CN106250827B - Vehicle color identification method and device - Google Patents
Vehicle color identification method and device Download PDFInfo
- Publication number
- CN106250827B CN106250827B CN201610584806.9A CN201610584806A CN106250827B CN 106250827 B CN106250827 B CN 106250827B CN 201610584806 A CN201610584806 A CN 201610584806A CN 106250827 B CN106250827 B CN 106250827B
- Authority
- CN
- China
- Prior art keywords
- color
- vehicle
- number plate
- shooting
- target vehicle
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/50—Context or environment of the image
- G06V20/56—Context or environment of the image exterior to a vehicle by using sensors mounted on the vehicle
- G06V20/58—Recognition of moving objects or obstacles, e.g. vehicles or pedestrians; Recognition of traffic objects, e.g. traffic signs, traffic lights or roads
- G06V20/584—Recognition of moving objects or obstacles, e.g. vehicles or pedestrians; Recognition of traffic objects, e.g. traffic signs, traffic lights or roads of vehicle lights or traffic lights
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/40—Extraction of image or video features
- G06V10/50—Extraction of image or video features by performing operations within image blocks; by using histograms, e.g. histogram of oriented gradients [HoG]; by summing image-intensity values; Projection analysis
- G06V10/507—Summing image-intensity values; Histogram projection analysis
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Multimedia (AREA)
- Theoretical Computer Science (AREA)
- Image Processing (AREA)
- Traffic Control Systems (AREA)
Abstract
The application is about a kind of vehicle color identification method and device, which comprises obtains the shooting number plate color for the target vehicle for including in target vehicle picture;According to the color relationship between the shooting number plate color and the real vehicles color of the target vehicle, the real vehicles color of the target vehicle is determined.Vehicle pictures captured by the equipment such as alert to bayonet, electricity may be implemented using this method and carry out color correct, so that the color of the vehicle pictures of shooting is closer to true vehicle color, thus when being retrieved according to vehicle color, so that search result is accurate as much as possible.
Description
Technical field
This application involves technical field of image processing more particularly to vehicle color identification methods and device.
Background technique
With the development of network technology and surging for vehicle fleet size, intelligentized traffic administration is particularly important.In intelligence
In the traffic administration of change, road vehicle intelligent monitoring and recording system (intelligent monitoring and recording
System of vehicles on highways, abbreviation bayonet) and red-lamp running automatic recording system (automatic
Detecting and recording system for violation of traffic signal, referred to as electricity are alert) application
It is very extensive.Automatic collection and processing can be carried out to the information of vehicles on monitored road surface by tollgate devices, by electricity, police is set
The standby behavior that can make a dash across the red light to motor vehicle in specific lane carries out uninterrupted automatic detection and record.
Vehicle color can be management, tracking, the retrieval etc. of vehicle as an important and very intuitive vehicle attribute
Important evidence is provided.However, tollgate devices and the alert equipment of electricity are highly susceptible to the influence of the factors such as environment, illumination, cause to be adopted
There are larger differences compared to true vehicle color for the vehicle pictures collected.To be set retrieving bayonet according to vehicle color
When the standby or electric alert collected vehicle pictures of equipment institute, it is easy to the problems such as erroneous detection, missing inspection occur.
Summary of the invention
To overcome the problems in correlation technique, this application provides vehicle color identification method and devices, it is therefore an objective to
Realize that vehicle pictures captured by the equipment such as alert to bayonet, electricity carry out color correct, so that the color of the vehicle pictures of shooting is more
Close to true vehicle color, thus when being retrieved according to vehicle color, so that search result is accurate as much as possible.
According to the embodiment of the present application in a first aspect, providing a kind of vehicle color identification method, which comprises
Obtain the shooting number plate color for the target vehicle for including in target vehicle picture;
According to the color relationship between the shooting number plate color and the real vehicles color of the target vehicle, institute is determined
State the real vehicles color of target vehicle.
Optionally, the color according between the shooting number plate color and the real vehicles color of the target vehicle
Relationship determines the real vehicles color of the target vehicle, comprising:
Picture analyzing is carried out to the target vehicle picture, obtains the vehicle of the target vehicle;
According to the corresponding relationship of the vehicle and number plate color, the true number plate color of the target vehicle is obtained;It is described
Number plate color include it is following at least one: number plate background colour, number plate character color;
According to the shooting number plate color of the target vehicle and the true number plate color, the shooting number plate is obtained
Color component proportionate relationship between color and the true number plate color;
According to the color component of the color component proportionate relationship and the shooting vehicle color of the target vehicle, obtain
To the color component of the real vehicles color of the target vehicle, the real vehicles color of the target vehicle is determined.
Optionally, the color according between the shooting number plate color and the real vehicles color of the target vehicle
Relationship determines the real vehicles color of the target vehicle, comprising:
According to the shooting number plate color, search the colour atla library pre-saved, and find in the colour atla library including
The target colour atla of the shooting number plate color;
The shooting vehicle color of the target vehicle is found on the target colour atla;
According to the colour atla position of the shooting vehicle color in the target colour atla, search standard color card, determine described in
The real vehicles color of target vehicle;The target colour atla and standard color card be respectively on same colour atla position shooting color and
Corresponding true colors.
Optionally, the method also includes:
After the real vehicles color for determining the target vehicle, the true brightness value of the real vehicles color is obtained, and
Obtain the shooting brightness value of the shooting vehicle color of the target vehicle;
Intensity deviation value is determined according to the true brightness value and the shooting brightness value;
The vehicle pictures are converted into YUV image;
According to the intensity deviation value, intensity deviation is carried out in the Y-component of the YUV image;
YUV image after intensity deviation is converted into RGB image;
The vehicle color for determining the target vehicle for including in the RGB image is the target vehicle after brightness correction
Real vehicles color.
Optionally, the method also includes:
When determining the target vehicle picture is gray level image, the ash of the shooting vehicle color of the target vehicle is determined
Angle value;
It, will be locating for the gray value of the shooting vehicle color according to the corresponding relationship of preset intensity value ranges and color
Color corresponding to intensity value ranges is determined as the real vehicles color of the target vehicle.
According to the second aspect of the embodiment of the present application, a kind of vehicle color identification device is provided, described device includes:
Module is obtained, for obtaining the shooting number plate color for the target vehicle for including in target vehicle picture;
Determining module, for according to the face between the shooting number plate color and the real vehicles color of the target vehicle
Color relationship determines the real vehicles color of the target vehicle.
Optionally, the determining module, comprising:
Model analysis submodule obtains the target vehicle for carrying out picture analyzing to the target vehicle picture
Vehicle;
Number plate color determines submodule, for the corresponding relationship according to the vehicle and number plate color, obtains the target
The true number plate color of vehicle;The number plate color include it is following at least one: number plate background colour, number plate character color;
Relationship determines submodule, for the shooting number plate color and the true number plate face according to the target vehicle
Color obtains the color component proportionate relationship between the shooting number plate color and the true number plate color;
Vehicle color determines submodule, for the bat according to the color component proportionate relationship and the target vehicle
The color component for taking the photograph vehicle color obtains the color component of the real vehicles color of the target vehicle, determines the target carriage
Real vehicles color.
Optionally, the determining module, comprising:
First searches submodule, for searching the colour atla library pre-saved, and described according to the shooting number plate color
The target colour atla including the shooting number plate color is found in colour atla library;
Second searches submodule, for finding the shooting vehicle color of the target vehicle on the target colour atla;
Third searches submodule, for the colour atla position according to the shooting vehicle color in the target colour atla, looks into
Standard color card is looked for, determines the real vehicles color of the target vehicle;The target colour atla and standard color card are in same colour atla position
Set is shooting color and corresponding true colors respectively.
Optionally, described device further include:
Brightness calculation module after the real vehicles color for determining the target vehicle, obtains the real vehicles face
The true brightness value of color, and obtain the shooting brightness value of the shooting vehicle color of the target vehicle;
Calculations of offset module, for determining intensity deviation value according to the true brightness value and the shooting brightness value;
Image conversion module, for the vehicle pictures to be converted to YUV image;
Intensity deviation module, for it is inclined to carry out brightness in the Y-component of the YUV image according to the intensity deviation value
It moves;
Described image conversion module, is also used to: the YUV image after intensity deviation is converted to RGB image;
The determining module, is also used to: determining that the vehicle color for the target vehicle for including in the RGB image is repaired for brightness
The real vehicles color of the target vehicle after just.
Optionally, described device further include:
Gray value determining module, for determining the target carriage when determining the target vehicle picture is gray level image
Shooting vehicle color gray value;
The determining module, is also used to: according to the corresponding relationship of preset intensity value ranges and color, by the shooting vehicle
Color corresponding to intensity value ranges locating for the gray value of color, is determined as the real vehicles color of the target vehicle.
As seen from the above-described embodiment, by analyzing target vehicle picture, the available shooting to target vehicle
Number plate color can determine target vehicle further according to the color relationship between the shooting number plate color and real vehicles color
Real vehicles color.When being retrieved according to vehicle color, due to the available target for including into target vehicle picture
, there is the problems such as erroneous detection, missing inspection so as to be effectively prevented from the real vehicles color of vehicle.
Detailed description of the invention
The drawings herein are incorporated into the specification and forms part of this specification, and shows and meets implementation of the invention
Example, and be used to explain the principle of the present invention together with specification.
Fig. 1 is one embodiment flow chart of the application vehicle color identification method;
Fig. 2A is another embodiment flow chart of the application vehicle color identification method;
Fig. 2 B shows Part VII in " GA24.7-2005 motor vehicle register information code ": the portion of number plate category codes
Divide information;
Fig. 3 is another embodiment flow chart of the application vehicle color identification method;
Fig. 4 is the further embodiment flow chart of the application vehicle color identification method;
Fig. 5 is the hardware block diagram of the network equipment in the embodiment of the present invention;
Fig. 6 is one embodiment block diagram of the application vehicle color identification device;
Fig. 7 is another embodiment block diagram of the application vehicle color identification device;
Fig. 8 is another embodiment block diagram of the application vehicle color identification device.
Specific embodiment
Example embodiments are described in detail here, and the example is illustrated in the accompanying drawings.Following description is related to
When attached drawing, unless otherwise indicated, the same numbers in different drawings indicate the same or similar elements.Following exemplary embodiment
Described in embodiment do not represent all embodiments consistented with the present invention.On the contrary, they be only with it is such as appended
The example of device and method being described in detail in claims, some aspects of the invention are consistent.
With the development of internet, in the construction of safe city, commonly used Intelligent traffic management systems.In intelligence
In traffic control system, can by the network equipment, for example, tollgate devices, the alert equipment of electricity, etc., to road vehicle into
Row Image Acquisition, can be with for example, make a dash across the red light to detect whether a vehicle commits unlawful acts by the collected vehicle pictures of institute
Vehicle is retrieved by institute's collected vehicle pictures, positioned, is tracked.In management, retrieval, the tracking etc. for realizing vehicle
When function, vehicle color is as an important and very intuitive vehicle attribute.
However, since the above-mentioned described network equipment is in the actual environment, to carry out Image Acquisition to road vehicle,
For example, the alert equipment of electricity is mountable on the intersection and section of signal control, to carry out to the vehicle travelled on specific lane
Monitoring and record.To which the collected vehicle pictures of the network equipment will likely be by the colour temperature shadow in actual environment
It rings, causes the color of vehicle in vehicle pictures not to be consistent with the true colors of the vehicle, or even differ greatly, for example, vehicle
True colors be yellow, and due to the colour temperature of shooting environmental influence, in vehicle pictures, it is white that the color of the vehicle, which is taken,
Color.To when being retrieved according to vehicle color to vehicle, it is easy to the problems such as erroneous detection, missing inspection occur.
In order to efficiently solve the above problem, this application provides vehicle color identification method and devices, to realize basis
The collected vehicle pictures of network equipment institute are analyzed, and identify target carriage included in vehicle pictures as precisely as possible
Real vehicles color occur erroneous detection, missing inspection etc. to be effectively prevented from when being retrieved according to color to vehicle and ask
Topic.
Following examples provided herein are illustrated vehicle color identification method provided herein.
Referring to Fig. 1, it is one embodiment flow chart of the application vehicle color identification method, may comprise steps of:
Step S101: the shooting number plate color for the target vehicle for including in target vehicle picture is obtained.
In this application, described target vehicle picture can refer to the network equipment, for example, the alert equipment of tollgate devices, electricity
Deng carrying out shooting obtained picture to target vehicle to be identified on road.
After shooting obtains target vehicle picture, picture analyzing can be carried out according to the target vehicle picture, to obtain the mesh
The number plate color for marking target vehicle included in vehicle pictures for convenience, will be in target vehicle picture in the application
Number plate color be known as shoot number plate color.
Under normal conditions, number plate color may include two parts: number plate background colour, number plate character color.For example, kart
Number plate background colour be blue, and number plate character color be white;In another example the number plate background colour of full-sized car is yellow, and number
Board character color is black.In this application, described shooting number plate color includes at least shooting number plate background colour and shooting number plate
Character color one of those.
Step S102: it is closed according to the color between the shooting number plate color and the real vehicles color of the target vehicle
System, determines the real vehicles color of the target vehicle.
In an optional implementation, true number plate color first can be obtained according to shooting number plate color, then determine
The color relationship between number plate color and true number plate color is shot, also, due to shooting number plate color and shooting vehicle color
It is that the color relationship is also complied between vehicle color and real vehicles color to shoot in same target vehicle picture.
So as to color relationship accordingly, the real vehicles color of target vehicle is determined.The detailed process of which may refer to following
Fig. 2A is detailed description of the illustrated embodiment, is not described further first herein.
In another optional implementation, shooting number plate color and the target carriage can be indicated by standard color card
Real vehicles color between color relationship.For example, shooting number plate color and the shooting vehicle color of target vehicle can be with
On same target colour atla, further according to the relationship between the target colour atla and standard color card, target vehicle can be determined
Real vehicles color.The detailed process of the implementation may refer to the detailed description of following embodiment illustrated in fig. 3, herein first not
It is described further.
In addition, in certain extreme cases, may and color difference, example be not present in taken target vehicle picture
If taken target vehicle picture is gray level image, signified " gray level image " be can be understood as usually said herein
" black white image ".When determining the target vehicle picture is gray level image, in this application, can continue to determine the target carriage
Shooting vehicle color gray value, it is assumed that obtained gray value be 25.
It is subsequent, the corresponding relationship of the intensity value ranges pre-saved and color, this place can be searched according to the gray value
" corresponding relationship " referred to, which can represent, belongs to true colors corresponding to the shooting color of a certain intensity value ranges, for example, grey
When angle value range is 10~20, corresponding color is yellow;When intensity value ranges are 21~30, corresponding color is indigo plant
Color.So, which is searched by obtained gray value, it can be deduced that real vehicles corresponding to the shooting vehicle color
Color is blue.
As seen from the above-described embodiment, by analyzing target vehicle picture, the available shooting to target vehicle
Number plate color can determine target vehicle further according to the color relationship between the shooting number plate color and real vehicles color
Real vehicles color.When being retrieved according to vehicle color, due to the available target for including into target vehicle picture
, there is the problems such as erroneous detection, missing inspection so as to be effectively prevented from the real vehicles color of vehicle.
Fig. 2A is referred to, is another embodiment flow chart of the application vehicle color identification method, shown in the Fig. 2A
On the basis of process process shown in above-mentioned Fig. 1, describes emphatically and true number plate color is determined according to shooting number plate color, then
It determines color relationship between the two, the process of the real vehicles color of target vehicle is determined according to the color relationship, can wrap
Include following steps:
Step S201: picture analyzing is carried out to target vehicle picture, obtains the vehicle of target vehicle.
In this step, it is how to determine the vehicle for the target vehicle for including in target vehicle picture, may refer to existing
There is the associated description in technology, the application is no longer described further this.
Step S202: according to the corresponding relationship of the vehicle and number plate color, the true number plate of the target vehicle is obtained
Color;The number plate color include it is following at least one: number plate background colour, number plate character color.
Following Fig. 2 B shows Part VII in " GA24.7-2005 motor vehicle register information code ": number plate type generation
The partial information of code.
When executing the step S201, after getting the vehicle of target vehicle, then can be searched according to the vehicle shown in Fig. 2 B
Number plate category codes.For example, the vehicle of target vehicle is full-sized car, then the true number plate of the available target vehicle is carried on the back
Scenery is yellow, and true number plate character color is black.It in this application, can be by true number plate background colour, true number plate character
True number plate color of at least one of the color as the target vehicle.
In addition, under certain conditions, only possibly can not confirm true number plate color by vehicle.For example, getting target
The vehicle of vehicle is kart, but at this point, only possibly true number plate color can not be confirmed by vehicle, for example, consulate's automobile
Kart may also be belonged to, pass through number plate category codes shown in Fig. 2 B, it can be deduced that the number plate background of common compact automobile
Color is blue, and the number plate background colour of consulate's kart is black, thus, only pass through vehicle and can not confirm true number plate face
Color.At this point, then can be according to shooting number plate color and vehicle, the common true number plate color for confirming target vehicle.Specifically may be used
To include: to shoot number plate color by analysis, determining in the shooting character color in shooting number plate color whether there is red, when
There are when red, then can be confirmed as consulate's kart, so that it is determined that the true number plate background colour of target vehicle is black, very
Real number plate character color is white, and has red " neck " word.
Step S203: it according to the shooting number plate color of the target vehicle and the true number plate color, obtains described
Shoot the color component proportionate relationship between number plate color and the true number plate color.
It is as follows, it will be illustrated so that number plate color includes number plate background colour as an example:
Assuming that the shooting number plate background colour got be (Rrp, Grp, Brp), true number plate background colour be (Rtp, Gtp,
Btp)。
In an optional implementation, shooting number plate color and true number plate can be obtained by following calculating process
Color component proportionate relationship between color:
It can be concluded that yield value Rgain of the shooting number plate background colour compared to true number plate background colour, in R component
=Rrp/Rtp;
Yield value Ggain=Grp/Gtp on G component;
Yield value Bgain=Brp/Btp on B component.
At this point it is possible to respectively indicate out shooting number plate color and true number plate color in R using Rgain, Ggain, Bgain
Component, G component, the proportionate relationship on B component.
It, can be empty to RGB image color in order to remove the influence of illumination and shade in another optional implementation
Between be normalized.It specifically, may include: that shooting number plate background colour after normalized isReally number plate background colour is
The color component between shooting number plate color and true number plate color is calculated according to the color value after normalization
Proportionate relationship:
Yield value in R component
Yield value Ggain=1 on G component;
Yield value on B component
Step S204: according to the color component proportionate relationship and the face of the shooting vehicle color of the target vehicle
Colouring component obtains the color component of the real vehicles color of the target vehicle, determines the real vehicles face of the target vehicle
Color.
Since shooting number plate color and shooting vehicle color are located on same target vehicle picture, thus, shoot vehicle
Between color and real vehicles color, the color relationship between shooting number plate color and true number plate color is also complied with.It then can be with
Real vehicles color is calculated according to the color relationship, and shooting vehicle color.
Assuming that the shooting vehicle color of target vehicle be (Rrb, Grb, Brb), real vehicles color be (Rtb, Gtb,
Btb)。
In an optional implementation, with the calculating in first shown in step S203 optional implementation
As a result based on, it can be deduced that:
Rtb=Rrb/Rgain=(Rrb*Rtp)/Rrp;
Gtb=Grb/Ggain=(Grb*Gtp)/Grp;
Btb=Brb/Bgain=(Brb*Btp)/Brp.
So as to obtain target vehicle real vehicles color R component value, G component value, B component value, accordingly can be with
Determine the real vehicles color of target vehicle.
It is consistent with the normalized in step S203 in another optional implementation, after normalized
Shooting vehicle color isReal vehicles color after normalized is
Based on calculated result in second optional implementation shown in the step S203, it can be deduced that:
The real vehicles color after above-mentioned resulting normalized is subjected to brightness transition again.It should be noted that by
In in the present embodiment, only Grb be it is known, then by after normalized real vehicles color carry out brightness transition after,
Obtained true vehicle color isIt is understood that is obtained at this time is true
Real vehicle color may be inaccurate in brightness, at this point, brightness correction can also be carried out to the true vehicle color.It is bright
Spending modified detailed process may refer to associated description in following embodiments, not be described further first herein.
As seen from the above-described embodiment, by obtaining the true number plate color of target vehicle according to the vehicle of target vehicle, really
Color relationship between fixed true number plate color and shooting number plate color obtains very according to the color relationship and shooting vehicle color
Real vehicle color.When being retrieved according to vehicle color, due to the available target vehicle for including into vehicle pictures
There is the problems such as erroneous detection, missing inspection so as to be effectively prevented from real vehicles color.
Fig. 3 is referred to, is another embodiment flow chart of the application vehicle color identification method, the stream shown in Fig. 3
On the basis of journey process shown in above-mentioned Fig. 1, the mistake that the real vehicles color of target vehicle is obtained by colour atla is described emphatically
Journey may comprise steps of:
Step S301: the shooting number plate color for the target vehicle for including in target vehicle picture is obtained.
The detailed description of the step may refer to the description in above-mentioned steps S101, does not remake be described in detail herein.
Step S302: according to the shooting number plate color, the colour atla library pre-saved is searched, and is looked into the colour atla library
Find the target colour atla including the shooting number plate color.
In this application, colour atla library can be preset, may include captured under various circumstances in the colour atla library
24 colour atla of standard, or the colour atla including more polychrome.It is shot for example, being included in the environment of colour temperature is 2000K in the colour atla library
24 colour atla of standard, colour temperature be 5000K in the environment of shoot 24 colour atla of standard, etc..Also, it can also be according to different rings
The frequency of occurrence in border is ranked up the colour atla in colour atla library with the sequence of frequency of occurrence from high to low.
According to the above-mentioned described colour atla library of the shooting number plate color lookup of target vehicle, found in colour atla library including
The target colour atla of the shooting number plate color.It should be noted that since the colour atla in colour atla library has been carried out according to set rule
Sequence, to will can initially find the colour atla including the shooting number plate color as aim colour when searching target colour atla
Card, once find target colour atla, it can stop searching.To preferentially select the higher colour atla of frequency of occurrence as target
Colour atla.
Step S303: the shooting vehicle color of the target vehicle is found on the target colour atla.
Step S304: according to colour atla position of the shooting vehicle color in the target colour atla, searching standard color card,
Determine the real vehicles color of the target vehicle;The target colour atla and standard color card are to clap respectively on same colour atla position
Take the photograph color and corresponding true colors.
In step S303 and step S304, in step s 302 on identified target colour atla, target vehicle is found
Shooting vehicle color, obtain shooting vehicle color colour atla position locating on target colour atla.Then, according to the colour atla position
Lookup standard color card is set, such as searches 24 colour atlas of standard, color of the standard color card on the colour atla position is determined as target
The real vehicles color of vehicle.
As seen from the above-described embodiment, by the shooting number plate color according to target vehicle, the colour atla library pre-saved is searched,
It determines the target colour atla including the shooting number plate color, the shooting vehicle color of target vehicle is found on the target colour atla,
Colour atla position of the shooting vehicle color on the target colour atla is obtained, according to the colour atla location lookup standard color card, determines mesh
Mark the real vehicles color of vehicle.When being retrieved according to vehicle color, include into vehicle pictures due to available
, there is the problems such as erroneous detection, missing inspection so as to be effectively prevented from the real vehicles color of target vehicle.
In addition, in this application, by executing above-mentioned Fig. 1, Fig. 2A, process shown in Fig. 3, can determine target carriage substantially
Real vehicles color.In subsequent practical application, when being retrieved according to vehicle color, in order to eliminate brightness of image to inspection
Vehicle color to be retrieved can also be normalized in the influence of hitch fruit.Detailed process may include:
Assuming that vehicle color to be retrieved is (Rs, Gs, Bs), then the vehicle color to be retrieved after normalized can be with
ForUsing the vehicle color to be retrieved after the normalized to by above-mentioned Fig. 1, Fig. 2A or Fig. 3
Shown in method treated that vehicle pictures are retrieved, can be effectively avoided since colour brightness influences, and caused by leak
Inspection problem.
By executing above-mentioned Fig. 1, Fig. 2A, process shown in Fig. 3, the real vehicles color of target vehicle can be determined.For
Further accurately determine that the real vehicles color of target vehicle in this application can also be to the real vehicles obtained
Color, which carries out brightness correction, can make basis for example, determining real vehicles color is modified to dark red or azarin from red
The search result that vehicle color carries out vehicle retrieval is more accurate.
Fig. 4 is referred to, is the further embodiment flow chart of the application vehicle color identification method, the stream shown in Fig. 4
Journey can the process shown in above-mentioned Fig. 1 or Fig. 2A or Fig. 3 on the basis of, describe emphatically in the true vehicle for determining target vehicle
After color, the process of brightness correction is carried out to real vehicles color, may comprise steps of:
Step S401: after the real vehicles color for determining the target vehicle, the true of the real vehicles color is obtained
Brightness value, and obtain the shooting brightness value of the shooting vehicle color of the target vehicle.
In this application, the true brightness value for obtaining real vehicles color can be calculated separately according to the following formula, with bat
Take the photograph the shooting brightness value of vehicle color:
Brightness value=0.299*R+0.587*G+0.114*B.
It is assumed that the true brightness value being calculated is 20, shooting brightness value is 15.
Step S402: intensity deviation value is determined according to the true brightness value and the shooting brightness value.
Intensity deviation value=true brightness value-shooting brightness value
Assuming that the intensity deviation value being calculated in the step is+5.
Step S403: the vehicle pictures are converted into YUV image.
Step S404: according to the intensity deviation value, intensity deviation is carried out in the Y-component of the YUV image.
Under normal conditions, the Y-component in YUV image can indicate brightness value.It in this application, then can be according to calculating
The intensity deviation value obtained is to carrying out intensity deviation in the Y-component of the YUV image of target vehicle picture.
Step S405: the YUV image after intensity deviation is converted into RGB image.
Step S406: determine that the vehicle color for the target vehicle for including in the RGB image is described after brightness correction
The real vehicles color of target vehicle.
YUV image after intensity deviation is reconverted into RGB image, thus then can be according to R included in RGB image
Component, G component, B component determine the real vehicles color of the target vehicle after brightness correction.
As seen from the above-described embodiment, after determining the real vehicles color of vehicle, pass through the brightness value of real vehicles color
With the brightness value of shooting vehicle color, intensity deviation value can be determined, real vehicles color is carried out according to the intensity deviation value
Brightness correction allows the real vehicles color after brightness correction more accurate.When being retrieved according to vehicle color, by
In the real vehicles color of the available target vehicle for including into vehicle pictures, missed so as to be effectively prevented from
The problems such as inspection, missing inspection.
Based on the inventive concept same with the above method, the embodiment of the invention also provides the realities of vehicle color identification device
Example is applied, which can be used in the equipment such as the network equipment, such as the alert equipment of tollgate devices, electricity.Wherein, which can pass through
Software realization can also be realized by way of hardware or software and hardware combining.Taking software implementation as an example, it anticipates as a logic
Device in justice is to be referred to computer program corresponding in nonvolatile memory by the processor of the network equipment where it
It enables and is read into memory what operation was formed.For hardware view, as shown in figure 5, being the network equipment in the embodiment of the present invention
Hardware block diagram, other than processor 51 shown in fig. 5, memory 52, network interface 53 and nonvolatile memory 54, also
It may include other hardware, such as photographic device, the forwarding chip for being responsible for processing message.
Fig. 6 is referred to, is one embodiment block diagram of the application vehicle color identification device, the apparatus may include: it obtains
Modulus block 61, determining module 62.
Wherein, the acquisition module 61 can be used for obtaining the shooting number plate for the target vehicle for including in target vehicle picture
Color;
The determining module 62 can be used for the real vehicles color according to shooting the number plate color and the target vehicle
Between color relationship, determine the real vehicles color of the target vehicle.
In one embodiment, which can also include (not showing that in Fig. 6): gray value determining module.
The gray value determining module can be used for when determining the target vehicle picture is gray level image, described in determination
The gray value of the shooting vehicle color of target vehicle;
In this embodiment, the determining module 62, can be also used for: according to pair of preset intensity value ranges and color
It should be related to, color corresponding to intensity value ranges locating for the gray value of the shooting vehicle color is determined as the target carriage
Real vehicles color.
Fig. 7 is referred to, is another embodiment block diagram of the application vehicle color identification device, which exists
On the basis of above-mentioned Fig. 6 shown device, in which:
The determining module 62 may include: that model analysis submodule 621, number plate color determine submodule 622, relationship
Determine that submodule 623, vehicle color determine submodule 624.
The model analysis submodule 621 can be used for carrying out picture analyzing to the target vehicle picture, obtain the mesh
Mark the vehicle of vehicle;
The number plate color determines submodule 622, can be used for being obtained according to the corresponding relationship of the vehicle and number plate color
The true number plate color of the target vehicle;The number plate color include it is following at least one: number plate background colour, number plate character
Color;
The relationship determines submodule 623, can be used for according to the shooting number plate color of the target vehicle and described true
Real number plate color obtains the color component proportionate relationship between the shooting number plate color and the true number plate color;
The vehicle color determines submodule 624, can be used for according to the color component proportionate relationship and the target
The color component of the shooting vehicle color of vehicle obtains the color component of the real vehicles color of the target vehicle, determines institute
State the real vehicles color of target vehicle.
Refer to Fig. 8, be the application vehicle color identification device another embodiment block diagram, the device shown in Fig. 8,
On the basis of above-mentioned Fig. 6 shown device, in which:
The determining module 62 may include: that the first lookup submodule 625, second searches submodule 626, third is searched
Submodule 627.
Wherein, this first searches submodule 625, can be used for according to the shooting number plate color, searches and pre-saves
Colour atla library, and the target colour atla including the shooting number plate color is found in the colour atla library;
The second lookup submodule 626, can be used for finding the shooting of the target vehicle on the target colour atla
Vehicle color;
The third searches submodule 627, can be used for the color according to the shooting vehicle color in the target colour atla
Card position searches standard color card, determines the real vehicles color of the target vehicle;The target colour atla and standard color card are same
It is shooting color and corresponding true colors respectively in card position of the same colour.
In one embodiment, on the basis of above-mentioned Fig. 6, Fig. 7, Fig. 8 shown device, vehicle color provided herein
Identification device can also include (being not shown in Fig. 6, Fig. 7, Fig. 8): brightness calculation module, calculations of offset module, image conversion
Module, intensity deviation module.
The brightness calculation module after being determined for the real vehicles color of the target vehicle, obtains described true
The true brightness value of vehicle color, and obtain the shooting brightness value of the shooting vehicle color of the target vehicle;
The calculations of offset module can be used for determining intensity deviation according to the true brightness value and the shooting brightness value
Value;
The image conversion module can be used for the vehicle pictures being converted to YUV image;
The intensity deviation module can be used for being carried out in the Y-component of the YUV image according to the intensity deviation value
Intensity deviation;
The image conversion module, can be also used for: the YUV image after intensity deviation is converted to RGB image;
In this embodiment, the determining module 62, can be also used for: determine the target carriage for including in the RGB image
Vehicle color be brightness correction after the target vehicle real vehicles color.
In another embodiment, which can also include (being not shown in Fig. 6, Fig. 7, Fig. 8): normalization module, inspection
Rope module.
Wherein, the normalization module can be used for working as according to vehicle color to be retrieved to the true of the target vehicle
When vehicle color is retrieved, the vehicle color to be retrieved is normalized;
The retrieval module can be used for according to the vehicle color to be retrieved after normalization, to the true of the target vehicle
Real vehicle color is retrieved.
Those skilled in the art after considering the specification and implementing the invention disclosed here, will readily occur to of the invention its
Its embodiment.The present invention is directed to cover any variations, uses, or adaptations of the invention, these modifications, purposes or
Person's adaptive change follows general principle of the invention and including the undocumented common knowledge in the art of the present invention
Or conventional techniques.The description and examples are only to be considered as illustrative, and true scope and spirit of the invention are by following
Claim is pointed out.
It should be understood that the present invention is not limited to the precise structure already described above and shown in the accompanying drawings, and
And various modifications and changes may be made without departing from the scope thereof.The scope of the present invention is limited only by the attached claims.
The foregoing is merely illustrative of the preferred embodiments of the present invention, is not intended to limit the invention, all in essence of the invention
Within mind and principle, any modification, equivalent substitution, improvement and etc. done be should be included within the scope of the present invention.
Claims (8)
1. a kind of vehicle color identification method, which is characterized in that the described method includes:
Obtain the shooting number plate color for the target vehicle for including in target vehicle picture;
According to the color relationship between the shooting number plate color and the real vehicles color of the target vehicle, the mesh is determined
Mark the real vehicles color of vehicle, comprising:
Picture analyzing is carried out to the target vehicle picture, obtains the vehicle of the target vehicle;
According to the corresponding relationship of the vehicle and number plate color, the true number plate color of the target vehicle is obtained;The number plate
Color include it is following at least one: number plate background colour, number plate character color;
According to the shooting number plate color of the target vehicle and the true number plate color, the shooting number plate color is obtained
With the color component proportionate relationship between the true number plate color;
According to the color component of the color component proportionate relationship and the shooting vehicle color of the target vehicle, institute is obtained
The color component for stating the real vehicles color of target vehicle determines the real vehicles color of the target vehicle.
2. the method according to claim 1, wherein described according to the shooting number plate color and the target carriage
Real vehicles color between color relationship, determine the real vehicles color of the target vehicle, comprising:
According to the shooting number plate color, the colour atla library pre-saved is searched, and is found in the colour atla library including described
Shoot the target colour atla of number plate color;
The shooting vehicle color of the target vehicle is found on the target colour atla;
According to colour atla position of the shooting vehicle color in the target colour atla, standard color card is searched, determines the target
The real vehicles color of vehicle;The target colour atla and standard color card are shooting color and correspondence respectively on same colour atla position
True colors.
3. any method according to claim 1~2, which is characterized in that the method also includes:
After the real vehicles color for determining the target vehicle, the true brightness value of the real vehicles color is obtained, and is obtained
The shooting brightness value of the shooting vehicle color of the target vehicle;
Intensity deviation value is determined according to the true brightness value and the shooting brightness value;
The vehicle pictures are converted into YUV image;
According to the intensity deviation value, intensity deviation is carried out in the Y-component of the YUV image;
YUV image after intensity deviation is converted into RGB image;
The vehicle color for determining the target vehicle for including in the RGB image is the true of the target vehicle after brightness correction
Vehicle color.
4. the method according to claim 1, wherein the method also includes:
When determining the target vehicle picture is gray level image, the gray scale of the shooting vehicle color of the target vehicle is determined
Value;
According to the corresponding relationship of preset intensity value ranges and color, by gray scale locating for the gray value of the shooting vehicle color
It is worth color corresponding to range, is determined as the real vehicles color of the target vehicle.
5. a kind of vehicle color identification device, which is characterized in that described device includes:
Module is obtained, for obtaining the shooting number plate color for the target vehicle for including in target vehicle picture;
Determining module, for being closed according to the color between the shooting number plate color and the real vehicles color of the target vehicle
System, determines the real vehicles color of the target vehicle, comprising:
Model analysis submodule obtains the vehicle of the target vehicle for carrying out picture analyzing to the target vehicle picture;
Number plate color determines submodule, for the corresponding relationship according to the vehicle and number plate color, obtains the target vehicle
True number plate color;The number plate color include it is following at least one: number plate background colour, number plate character color;
Relationship determines submodule, for the shooting number plate color and the true number plate color according to the target vehicle, obtains
Color component proportionate relationship between the shooting number plate color and the true number plate color;
Vehicle color determines submodule, for the shooting vehicle according to the color component proportionate relationship and the target vehicle
The color component of color, obtains the color component of the real vehicles color of the target vehicle, determines the target vehicle
Real vehicles color.
6. device according to claim 5, which is characterized in that the determining module, comprising:
First searches submodule, for searching the colour atla library pre-saved, and in the colour atla according to the shooting number plate color
The target colour atla including the shooting number plate color is found in library;
Second searches submodule, for finding the shooting vehicle color of the target vehicle on the target colour atla;
Third searches submodule, for the colour atla position according to the shooting vehicle color in the target colour atla, searches mark
Quasi- colour atla determines the real vehicles color of the target vehicle;The target colour atla and standard color card are on same colour atla position
It is shooting color and corresponding true colors respectively.
7. according to any device of claim 5~6, which is characterized in that described device further include:
Brightness calculation module after the real vehicles color for determining the target vehicle, obtains the real vehicles color
True brightness value, and obtain the shooting brightness value of the shooting vehicle color of the target vehicle;
Calculations of offset module, for determining intensity deviation value according to the true brightness value and the shooting brightness value;
Image conversion module, for the vehicle pictures to be converted to YUV image;
Intensity deviation module, for carrying out intensity deviation in the Y-component of the YUV image according to the intensity deviation value;
Described image conversion module, is also used to: the YUV image after intensity deviation is converted to RGB image;
The determining module, is also used to: determining the vehicle color for the target vehicle for including in the RGB image for after brightness correction
The target vehicle real vehicles color.
8. device according to claim 5, which is characterized in that described device further include:
Gray value determining module, for determining the target vehicle when determining the target vehicle picture is gray level image
Shoot the gray value of vehicle color;
The determining module, is also used to: according to the corresponding relationship of preset intensity value ranges and color, by the shooting vehicle face
Color corresponding to intensity value ranges locating for the gray value of color is determined as the real vehicles color of the target vehicle.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610584806.9A CN106250827B (en) | 2016-07-20 | 2016-07-20 | Vehicle color identification method and device |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610584806.9A CN106250827B (en) | 2016-07-20 | 2016-07-20 | Vehicle color identification method and device |
Publications (2)
Publication Number | Publication Date |
---|---|
CN106250827A CN106250827A (en) | 2016-12-21 |
CN106250827B true CN106250827B (en) | 2019-06-11 |
Family
ID=57604554
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201610584806.9A Active CN106250827B (en) | 2016-07-20 | 2016-07-20 | Vehicle color identification method and device |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN106250827B (en) |
Families Citing this family (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
AU2017399473B2 (en) | 2017-02-15 | 2021-03-11 | Beijing Didi Infinity Technology And Development Co., Ltd. | Systems and methods for on-demand service |
CN108537714A (en) * | 2017-02-24 | 2018-09-14 | 北京嘀嘀无限科技发展有限公司 | Assist in identifying the method and server of vehicle |
CN115002279A (en) | 2017-02-15 | 2022-09-02 | 北京嘀嘀无限科技发展有限公司 | Method and system for providing information on terminal equipment |
CN110569692B (en) * | 2018-08-16 | 2023-05-12 | 创新先进技术有限公司 | Multi-vehicle identification method, device and equipment |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN1972457A (en) * | 2005-11-24 | 2007-05-30 | 株式会社日立制作所 | Video processing apparatus and portable mobile terminal |
CN101635091A (en) * | 2008-07-23 | 2010-01-27 | 上海弘视通信技术有限公司 | Device for detecting vehicle and identifying color |
CN102231206A (en) * | 2011-07-14 | 2011-11-02 | 浙江理工大学 | Colorized night vision image brightness enhancement method applicable to automotive assisted driving system |
CN103425989A (en) * | 2013-08-07 | 2013-12-04 | 中山大学 | Vehicle color identification method based on significance analysis |
CN104978746A (en) * | 2015-06-26 | 2015-10-14 | 西安理工大学 | Running vehicle body color identification method |
-
2016
- 2016-07-20 CN CN201610584806.9A patent/CN106250827B/en active Active
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN1972457A (en) * | 2005-11-24 | 2007-05-30 | 株式会社日立制作所 | Video processing apparatus and portable mobile terminal |
CN101635091A (en) * | 2008-07-23 | 2010-01-27 | 上海弘视通信技术有限公司 | Device for detecting vehicle and identifying color |
CN102231206A (en) * | 2011-07-14 | 2011-11-02 | 浙江理工大学 | Colorized night vision image brightness enhancement method applicable to automotive assisted driving system |
CN103425989A (en) * | 2013-08-07 | 2013-12-04 | 中山大学 | Vehicle color identification method based on significance analysis |
CN104978746A (en) * | 2015-06-26 | 2015-10-14 | 西安理工大学 | Running vehicle body color identification method |
Also Published As
Publication number | Publication date |
---|---|
CN106250827A (en) | 2016-12-21 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
KR102266830B1 (en) | Lane determination method, device and storage medium | |
CN106250827B (en) | Vehicle color identification method and device | |
US8886634B2 (en) | Apparatus for displaying result of analogous image retrieval and method for displaying result of analogous image retrieval | |
CN108091141B (en) | License plate recognition system | |
Liu et al. | “Seeing is not always believing”: detecting perception error attacks against autonomous vehicles | |
US8831284B2 (en) | Object identification from image data captured from a mobile aerial platforms | |
CN105321350A (en) | Method and device for detection of fake plate vehicles | |
US20190156140A1 (en) | Method and system for recognizing license plate | |
CN109284801B (en) | Traffic indicator lamp state identification method and device, electronic equipment and storage medium | |
CN105608417A (en) | Traffic signal lamp detection method and device | |
CN111369801B (en) | Vehicle identification method, device, equipment and storage medium | |
CN107003671A (en) | Positioning and mapping method and system | |
CN111382735B (en) | Night vehicle detection method, device, equipment and storage medium | |
CN105894819A (en) | Fake-license-plate vehicle identification method based on twice verification | |
CN112861673A (en) | False alarm removal early warning method and system for multi-target detection of surveillance video | |
CN115761655A (en) | Target tracking method and device | |
CN109727268A (en) | Method for tracking target, device, computer equipment and storage medium | |
JPWO2020194570A1 (en) | Sign positioning system and program | |
CN111639640B (en) | License plate recognition method, device and equipment based on artificial intelligence | |
Tang | Development of a multiple-camera tracking system for accurate traffic performance measurements at intersections | |
CN111402185A (en) | Image detection method and device | |
WO2023275544A1 (en) | Methods and systems for detecting vessels | |
CN111601041B (en) | Camera scene mode switching method and device and intelligent equipment | |
Chen et al. | Multi-camera vehicle identification in tunnel surveillance system | |
CN111860050B (en) | Loop detection method and device based on image frames and vehicle-mounted terminal |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant |