process that vehicle to be measured passes through height detecting device
In, Bus- Speed Monitoring and multiple height detection are carried out to it, multiple vehicle bodies with determination vehicle to be measured from headstock portion to tailstock portion are high
Degree evidence, and treat the speed that measuring car carries out vehicle to be measured during height detection.Due to height detection and Bus- Speed Monitoring
It is all easily achieved, it is thus possible to reduce the enforcement difficulty of the model recognizing method；Also, during detection, it is not necessarily to
Vehicle to be measured is set to remain static, thus when method to be applied to the vehicle cab recognition of road vehicle, it can be avoided to road
Smooth influence.
In one possible implementation, according to multiple bodywork height data, it is each to determine that first coordinate points are concentrated
The vertical coordinate parameter of coordinate points, comprising: determine that multiple bodywork height data are that first coordinate points concentrate each coordinate points
Vertical coordinate parameter；Or, determining that the product of multiple bodywork height data and the weight set is concentrated respectively as first coordinate points
The vertical coordinate parameter of coordinate points, wherein for the weight of the bodywork height data setting in characterization headstock portion, be higher than for the characterization tailstock
The weight of the bodywork height data setting in portion.
The model recognizing method provided according to embodiments of the present invention determines that each coordinate points are hung down according to multiple bodywork height data
Straight coordinate parameters include at least two schemes, wherein use and determine multiple bodywork height data for the vertical seat of each coordinate points
When marking this scheme of parameter, the calculation amount during vehicle cab recognition can be reduced, and uses and determines multiple bodywork height data
When with the product of the weight set as this scheme of the vertical coordinate parameter of each coordinate points, characteristic pair in vehicle body can be improved
The influence of recognition result reduces influence of the non-characteristic to recognition result in vehicle body, to improve the accurate of vehicle cab recognition result
Property.
In one possible implementation, it according to the similarity distance of the difference, determines the vehicle of the vehicle to be measured, wraps
It includes: determining similarity distance of the various respectively with the first coordinate point set, wherein any vehicle and first coordinate points
The similarity distance of collection, for the smallest phase of at least one template coordinate point set and the first coordinate point set that this kind of vehicle includes
Like degree distance；According to the various similarity distance with the first coordinate point set respectively, determine that the vehicle to be measured is respectively
The probability of this every kind vehicle, wherein similarity distance and probability negative correlation；Determine the vehicle to be measured for institute in multiple vehicle really
The vehicle for the maximum probability made.
The model recognizing method provided according to embodiments of the present invention, the similarity of any vehicle and the first coordinate point set away from
From, be each template coordinate point set and the first coordinate point set that this kind of vehicle includes similarity distance in minimum similarity degree away from
From.Similarity distance and the probability that vehicle to be measured is this kind of vehicle are negatively correlated, thus, it is sat respectively with first according to the various
The similarity distance of punctuate collection can accurately determine out the probability that vehicle to be measured is respectively every kind of vehicle.
In one possible implementation, according to the various and the first coordinate point set similarity respectively away from
From determining that the vehicle to be measured is respectively the probability of this every kind vehicle, comprising: pass through formula Pi=e-di/μDetermine the vehicle to be measured point
Not Wei this every kind vehicle relative probability, wherein PiIndicate that the vehicle to be measured is respectively the general of i-th kind of vehicle in the various
Rate, diIndicate the similarity distance of i-th kind of vehicle and the first coordinate point set in the various, μ indicate the various with
The mean value of the similarity distance of the first coordinate point set respectively；The relative probability is normalized, it is to be measured to obtain this
Vehicle is respectively the probability of this every kind vehicle.
The model recognizing method provided according to embodiments of the present invention, using formula Pi=e-di/μDetermining relative probability, more
Add the recognition result met in practical application.Also, after relative probability is normalized, it can obtain and compare relative probability
More intuitive probability data.
In one possible implementation, any template coordinate point set that multiple template coordinate points are concentrated passes through as follows
Mode obtains: multiple height detection is carried out to by the template vehicle of height detecting device with prefixed time interval, to be somebody's turn to do
Multiple template height of vehicle body data of the template vehicle from headstock portion to tailstock portion；And the template vehicle is detected by being somebody's turn to do
Template vehicle speed during height detecting device；Wherein, which is the vehicle of any template coordinate point set characterization
Push up vehicle belonging to contour line；According to multiple template height of vehicle body data, it is each to determine that any template coordinate points are concentrated
The vertical coordinate parameter of coordinate points；According to the template vehicle speed and the prefixed time interval, any template coordinate points are determined
Concentrate the horizontal coordinate parameter of each coordinate points.
The model recognizing method provided according to embodiments of the present invention, can be using side identical with the first coordinate point set is determined
Method predefines the template coordinate point set of multiple template vehicle, in this way, during vehicle cab recognition, it can be directly according to more
A template coordinate point set, quickly and accurately determines the vehicle of vehicle to be measured.
Second aspect provides a kind of vehicle type recognition device, which includes:
Coordinate point set determining module, for determining first sitting for characterizing vehicle roof contour line to be measured under side view visual angle
Punctuate collection, each coordinate points which concentrates represent a coordinate position；
Similarity is apart from determining module, for being based on preset similarity distance algorithm, determine the first coordinate point set with
Belong to similarity distance respectively between the multiple template coordinate point set of various, wherein every kind of vehicle in the various
Including at least one template coordinate point set, each template coordinate point set characterizes vehicle vehicle roof outline line under a kind of side view visual angle, phase
Distance like degree distance to indicate the similarity between point set；
Vehicle determining module determines the vehicle of the vehicle to be measured for the similarity distance according to the difference.
In one possible implementation, which is used for:
Multiple height detection is carried out to by the vehicle to be measured of height detecting device with prefixed time interval, to be somebody's turn to do
Multiple bodywork height data of the vehicle to be measured from headstock portion to tailstock portion；And it detects the vehicle to be measured and passes through the height detection
Speed during device；
According to multiple bodywork height data, determine that first coordinate points concentrate the vertical coordinate parameter of each coordinate points；
According to the speed and the prefixed time interval, the horizontal coordinate parameter of each coordinate points is determined.
In one possible implementation, which is used for:
Determine that multiple bodywork height data are the vertical coordinate parameter that first coordinate points concentrate each coordinate points；Or,
Determine that the product of multiple bodywork height data and the weight set concentrates each coordinate points as first coordinate points
Vertical coordinate parameter, wherein be the weight of the bodywork height data setting in characterization headstock portion, higher than the vehicle body to characterize tailstock portion
The weight of altitude information setting.
In one possible implementation, which is used for:
Determine similarity distance of the various respectively with the first coordinate point set, wherein any vehicle and this first
The similarity distance of coordinate point set, at least one the template coordinate point set and the first coordinate point set for including for this kind of vehicle are most
Small similarity distance；
According to the various similarity distance with the first coordinate point set respectively, determining that the vehicle to be measured is respectively should
The probability of every kind of vehicle, wherein similarity distance and probability negative correlation；
Determine the vehicle to be measured by the vehicle for the maximum probability determined in multiple vehicle.
In one possible implementation, which is used for:
Pass through formula Pi=e-di/μDetermine that the vehicle to be measured is respectively the relative probability of this every kind vehicle, wherein PiIt indicates
The vehicle to be measured is respectively the probability of i-th kind of vehicle in the various, diIndicate in the various i-th kind of vehicle and this
The similarity distance of one coordinate point set, μ indicate the mean value of the similarity distance of the various and the first coordinate point set respectively；
The relative probability is normalized, to obtain the probability that the vehicle to be measured is respectively this every kind vehicle.
In one possible implementation, which further includes template coordinate point set determining module, for by as follows
Mode obtains any template coordinate point set that multiple template coordinate points are concentrated:
Multiple height detection is carried out to by the template vehicle of height detecting device with prefixed time interval, to obtain the mould
Multiple template height of vehicle body data of the wooden handcart from headstock portion to tailstock portion；And it detects the template vehicle and passes through the height
Spend the template vehicle speed during detection device；Wherein, which is the roof of any template coordinate point set characterization
Vehicle belonging to contour line；
According to multiple template height of vehicle body data, determine that any template coordinate points concentrate the vertical of each coordinate points
Coordinate parameters；
According to the template vehicle speed and the prefixed time interval, determine that any template coordinate points concentrate each coordinate points
Horizontal coordinate parameter.
The technical effect of vehicle type recognition device provided in an embodiment of the present invention may refer to above-mentioned first aspect or first party
The technical effect of the various implementations in face, details are not described herein again.
The third aspect provides a kind of computer equipment, which includes:
At least one processor, and
The memory being connect at least one processor；
Wherein, which is stored with the instruction that can be executed by least one processor, at least one processor is logical
The instruction for executing memory storage is crossed, is known with executing the vehicle as described in the various implementations of first aspect or first aspect
Other method.
The technical effect of computer equipment provided in an embodiment of the present invention may refer to above-mentioned first aspect or first aspect
The technical effect of various implementations, details are not described herein again.
Fourth aspect provides a kind of computer readable storage medium, in which:
The computer-readable recording medium storage has computer instruction, when the computer instruction is run on computers,
So that computer executes method described in the various implementations of first aspect or first aspect.
The technical effect of computer equipment provided in an embodiment of the present invention may refer to above-mentioned first aspect or first aspect
The technical effect of various implementations, details are not described herein again.
Detailed description of the invention
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 technical description to be briefly described, it should be apparent that, the accompanying drawings in the following description is only this
Inventive embodiments for those of ordinary skill in the art without creative efforts, can also be according to mentioning
The attached drawing of confession obtains other attached drawings.
Fig. 1 is a kind of schematic diagram of vehicle vehicle roof outline line under side view visual angle in the embodiment of the present invention；
Fig. 2 is a kind of schematic diagram of template library Establishing process in the embodiment of the present invention；
Fig. 3 is the schematic diagram of a scenario that a kind of pair of vehicle is detected in the embodiment of the present invention；
Fig. 4 is the schematic diagram of vehicle vehicle roof outline line under another side view visual angle in the embodiment of the present invention；
Fig. 5 is a kind of schematic diagram of vehicle cab recognition process in the embodiment of the present invention；
Fig. 6 is a kind of schematic diagram of the template coordinate point set of not weighted processing in the embodiment of the present invention；
Fig. 7 is a kind of schematic diagram of weighting treated template coordinate point set in the embodiment of the present invention；
Fig. 8 is the schematic diagram of the template coordinate point set of another not weighted processing in the embodiment of the present invention；
Fig. 9 is the schematic diagram of another weighting treated template coordinate point set in the embodiment of the present invention；
Figure 10 is a kind of schematic diagram of vehicle type recognition device in the embodiment of the present invention.
Specific embodiment
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
Every other embodiment obtained without creative efforts, shall fall within the protection scope of the present invention.
In addition, the terms "and/or", only a kind of incidence relation for describing affiliated partner, indicates may exist
Three kinds of relationships, for example, A and/or B, can indicate: individualism A exists simultaneously A and B, these three situations of individualism B.Separately
Outside, character "/" herein typicallys represent the relationship that forward-backward correlation object is a kind of "or" in the case where not illustrating.
In addition, it is necessary to understand, in the description of the embodiment of the present invention, the vocabulary such as " first ", " second " are only used for distinguishing description
Purpose is not understood to indicate or imply relative importance, can not be interpreted as indication or suggestion sequence.
Model recognizing method and device provided in an embodiment of the present invention can be used for identifying vehicle.Wherein, vehicle refers to vehicle
Type.For example, can be according to motor vehicle parameters such as vehicle commander, overall height, the load-bearing number of axle, specified seating capacity, loading capacity, by vehicle
It is divided into following 6 kinds of vehicles: middle minibus, motor bus, jubilee wagen, medium truck, large-sized truck, super-huge lorry.Certainly,
It in the specific implementation process, can also be according to actual needs using other vehicle classification modes.
Embodiment one
The embodiment of the present invention provides a kind of model recognizing method, which can be divided into template library foundation and vehicle
Type identifies the two parts.Wherein, the template library established based on template established part of the library can be used as the identification of vehicle cab recognition part
The basis of vehicle, and identify the process of vehicle, result and can be used for updating, improve template library.The two parts are carried out below
It introduces:
First part: template library is established
In the embodiment of the present invention, template library includes the multiple template coordinate point set for belonging to various, wherein a variety of vehicles
Every kind of vehicle in type includes at least one template coordinate point set, and each template coordinate point set characterizes vehicle under a kind of side view visual angle
Vehicle roof outline line.Fig. 1 show a kind of schematic diagram of vehicle vehicle roof outline line under side view visual angle, wherein x-axis corresponds to vehicle commander, y-axis
Overall height is corresponded to, each coordinate points marked on the vehicle roof outline line 101 in Fig. 1 form a coordinate point set.
Fig. 2 is referred to, it, can be with for any template vehicle for establishing template library during establishing template library
By following step 201- step 203, to determine the template for characterizing the vehicle module vehicle roof outline line under side view visual angle
Coordinate point set:
Step 201: multiple height detection is carried out to by the template vehicle of height detecting device with prefixed time interval, with
Obtain multiple template height of vehicle body data of the template vehicle from headstock portion to tailstock portion；And detect the template vehicle
Pass through the template vehicle speed during height detecting device.
For example, as shown in figure 3, height detecting device may include bracket 301 and rangefinder 302.With rangefinder 302
For ultrasonic range finder, rangefinder 302 can emit ultrasonic wave vertically downward, to detect rangefinder to reflecting ultrasonic wave
Reflect the distance of object.During vehicle passes through height detecting device, rangefinder 302 is with prefixed time interval (such as 40ms)
Transmitting excusing from death wave, can obtain each section roof of the vehicle from headstock portion to tailstock portion respectively at a distance from rangefinder 302, into
And each section roof is subtracted at a distance from rangefinder 302 away from the distance on ground with rangefinder 302 respectively, so that it may obtain multiple
Bodywork height data.In the specific implementation process, bracket 301 could alternatively be other objects with certain altitude, such as
Overline bridge, portal frame, tunnel etc..
In Fig. 3, tachymeter 303 can detecte vehicle and pass through the vehicle speed during height detecting device.Wherein, it surveys
Fast instrument 303 can be Microwave Velocity instrument, radar meter, laser velocimeter etc..Also, tachymeter 303 can pacify as shown in Figure 3
It sets in the road roadside that vehicle passes through, tachymeter 303 can also be placed on bracket 301, etc..
Step 202: according to multiple template height of vehicle body data, determining the corresponding template coordinate points of the template vehicle
Concentrate the vertical coordinate parameter of each coordinate points.
In the embodiment of the present invention, according to multiple template height of vehicle body data, the corresponding mould of template vehicle is determined
Plate coordinate points concentrate the specific embodiment of the vertical coordinate parameter of each coordinate points, at least may include two kinds:
The first determines the mode of vertical coordinate parameter: determining that multiple template height of vehicle body data are the template vehicle
Corresponding template coordinate points concentrate the vertical coordinate parameter of each coordinate points.
The mode of second of determining vertical coordinate parameter: the power of multiple template height of vehicle body data and setting is determined
The product of weight is the vertical coordinate parameter that the corresponding template coordinate points of the template vehicle concentrate each coordinate points, wherein for characterization vehicle
The weight of the bodywork height data setting on head, higher than the weight to characterize the bodywork height data setting in tailstock portion.
In the mode of second of determining vertical coordinate parameter, by being set higher than the weight in tailstock portion for headstock portion, make
The difference in height for obtaining headstock portion is reinforced, and the difference in height in tailstock portion is then weakened.By taking lorry as an example, when lorry loading and not
The difference in height of wagon box is very big when loading, thus strengthens the difference in height in headstock portion and weaken the difference in height in tailstock portion
It is different, the accuracy of vehicle cab recognition result will be promoted.
In the specific implementation process, headstock portion and tailstock portion can be set and respectively accounts for the percentage of vehicle body, for example, can set
Determine preceding 30% part that headstock portion is vehicle body, tailstock portion is rear 70% part of vehicle body.It also, is characterization headstock portion and tailstock portion
The weight that is set separately of bodywork height data, can be the data obtained by multiple test optimization, for example, can be characterization
The weight of the bodywork height data setting 140% in headstock portion, for the weight of the bodywork height data setting 80% in characterization tailstock portion.
Step 203: according to the template vehicle speed and the prefixed time interval, determining the corresponding template of template vehicle
Coordinate points concentrate the horizontal coordinate parameter of each coordinate points.
For example, as shown in figure 4, coordinate points 41 are first coordinate for characterizing the coordinate points of vehicle roof outline line and concentrating
Point, in the specific implementation process, the horizontal coordinate parameter of first coordinate points can be arbitrary preset value, such as can be pre-
It is set as 0,10 etc..That is, in Fig. 4, X41It can be preset value.And in Fig. 4, X42For (X41+ c1), X43For (X42+ c2),
Wherein, c1 indicates to carry out the distance that vehicle driving is crossed between first time height detection and second of height detection, c2 table to vehicle
Show to the distance that vehicle driving is crossed between second of height detection of vehicle and third time height detection.
There is vehicle by the speed during height detecting device due to measuring, and during the multiple height detection of progress
Prefixed time interval is used, therefore, it is determined the distance of vehicle driving is prefixed time interval between adjacent height detection twice
Product between speed, that is, c1 and c2 can be determined according to speed and prefixed time interval, in turn, be based on (X41+ c1) it can
To determine X42, it is based on (X42+ c2) it can determine X43, and the horizontal coordinate parameter for other coordinate points that coordinate points are concentrated can also be with
It is successively determined by the way of similar.
Pass through the flat of height detecting device when vehicle at the uniform velocity passes through height detecting device or detected speed for vehicle
When equal speed, vehicle driving is equidistant between adjacent height detection twice, that is, in that case, c1=shown in Fig. 4
c2.Then, X42It can be expressed as (X41+ 1*c), X43For (X41+ 2*c), wherein c is the product of prefixed time interval and speed, and
The horizontal coordinate parameter for other coordinate points that coordinate points are concentrated can also successively be determined by the way of similar.
In the embodiment of the present invention, when being determined by the way of vertical coordinate parameter using the first above-mentioned, the mould of acquisition
Plate coordinate point set can be indicated with set H:
H={ (xk,yk)|k∈[1,F])}
Wherein, F indicates the coordinate points quantity that the template coordinate points are concentrated, xkIndicate that the template coordinate points concentrate k-th of seat
The horizontal coordinate parameter of punctuate, ykIndicate that the template coordinate points concentrate the vertical coordinate parameter of k-th of coordinate points.
It is each in set H when by the way of above-mentioned second determining vertical coordinate parameter in the embodiment of the present invention
The weight of coordinate points setting can be indicated with set W:
W={ wk|k∈[1,F]}
Wherein, F indicates the coordinate points quantity that the template coordinate points are concentrated, wkThe template coordinate points are expressed as to concentrate k-th
The weight of the bodywork height data setting of coordinate points can be determined using above-mentioned second really then according to set H and set W
When determining the mode of vertical coordinate parameter, the template coordinate point set of acquisition can be indicated with set H:
H '={ (xk,wkyk)|k∈[1,F])}
Wherein, F indicates the coordinate points quantity that the template coordinate points are concentrated, xkIndicate that the template coordinate points concentrate k-th of seat
The horizontal coordinate parameter of punctuate, wkykIndicate that the template coordinate points concentrate the vertical coordinate parameter of k-th of coordinate points.
In the embodiment of the present invention, according to the difference of the vehicle roof outline line under vehicle side view visual angle each in any vehicle, by this
Kind vehicle is divided into multiple subclasses, and each subclass includes at least one template coordinate point set, the template coordinate point set group of each subclass
At the template set of this kind of vehicle.
Wherein, a kind of template set of vehicle can be indicated with set S:
S=V (n) | n ∈ [1, N] }
Wherein, N indicates the quantity for the subclass that this kind of vehicle includes, and V (n) indicates n-th of subclass.
For example, if with set S1Indicate the template set of middle-size and small-size this vehicle of car, and middle-size and small-size car includes small
Car V1(1), minibus V1(2), jeep V1(3), station wagon V1(4), middle bus V1(5) this five subclasses, then it is medium and small
The template set S of type car1={ V1(1),V1(2),V1(3),V1(4),V1(5)}。
Further, if by vehicle be divided into middle-size and small-size car, jubilee wagen, medium truck, motorbus, large-sized truck and
This six kinds of vehicles of super-huge lorry, and the template set of this six kinds of vehicles uses S respectively1、S2、S3、S4、S5And S6It indicates, then summarizes
Template library can be obtained in the template set of this six kinds of vehicles, and template library can be indicated with set L:
L={ S1,S2,S3,S4,S5,S6}
Second part: vehicle cab recognition
Fig. 5 is referred to, the process of vehicle cab recognition part is described as follows:
Step 501: determining the first coordinate point set for characterizing vehicle roof contour line to be measured under side view visual angle, first sits
Each coordinate points that punctuate is concentrated represent a coordinate position.
Wherein it is determined that the concrete mode of the first coordinate point set can be and treat measuring car and measure calculating, it is also possible to
The image for shooting vehicle to be measured carries out image analysis, etc..
In a kind of possible embodiment, first for characterizing vehicle roof contour line to be measured under side view visual angle is determined
Coordinate point set may include following step 5011- step 5013:
Step 5011: multiple height detection is carried out to by the vehicle to be measured of height detecting device with prefixed time interval,
To obtain multiple bodywork height data of the vehicle to be measured from headstock portion to tailstock portion；And it detects vehicle to be measured and is examined by height
Survey the speed during device.
Step 5012: according to multiple bodywork height data, determining that the first coordinate points concentrate the vertical coordinate of each coordinate points to join
Number.
Step 5013: according to speed and prefixed time interval, determining the horizontal coordinate parameter of each coordinate points.
In a kind of possible embodiment, according to multiple bodywork height data, determine that the first coordinate points concentrate each coordinate
The concrete mode of the vertical coordinate parameter of point, at least may include two kinds:
The first determines the mode of vertical coordinate parameter: determining that multiple bodywork height data are that the first coordinate points are concentrated respectively
The vertical coordinate parameter of coordinate points.
The mode of second of determining vertical coordinate parameter: the product of the weight of multiple bodywork height data and setting is determined
The vertical coordinate parameter of each coordinate points is concentrated for the first coordinate points, wherein be the bodywork height data setting in characterization headstock portion
Weight, higher than the weight to characterize the bodywork height data setting in tailstock portion.
In the specific implementation process, the first coordinate for characterizing vehicle roof contour line to be measured under side view visual angle is determined
The mode of point set, can with determine the mode for characterizing the template coordinate point set of side view visual angle lower template vehicle roof contour line
It is identical, determine that the specific embodiment of the first coordinate point set may refer to sit in above-mentioned template established part of the library about determining template
The description of punctuate collection, details are not described herein.
Step 502: being based on preset similarity distance algorithm, determine the first coordinate point set and belong to the multiple of various
Similarity distance between template coordinate point set respectively, wherein every kind of vehicle includes at least one template coordinate in various
Point set, each template coordinate point set characterize vehicle vehicle roof outline line under a kind of side view visual angle, similarity distance for indicate point set it
Between similarity distance.
Wherein, preset similarity distance algorithm for example can be for calculating amendment Hausdorff distance (Modified
Hausdorff Distance, MHD) algorithm, MHD can be used for measuring the similarity of two vehicle roof outline lines.Compared to person of outstanding talent
Si Duofu uses the average value of distance as measurement standard in (Hausdorff Distance, HD), MHD, by average
Algorithm is shared, and MHD reduces the susceptibility for calculating noise in data.
Step 503: according to similarity distance respectively, determining the vehicle of vehicle to be measured.
In the embodiment of the present invention, similarity distance is for indicating the similarity between point set, wherein similarity is apart from smaller
Similarity between indicates coordinate point set is smaller, i.e. the vehicle roof outline line of two coordinate point sets characterization is closer.Thus, specific
Implementation process in, can determine that the vehicle of vehicle to be measured is similarity template vehicle belonging to the smallest template coordinate point set
Vehicle, it is of course also possible to according between the first coordinate point set and multiple template coordinate point set respectively similarity distance, first
It determines that vehicle to be measured is respectively the probability of each vehicle, the vehicle, etc. of vehicle to be measured is determined further according to the probability determined.
In a kind of possible embodiment, according to the similarity distance respectively, the vehicle of vehicle to be measured is determined, it can be with
Include the following steps, namely 5031- step 5033:
Step 5031: determining similarity distance of the various respectively with the first coordinate point set, wherein any vehicle and the
The similarity distance of one coordinate point set, at least one the template coordinate point set and the first coordinate point set for including for this kind of vehicle are most
Small similarity distance.
As an example it is assumed that this vehicle of middle-size and small-size car includes 6 template coordinate point sets, wherein each template is sat
Punctuate collection and the first coordinate point set have a similarity distance, totally 6 similarity distances, then this vehicle of middle-size and small-size car and first
The similarity distance of coordinate point set is the smallest similarity distance in this 6 similarity distances.
Step 5032: according to the various similarity distance with the first coordinate point set respectively, determining vehicle difference to be measured
For the probability of every kind of vehicle, wherein similarity distance and probability negative correlation.
In the embodiment of the present invention, similarity distance indicates that the similarity between point set, the more big then template of similarity distance are sat
The vehicle roof outline line and the difference of the vehicle roof outline line of vehicle to be measured of punctuate collection characterization are bigger, and similarity is sat apart from smaller then template
The vehicle roof outline line and the difference of the vehicle roof outline line of vehicle to be measured of punctuate collection characterization are bigger.
In a kind of possible embodiment, according to the similarity distance of various and the first coordinate point set respectively, really
Fixed vehicle to be measured is respectively the detailed process of the probability of every kind of vehicle, be can be through formula Pi=e-di/μDetermine vehicle to be measured point
Not Wei every kind of vehicle relative probability, wherein PiIndicate that vehicle to be measured is respectively the probability of i-th kind of vehicle in various, diTable
Show the similarity distance of i-th kind of vehicle and the first coordinate point set in various, μ indicates various and the first coordinate point set point
The mean value of other similarity distance；In turn, relative probability is normalized, is respectively every kind of vehicle to obtain vehicle to be measured
The probability of type.
Wherein, it is which kind of vehicle, the phase of each vehicle that relative probability, which can be used between each vehicle being compared to each other vehicle to be measured,
To the sum of probability not necessarily 1；The probability obtained after normalized is that vehicle to be measured is respectively the naturally general of every kind of vehicle
Rate, the sum of probability of each vehicle obtained after processing are 1.
Step 5033: determining vehicle to be measured by the vehicle for the maximum probability determined in multiple vehicles.
In the specific implementation process, the output result after vehicle cab recognition can be the vehicle for the vehicle to be measured determined,
It is also possible to the probability that vehicle to be measured is respectively every kind of vehicle.
For ease of understanding, the model recognizing method in the embodiment of the present invention is lifted with a specific implementation process below
Example explanation:
Assuming that in the L of module library by vehicle be divided into middle-size and small-size car and jubilee wagen both.
During establishing template library L, detection sampling is carried out to 4 middle-size and small-size car template vehicle, is obtained without adding
Weigh the middle-size and small-size car template set H of processing1={ H1(1),H1(2),H1(3),H1(4) }, including each not weighted processing
Template coordinate point set it is as shown in Figure 6.
Assuming that the headstock portion of 30% part is characterized region before vehicle body, the tailstock portion of 70% part is non-characteristic area after vehicle body
Domain, for characterize headstock portion bodywork height data setting 140% weight, for characterize tailstock portion bodywork height data setting
80% weight.Then centering station wagon template set H1After being weighted processing, template set S is obtained1={ V1(1),V1(2),V1
(3),V1(4) }, as shown in Figure 7.
Detection sampling is carried out to 5 jubilee wagen template vehicles, obtains the template set H of not weighted processing2={ H2(1),
H2(2),H2(3),H2(4),H2(5) }, jubilee wagen template set H2Including each not weighted processing template coordinate point set as scheme
Shown in 8.
It is similarly assumed that the headstock portion of 30% part is characterized region before vehicle body, the tailstock portion of 70% part is non-after vehicle body
Characteristic area, for characterize headstock portion bodywork height data setting 140% weight, for characterize tailstock portion bodywork height data
The weight of setting 80%.To jubilee wagen template set H2After being weighted processing, template set S is obtained2={ V2(1),V2(2),V2
(3),V2(4),V2(5) }, as shown in Figure 9.
During vehicle cab recognition, treats measuring car and carry out detection sampling, obtain the weighted processing of vehicle not to be measured
Original coordinates point set H, original coordinates point set H is as shown in table 1.
Table 1
| Coordinate point set | Point 1 | Point 2 | Point 3 | Point 4 | Point 5 | Point 6 | Point 7 | Point 8 | Point 9 |
| X | 38 | 114 | 190 | 266 | 342 | 418 | 494 | 570 | 646 |
| Y | 102 | 110 | 149 | 151 | 143 | 130 | 100 | 86 | 80 |
It is similarly assumed that the headstock portion of 30% part is characterized region before vehicle body, the tailstock portion of 70% part is non-after vehicle body
Characteristic area, for characterize headstock portion bodywork height data setting 140% weight, for characterize tailstock portion bodywork height data
The weight of setting 80%.After being weighted processing to original coordinates point set H, coordinate point set H ', coordinate point set H ' such as 2 institute of table are obtained
Show.
Table 2
| Coordinate point set | Point 1 | Point 2 | Point 3 | Point 4 | Point 5 | Point 6 | Point 7 | Point 8 | Point 9 |
| X | 38 | 114 | 190 | 266 | 342 | 418 | 494 | 570 | 646 |
| Y’ | 143 | 154 | 209 | 121 | 114 | 104 | 80 | 69 | 64 |
Then, according to preset MHD algorithm, coordinates computed point set H ' and template set S1With template set S2It is sat including each template
The similarity distance of punctuate collection respectively, calculated result are as shown in table 3.
Table 3
Wherein, the template set S of middle-size and small-size car1Including the minimum of 4 template coordinate point sets and coordinate point set H ' it is similar
Spend distance d1It is 30.9767.The template set S of jubilee wagen2Including 5 template coordinate point sets and coordinate point set H ' minimum phase
Like degree distance d2It is 9.7897.
In turn, in conjunction with formula Pi=e-di/μDetermine that vehicle to be measured is the relative probability P (d of middle-size and small-size car1)=
0.219, vehicle to be measured is the relative probability P (d of jubilee wagen2)=0.619.To P (d1) and P (d2) be normalized after,
Obtaining the probability that vehicle to be measured is middle-size and small-size car is 26.1%, and vehicle to be measured is that the probability of jubilee wagen is 73.9%.Then may be used
Using determination vehicle to be measured as jubilee wagen.
Embodiment two
Based on the same inventive concept, the embodiment of the present invention provides a kind of vehicle type recognition device.Referring to Figure 10, which knows
Other device includes at least coordinate point set determining module 1001, similarity apart from determining module 1002 and vehicle determining module 1003,
Wherein:
Coordinate point set determining module 1001, for determining for characterizing vehicle roof contour line to be measured under side view visual angle
One coordinate point set, each coordinate points which concentrates represent a coordinate position；
Similarity determines first coordinate points for being based on preset similarity distance algorithm apart from determining module 1002
Collect and belong to similarity distance respectively between the multiple template coordinate point set of various, wherein every kind in the various
Vehicle includes at least one template coordinate point set, and each template coordinate point set characterizes vehicle vehicle roof outline under a kind of side view visual angle
Line, the distance of similarity of the similarity distance between expression point set；
Vehicle determining module 1003 determines the vehicle of the vehicle to be measured for the similarity distance according to the difference.
In a kind of possible embodiment, which is used for:
Multiple height detection is carried out to by the vehicle to be measured of height detecting device with prefixed time interval, to be somebody's turn to do
Multiple bodywork height data of the vehicle to be measured from headstock portion to tailstock portion；And it detects the vehicle to be measured and passes through the height detection
Speed during device；
According to multiple bodywork height data, determine that first coordinate points concentrate the vertical coordinate parameter of each coordinate points；
According to the speed and the prefixed time interval, the horizontal coordinate parameter of each coordinate points is determined.
In a kind of possible embodiment, which is used for:
Determine that multiple bodywork height data are the vertical coordinate parameter that first coordinate points concentrate each coordinate points；Or,
Determine that the product of multiple bodywork height data and the weight set concentrates each coordinate points as first coordinate points
Vertical coordinate parameter, wherein be the weight of the bodywork height data setting in characterization headstock portion, higher than the vehicle body to characterize tailstock portion
The weight of altitude information setting.
In a kind of possible embodiment, which is used for:
Determine similarity distance of the various respectively with the first coordinate point set, wherein any vehicle and this first
The similarity distance of coordinate point set, at least one the template coordinate point set and the first coordinate point set for including for this kind of vehicle are most
Small similarity distance；
According to the various similarity distance with the first coordinate point set respectively, determining that the vehicle to be measured is respectively should
The probability of every kind of vehicle, wherein similarity distance and probability negative correlation；
Determine the vehicle to be measured by the vehicle for the maximum probability determined in multiple vehicle.
In a kind of possible embodiment, which is used for:
Pass through formula Pi=e-di/μDetermine that the vehicle to be measured is respectively the relative probability of this every kind vehicle, wherein PiIt indicates
The vehicle to be measured is respectively the probability of i-th kind of vehicle in the various, diIndicate in the various i-th kind of vehicle and this
The similarity distance of one coordinate point set, μ indicate the mean value of the similarity distance of the various and the first coordinate point set respectively；
The relative probability is normalized, to obtain the probability that the vehicle to be measured is respectively this every kind vehicle.
In a kind of possible embodiment, which further includes template coordinate point set determining module 1001,
Any template coordinate point set concentrated for obtaining multiple template coordinate points in the following way:
Multiple height detection is carried out to by the template vehicle of height detecting device with prefixed time interval, to obtain the mould
Multiple template height of vehicle body data of the wooden handcart from headstock portion to tailstock portion；And it detects the template vehicle and passes through the height
Spend the template vehicle speed during detection device；Wherein, which is the roof of any template coordinate point set characterization
Vehicle belonging to contour line；
According to multiple template height of vehicle body data, determine that any template coordinate points concentrate the vertical of each coordinate points
Coordinate parameters；
According to the template vehicle speed and the prefixed time interval, determine that any template coordinate points concentrate each coordinate points
Horizontal coordinate parameter.
Embodiment three
Based on the same inventive concept, the embodiment of the present invention provides a kind of computer equipment, comprising:
At least one processor, and
The memory being connect at least one processor；
Wherein, memory is stored with the instruction that can be executed by least one processor, at least one processor passes through execution
The instruction of memory storage, executes the method as described in embodiment one.
Example IV
Based on the same inventive concept, the embodiment of the present invention provides a kind of computer readable storage medium, this is computer-readable
Storage medium is stored with computer instruction, when computer instruction is run on computers, so that computer executes embodiment one
The method.
In the specific implementation process, computer readable storage medium includes: general serial bus USB
(Universal Serial Bus flash drive, USB), mobile hard disk, read-only memory (Read-Only Memory,
ROM), random access memory (Random Access Memory, RAM), magnetic or disk etc. are various can store program
The storage medium of code.
Above-mentioned one or more technical solutions, at least have the following beneficial effects:
In the embodiment of the present invention, the first coordinate points for characterizing vehicle roof contour line to be measured under side view visual angle are determined
Collection, each coordinate points which concentrates represent a coordinate position；Based on preset similarity distance algorithm, determine
Similarity distance between the first coordinate point set and the multiple template coordinate point set for belonging to various respectively, wherein this is more
Every kind of vehicle includes at least one template coordinate point set in kind of vehicle, and each template coordinate point set characterizes a kind of side view visual angle and gets off
Vehicle roof outline line；According to the similarity distance respectively, the vehicle of the vehicle to be measured is determined.
Since vehicle roof outline line of the vehicle under side view visual angle of different automobile types differs greatly, thus based on characterization side view
Comparison under visual angle between the coordinate point set of vehicle vehicle roof outline line, can accurately identify vehicle, also, coordinate point set it
Between similarity distance calculation amount it is smaller so that the enforcement difficulty of the model recognizing method is low, and can be improved vehicle cab recognition
Efficiency, furthermore, memory space needed for storing coordinate point set is small, and the implementation that further reduced the model recognizing method is difficult
Degree.
Further, in the embodiment of the present invention, during vehicle to be measured passes through height detecting device, vehicle is carried out to it
Speed detection and multiple height detection, multiple bodywork height data with determination vehicle to be measured from headstock portion to tailstock portion and right
Vehicle to be measured carries out the speed of vehicle to be measured during height detection.Since height detection and Bus- Speed Monitoring are all easily achieved
, it is thus possible to reduce the enforcement difficulty of the model recognizing method；Also, during detection, without making at vehicle to be measured
In stationary state, thus when method to be applied to the vehicle cab recognition of road vehicle, the influence to the coast is clear can be avoided.
Further, in the embodiment of the present invention, each coordinate points vertical coordinate parameter is determined according to multiple bodywork height data
Including at least two schemes, wherein use determine multiple bodywork height data for the vertical coordinate parameter of each coordinate points this
When scheme, the calculation amount during vehicle cab recognition can be reduced, and uses the power for determining multiple bodywork height data and setting
When the product of weight is this scheme of the vertical coordinate parameter of each coordinate points, characteristic can be improved in vehicle body to recognition result
It influences, influence of the non-characteristic to recognition result in vehicle body is reduced, to improve the accuracy of vehicle cab recognition result.
Further, in the embodiment of the present invention, the similarity distance of any vehicle and the first coordinate point set is this kind of vehicle
Minimum similarity degree distance in the similarity distance for each template coordinate point set and the first coordinate point set that type includes.Similarity distance
It is negatively correlated for the probability of this kind of vehicle with vehicle to be measured, thus, it is similar to the first coordinate point set respectively according to the various
Distance is spent, the probability that vehicle to be measured is respectively every kind of vehicle can be accurately determined out.
Further, in the embodiment of the present invention, using formula Pi=e-di/μDetermining relative probability, is more in line with and actually answers
Recognition result in.Also, after relative probability is normalized, it can obtain more more intuitive than relative probability general
Rate data.
Further, it in the embodiment of the present invention, can be predefined using method identical with the first coordinate point set is determined
The template coordinate point set of multiple template vehicle out, in this way, during vehicle cab recognition, it can be directly according to multiple template coordinate points
Collection, quickly and accurately determines the vehicle of vehicle to be measured.
The apparatus embodiments described above are merely exemplary, wherein units/modules as illustrated by the separation member
It may or may not be physically separated, the component shown as units/modules may or may not be
Physical unit/module, it can it is in one place, or may be distributed in multiple network unit/modules.It can basis
It is actual to need that some or all of the modules therein is selected to achieve the purpose of the solution of this embodiment.Ordinary skill people
Member is without paying creative labor, it can understands and implements.
Through the above description of the embodiments, those skilled in the art can be understood that each embodiment can
It realizes by means of software and necessary general hardware platform, naturally it is also possible to pass through hardware.Based on this understanding, on
Stating technical solution, substantially the part that contributes to existing technology can be embodied in the form of software products in other words, should
Computer software product may be stored in a computer readable storage medium, such as ROM/RAM, magnetic disk, CD, including several fingers
It enables and using so that a computer equipment (can be personal computer, server or the network equipment etc.) executes each implementation
Method described in certain parts of example or embodiment.
Finally, it should be noted that the above embodiments are merely illustrative of the technical solutions of the present invention, rather than its limitations；Although
Present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that: it still may be used
To modify the technical solutions described in the foregoing embodiments or equivalent replacement of some of the technical features；
And these are modified or replaceed, technical solution of various embodiments of the present invention that it does not separate the essence of the corresponding technical solution spirit and
Range.
Claims (14)
1. a kind of model recognizing method characterized by comprising
Determine the first coordinate point set for characterizing vehicle roof contour line to be measured under side view visual angle, first coordinate points are concentrated
Each coordinate points represent a coordinate position；
Based on preset similarity distance algorithm, determines the first coordinate point set and belong to the multiple template coordinate of various
Similarity distance between point set respectively, wherein every kind of vehicle includes at least one template coordinate point set in the various,
Each template coordinate point set characterizes vehicle vehicle roof outline line under a kind of side view visual angle, and similarity distance is indicates the phase between point set
Like the distance of degree；
According to the similarity distance of the difference, the vehicle of the vehicle to be measured is determined.
2. the method as described in claim 1, which is characterized in that determine for characterizing vehicle roof profile to be measured under side view visual angle
First coordinate point set of line, comprising:
Multiple height detection is carried out to by the vehicle to be measured of height detecting device with prefixed time interval, described in obtaining
Multiple bodywork height data of the vehicle to be measured from headstock portion to tailstock portion；And the detection vehicle to be measured passes through the height
Speed during detection device；
According to the multiple bodywork height data, determine that first coordinate points concentrate the vertical coordinate parameter of each coordinate points；
According to the speed and the prefixed time interval, the horizontal coordinate parameter of each coordinate points is determined.
3. method according to claim 2, which is characterized in that according to the multiple bodywork height data, determine described first
Coordinate points concentrate the vertical coordinate parameter of each coordinate points, comprising:
Determine that the multiple bodywork height data are the vertical coordinate parameter that first coordinate points concentrate each coordinate points；Or,
Determine that the product of the multiple bodywork height data and the weight set concentrates each coordinate points as first coordinate points
Vertical coordinate parameter, wherein be the weight of the bodywork height data setting in characterization headstock portion, higher than the vehicle body to characterize tailstock portion
The weight of altitude information setting.
4. method as claimed in any one of claims 1-3, which is characterized in that according to the similarity distance of the difference, really
The vehicle of the fixed vehicle to be measured, comprising:
Determine similarity distance of the various respectively with the first coordinate point set, wherein any vehicle and described the
The similarity distance of one coordinate point set, at least one the template coordinate point set and the first coordinate point set for including for this kind of vehicle
The smallest similarity distance；
According to the various similarity distance with the first coordinate point set respectively, determine that the vehicle to be measured is respectively
The probability of every kind of vehicle, wherein similarity distance and probability negative correlation；
Determine the vehicle to be measured by the vehicle for the maximum probability determined in the multiple vehicle.
5. method as claimed in claim 4, which is characterized in that distinguished according to the various and the first coordinate point set
Similarity distance, determine that the vehicle to be measured is respectively the probability of every kind of vehicle, comprising:
Pass through formula Pi=e-di/μDetermine that the vehicle to be measured is respectively the relative probability of every kind of vehicle, wherein PiIndicate institute
State the probability that vehicle to be measured is respectively i-th kind of vehicle in the various, diIndicate in the various i-th kind of vehicle with
The similarity distance of the first coordinate point set, μ indicate the similarity of the various and the first coordinate point set respectively
The mean value of distance；
The relative probability is normalized, to obtain the probability that the vehicle to be measured is respectively every kind of vehicle.
6. method as claimed in any one of claims 1-3, which is characterized in that any that the multiple template coordinate points are concentrated
Template coordinate point set obtains in the following way:
Multiple height detection is carried out to by the template vehicle of height detecting device with prefixed time interval, to obtain the template
Multiple template height of vehicle body data of the vehicle from headstock portion to tailstock portion；And the template vehicle is detected described in
Template vehicle speed during height detecting device；Wherein, the template vehicle is any template coordinate point set characterization
Vehicle roof outline line belonging to vehicle；
According to the multiple template height of vehicle body data, determine that any template coordinate points concentrate the vertical of each coordinate points
Coordinate parameters；
According to the template vehicle speed and the prefixed time interval, determine that any template coordinate points concentrate each coordinate points
Horizontal coordinate parameter.
7. a kind of vehicle type recognition device characterized by comprising
Coordinate point set determining module, for determining the first coordinate points for characterizing vehicle roof contour line to be measured under side view visual angle
Collection, each coordinate points that first coordinate points are concentrated represent a coordinate position；
Similarity is apart from determining module, for being based on preset similarity distance algorithm, determining the first coordinate point set and belonging to
Similarity distance between the multiple template coordinate point set of various respectively, wherein every kind of vehicle in the various
Including at least one template coordinate point set, each template coordinate point set characterizes vehicle vehicle roof outline line under a kind of side view visual angle, phase
Distance like degree distance to indicate the similarity between point set；
Vehicle determining module determines the vehicle of the vehicle to be measured for the similarity distance according to the difference.
8. device as claimed in claim 7, which is characterized in that the coordinate point set determining module is used for:
Multiple height detection is carried out to by the vehicle to be measured of height detecting device with prefixed time interval, described in obtaining
Multiple bodywork height data of the vehicle to be measured from headstock portion to tailstock portion；And the detection vehicle to be measured passes through the height
Speed during detection device；
According to the multiple bodywork height data, determine that first coordinate points concentrate the vertical coordinate parameter of each coordinate points；
According to the speed and the prefixed time interval, the horizontal coordinate parameter of each coordinate points is determined.
9. device as claimed in claim 8, which is characterized in that the coordinate point set determining module is used for:
Determine that the multiple bodywork height data are the vertical coordinate parameter that first coordinate points concentrate each coordinate points；Or,
Determine that the product of the multiple bodywork height data and the weight set concentrates each coordinate points as first coordinate points
Vertical coordinate parameter, wherein be the weight of the bodywork height data setting in characterization headstock portion, higher than the vehicle body to characterize tailstock portion
The weight of altitude information setting.
10. device as claimed in any one of claims 7-9, which is characterized in that the vehicle determining module is used for:
Determine similarity distance of the various respectively with the first coordinate point set, wherein any vehicle and described the
The similarity distance of one coordinate point set, at least one the template coordinate point set and the first coordinate point set for including for this kind of vehicle
The smallest similarity distance；
According to the various similarity distance with the first coordinate point set respectively, determine that the vehicle to be measured is respectively
The probability of every kind of vehicle, wherein similarity distance and probability negative correlation；
Determine the vehicle to be measured by the vehicle for the maximum probability determined in the multiple vehicle.
11. device as claimed in claim 10, which is characterized in that the vehicle determining module is used for:
Pass through formula Pi=e-di/μDetermine that the vehicle to be measured is respectively the relative probability of every kind of vehicle, wherein PiIndicate institute
State the probability that vehicle to be measured is respectively i-th kind of vehicle in the various, diIndicate in the various i-th kind of vehicle with
The similarity distance of the first coordinate point set, μ indicate the similarity of the various and the first coordinate point set respectively
The mean value of distance；
The relative probability is normalized, to obtain the probability that the vehicle to be measured is respectively every kind of vehicle.
12. device as claimed in any one of claims 7-9, which is characterized in that described device further includes template coordinate point set
Determining module, any template coordinate point set concentrated for obtaining the multiple template coordinate points in the following way:
Multiple height detection is carried out to by the template vehicle of height detecting device with prefixed time interval, to obtain the template
Multiple template height of vehicle body data of the vehicle from headstock portion to tailstock portion；And the template vehicle is detected described in
Template vehicle speed during height detecting device；Wherein, the template vehicle is any template coordinate point set characterization
Vehicle roof outline line belonging to vehicle；
According to the multiple template height of vehicle body data, determine that any template coordinate points concentrate the vertical of each coordinate points
Coordinate parameters；
According to the template vehicle speed and the prefixed time interval, determine that any template coordinate points concentrate each coordinate points
Horizontal coordinate parameter.
13. a kind of computer equipment, which is characterized in that the computer equipment includes:
At least one processor, and
The memory being connect at least one described processor；
Wherein, the memory is stored with the instruction that can be executed by least one described processor, at least one described processor
By executing the instruction of the memory storage, such as method of any of claims 1-6 is executed.
14. a kind of computer readable storage medium, it is characterised in that:
The computer-readable recording medium storage has computer instruction, when the computer instruction is run on computers,
So that computer executes such as method of any of claims 1-6.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810228691.9A CN110307809B (en) | 2018-03-20 | 2018-03-20 | Vehicle type recognition method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810228691.9A CN110307809B (en) | 2018-03-20 | 2018-03-20 | Vehicle type recognition method and device |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110307809A true CN110307809A (en) | 2019-10-08 |
| CN110307809B CN110307809B (en) | 2021-08-06 |
Family
ID=68073387
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810228691.9A Active CN110307809B (en) | 2018-03-20 | 2018-03-20 | Vehicle type recognition method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110307809B (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114882708A (en) * | 2022-07-11 | 2022-08-09 | 临沂市公路事业发展中心 | Vehicle identification method based on monitoring video |
Citations (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1996346A (en) * | 2005-12-29 | 2007-07-11 | 电子科技大学 | Vehicle recognition algorithm based on contour |
| CN201051310Y (en) * | 2007-06-21 | 2008-04-23 | 南京理工大学 | Millimeter wave and ultrasonic wave dual frequency traffic parameter detector with car type recognition function |
| CN101859491A (en) * | 2009-04-10 | 2010-10-13 | 张高军 | Method for obtaining longitudinal profile pattern of mobile vehicle and device thereof |
| CN102708694A (en) * | 2012-06-19 | 2012-10-03 | 公安部第三研究所 | Automatic motorcycle type identification system and method based on high-speed pulse laser scanning |
| CN102930242A (en) * | 2012-09-12 | 2013-02-13 | 上海交通大学 | Bus type identifying method |
| US20130182908A1 (en) * | 2012-01-17 | 2013-07-18 | Oki Electric Industry Co., Ltd. | Vehicle type identification device |
| CN104064030A (en) * | 2014-07-01 | 2014-09-24 | 武汉万集信息技术有限公司 | Vehicle type identification method and vehicle type identification system |
| CN104361752A (en) * | 2014-10-27 | 2015-02-18 | 北京握奇智能科技有限公司 | Laser scanning based vehicle type recognition method for free flow charging |
| CN105426922A (en) * | 2015-12-08 | 2016-03-23 | 同方威视技术股份有限公司 | Train type recognition method and system as well as safety inspection method and system |
| KR20160057059A (en) * | 2014-11-13 | 2016-05-23 | 주식회사한맥아이피에스 | Method and system for recognizing vehicle license plate |
| CN106469299A (en) * | 2016-08-31 | 2017-03-01 | 北京邮电大学 | A kind of vehicle search method and device |
| CN107273792A (en) * | 2017-04-27 | 2017-10-20 | 北京果毅科技有限公司 | Model recognizing method and device |
-
2018
- 2018-03-20 CN CN201810228691.9A patent/CN110307809B/en active Active
Patent Citations (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1996346A (en) * | 2005-12-29 | 2007-07-11 | 电子科技大学 | Vehicle recognition algorithm based on contour |
| CN201051310Y (en) * | 2007-06-21 | 2008-04-23 | 南京理工大学 | Millimeter wave and ultrasonic wave dual frequency traffic parameter detector with car type recognition function |
| CN101859491A (en) * | 2009-04-10 | 2010-10-13 | 张高军 | Method for obtaining longitudinal profile pattern of mobile vehicle and device thereof |
| US20130182908A1 (en) * | 2012-01-17 | 2013-07-18 | Oki Electric Industry Co., Ltd. | Vehicle type identification device |
| CN102708694A (en) * | 2012-06-19 | 2012-10-03 | 公安部第三研究所 | Automatic motorcycle type identification system and method based on high-speed pulse laser scanning |
| CN102930242A (en) * | 2012-09-12 | 2013-02-13 | 上海交通大学 | Bus type identifying method |
| CN104064030A (en) * | 2014-07-01 | 2014-09-24 | 武汉万集信息技术有限公司 | Vehicle type identification method and vehicle type identification system |
| CN104361752A (en) * | 2014-10-27 | 2015-02-18 | 北京握奇智能科技有限公司 | Laser scanning based vehicle type recognition method for free flow charging |
| KR20160057059A (en) * | 2014-11-13 | 2016-05-23 | 주식회사한맥아이피에스 | Method and system for recognizing vehicle license plate |
| CN105426922A (en) * | 2015-12-08 | 2016-03-23 | 同方威视技术股份有限公司 | Train type recognition method and system as well as safety inspection method and system |
| CN106469299A (en) * | 2016-08-31 | 2017-03-01 | 北京邮电大学 | A kind of vehicle search method and device |
| CN107273792A (en) * | 2017-04-27 | 2017-10-20 | 北京果毅科技有限公司 | Model recognizing method and device |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114882708A (en) * | 2022-07-11 | 2022-08-09 | 临沂市公路事业发展中心 | Vehicle identification method based on monitoring video |
| CN114882708B (en) * | 2022-07-11 | 2022-09-30 | 临沂市公路事业发展中心 | Vehicle identification method based on monitoring video |
Also Published As
| Publication number | Publication date |
|---|---|
| CN110307809B (en) | 2021-08-06 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109870680B (en) | Target classification method and device | |
| US9488725B2 (en) | Method and device for detecting objects in the surroundings of a vehicle | |
| CN100520853C (en) | Vehicle type classification method based on single frequency continuous-wave radar | |
| CN102248947B (en) | Object and vehicle detecting and tracking using a 3-D laser rangefinder | |
| CN112270252A (en) | Multi-vehicle target identification method for improving YOLOv2 model | |
| CN110794406B (en) | Multi-source sensor data fusion system and method | |
| CN106295459A (en) | Based on machine vision and the vehicle detection of cascade classifier and method for early warning | |
| CN112037521B (en) | Vehicle type identification method and hazardous chemical substance vehicle identification method | |
| CN114660330A (en) | Method for determining uncertainty estimate of estimated velocity | |
| CN106080397B (en) | Self-adaption cruise system and mobile unit | |
| CN107796373A (en) | A kind of distance-finding method of the front vehicles monocular vision based on track plane geometry model-driven | |
| CN109633686B (en) | Method and system for detecting ground obstacle based on laser radar | |
| CN113378751A (en) | Traffic target identification method based on DBSCAN algorithm | |
| CN104608768B (en) | A kind of objects ahead vehicle enters bend and carries out the condition discriminating apparatus and method of lane-change | |
| CN114241448A (en) | Method and device for obtaining heading angle of obstacle, electronic equipment and vehicle | |
| CN110307809A (en) | A kind of model recognizing method and device | |
| WO2007046336A1 (en) | Object recognizing device | |
| US20230245466A1 (en) | Vehicle Lidar System and Object Classification Method Therewith | |
| US20050177336A1 (en) | Method for identifying object constellations using distance signals | |
| CN113361079B (en) | Road surface flatness detection method, device, equipment and storage medium | |
| Xu et al. | Roadside estimation of a vehicle’s center of gravity height based on an improved single-stage detection algorithm and regression prediction technology | |
| CN114862944A (en) | Vehicle pose detection method and device and electronic equipment | |
| Lidman et al. | Clustering, shape extraction and velocity estimation applied to radar detections | |
| CN115015962A (en) | Collision detection method, system, equipment and storage medium based on shore bridge | |
| CN113589288A (en) | Target screening method, device and equipment based on millimeter wave radar and storage medium |
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 |