CN109447174A - A kind of lacquer painting recognition methods, device, storage medium and electronic equipment - Google Patents
A kind of lacquer painting recognition methods, device, storage medium and electronic equipment Download PDFInfo
- Publication number
- CN109447174A CN109447174A CN201811320189.7A CN201811320189A CN109447174A CN 109447174 A CN109447174 A CN 109447174A CN 201811320189 A CN201811320189 A CN 201811320189A CN 109447174 A CN109447174 A CN 109447174A
- Authority
- CN
- China
- Prior art keywords
- lacquer painting
- type
- lacquer
- characteristic
- 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.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/74—Image or video pattern matching; Proximity measures in feature spaces
- G06V10/75—Organisation of the matching processes, e.g. simultaneous or sequential comparisons of image or video features; Coarse-fine approaches, e.g. multi-scale approaches; using context analysis; Selection of dictionaries
- G06V10/757—Matching configurations of points or features
Landscapes
- Engineering & Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Theoretical Computer Science (AREA)
- Evolutionary Computation (AREA)
- Computing Systems (AREA)
- Databases & Information Systems (AREA)
- Artificial Intelligence (AREA)
- General Health & Medical Sciences (AREA)
- Medical Informatics (AREA)
- Software Systems (AREA)
- Health & Medical Sciences (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Multimedia (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The present invention provides a kind of lacquer painting recognition methods, device, storage medium and electronic equipments, wherein, this method comprises: obtaining the lacquer painting data of object to be measured, and lacquer painting characteristic is determined from lacquer painting data, lacquer painting characteristic is used to indicate the lacquer painting type of object to be measured;Lacquer painting type corresponding with lacquer painting characteristic is determined according to trained lacquer painting disaggregated model in advance, wherein lacquer painting disaggregated model is used to indicate the corresponding relationship between lacquer painting characteristic and lacquer painting type;Export the lacquer painting type of object to be measured.Lacquer painting recognition methods, device, storage medium and the electronic equipment provided through the embodiment of the present invention can more accurately judge lacquer painting type based on lacquer painting characteristic;It can be convenient based on trained lacquer painting disaggregated model and quickly determine corresponding lacquer painting type, the recognition efficiency of lacquer painting identification can be improved.
Description
Technical field
The present invention relates to lacquer painting identification technology fields, are situated between in particular to a kind of lacquer painting recognition methods, device, storage
Matter and electronic equipment.
Background technique
Used car has its particularity as nonstandard commodity, and every money vehicle vehicle condition is different, and price is also not quite similar.In assessment two
When handcart price, check that vehicle appearance lacquer painting is a kind of evaluation measures on basis.
It is directly detected currently, can generally directly adopt paint film instrument in each sample point, measures the average thickness of lacquer painting;It is existing
Some bluetooth paint film instrument can only judge the thickness of lacquer painting, can not automatically identify metal plate, spray painting or normal.Utilizing paint
Film instrument is measured after lacquer painting thickness, it is also necessary to lacquer painting type is judged by appraiser, this evaluation process is mainly by assessing
For the experience of teacher come what is determined, the experience ability difference of appraiser causes result that can be not quite similar, and can expend a large amount of manpower.
Summary of the invention
To solve the above problems, the embodiment of the present invention be designed to provide a kind of lacquer painting recognition methods, device, storage are situated between
Matter and electronic equipment.
In a first aspect, the embodiment of the invention provides a kind of lacquer painting recognition methods, comprising:
The lacquer painting data of object to be measured are obtained, and determine lacquer painting characteristic from the lacquer painting data, the lacquer painting is special
Sign data are used to indicate the lacquer painting type of the object to be measured;
Lacquer painting type corresponding with the lacquer painting characteristic is determined according to trained lacquer painting disaggregated model in advance,
In, the lacquer painting disaggregated model is used to indicate the corresponding relationship between lacquer painting characteristic and lacquer painting type;
Export the lacquer painting type of the object to be measured.
In one possible implementation, the lacquer painting disaggregated model includes the first lacquer painting classification submodel and the second paint
Face classification submodel;
Lacquer painting type packet corresponding with the lacquer painting characteristic is determined according to trained lacquer painting disaggregated model in advance
It includes:
Whether lacquer painting type corresponding with the lacquer painting characteristic is judged using first lacquer painting classification submodel
For the first lacquer painting type;
Lacquer painting type corresponding with the lacquer painting characteristic is being judged using first lacquer painting classification submodel
When being the first lacquer painting type, determine that the lacquer painting type of the object to be measured is the first lacquer painting type;
Lacquer painting type corresponding with the lacquer painting characteristic is being judged using first lacquer painting classification submodel
When not being the first lacquer painting type, judged using second lacquer painting classification submodel corresponding with the lacquer painting characteristic
Lacquer painting type whether be the second lacquer painting type;
Judging that lacquer painting type corresponding with the lacquer painting characteristic is using second lacquer painting classification submodel
When the second lacquer painting type, determine the object to be measured lacquer painting type be the second lacquer painting type, otherwise determine described in
The lacquer painting type of object to be measured is third lacquer painting type.
In one possible implementation, the first lacquer painting type, the second lacquer painting type and the third
Lacquer painting type is respectively one of type in lacquer painting type set, wherein the lacquer painting type set include metal plate type,
Normal type, spray painting type;The first lacquer painting type, the second lacquer painting type and the third lacquer painting type not phase
Together.
In one possible implementation, described special using first lacquer painting classification submodel judgement and the lacquer painting
Levy whether the corresponding lacquer painting type of data is that the first lacquer painting type includes:
It is input to the first lacquer painting classification submodel using the lacquer painting characteristic as input parameter, it is defeated to obtain first
Parameter out;
Judge whether first output parameter is greater than first threshold, wherein be greater than in first output parameter described
When first threshold, determine that corresponding with lacquer painting characteristic lacquer painting type is the first lacquer painting type, described the
When one output parameter is less than or equal to the first threshold, lacquer painting type corresponding with the lacquer painting characteristic is determined not
It is the first lacquer painting type.
In one possible implementation, described special using second lacquer painting classification submodel judgement and the lacquer painting
Levy whether the corresponding lacquer painting type of data is that the second lacquer painting type includes:
It is input to the second lacquer painting classification submodel using the lacquer painting characteristic as input parameter, it is defeated to obtain second
Parameter out;
Judge whether second output parameter is greater than second threshold, wherein be greater than in second output parameter described
When second threshold, determine that corresponding with lacquer painting characteristic lacquer painting type is the second lacquer painting type, described the
When two output parameters are less than or equal to the second threshold, determine that lacquer painting type corresponding with the lacquer painting characteristic is
The third lacquer painting type.
In one possible implementation, before the lacquer painting data for obtaining object to be measured, the method is also wrapped
It includes:
The lacquer painting type of the lacquer painting sample data and the sample object with corresponding relationship sample object is obtained, and from institute
It states and determines lacquer painting sample characteristics data in lacquer painting sample data;
According to the corresponding relationship between the lacquer painting sample characteristics data and the lacquer painting type of the sample object to initial
Disaggregated model carries out classification based training, obtains the lacquer painting disaggregated model, wherein described initial when carrying out the classification based training
The input parameter of disaggregated model is the lacquer painting sample characteristics data, and the output parameter of the preliminary classification model is used to indicate institute
State the lacquer painting type of sample object.
Second aspect, the embodiment of the invention also provides a kind of lacquer painting identification devices, comprising:
Preprocessing module for obtaining the lacquer painting data of object to be measured, and determines lacquer painting feature from the lacquer painting data
Data, the lacquer painting characteristic are used to indicate the lacquer painting type of the object to be measured;
Categorization module, for corresponding with the lacquer painting characteristic according to trained lacquer painting disaggregated model determination in advance
Lacquer painting type, wherein the lacquer painting disaggregated model is used to indicate the corresponding relationship between lacquer painting characteristic and lacquer painting type;
Output module, for exporting the lacquer painting type of the object to be measured.
The third aspect, the embodiment of the invention also provides a kind of storage medium, the storage medium is stored with computer can
It executes instruction, the computer executable instructions are for executing lacquer painting recognition methods described in above-mentioned any one.
Fourth aspect, the embodiment of the invention also provides a kind of electronic equipment, comprising: at least one processor;And with
The memory of at least one processor communication connection;Wherein,
The memory is stored with the instruction that can be executed by least one described processor, and described instruction is by described at least one
A processor executes, so that at least one described processor is able to carry out lacquer painting recognition methods described in above-mentioned any one.
In the scheme that the above-mentioned first aspect of the embodiment of the present invention provides, located in advance first after getting lacquer painting data
Reason extracts lacquer painting characteristic, can more accurately judge lacquer painting type based on lacquer painting characteristic;Based on trained paint
Face disaggregated model, which can be convenient, quickly determines corresponding lacquer painting type, and the recognition efficiency of lacquer painting identification can be improved.
To enable the above objects, features and advantages of the present invention to be clearer and more comprehensible, preferred embodiment is cited below particularly, and cooperate
Appended attached drawing, is described in detail below.
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
Some embodiments of invention for those of ordinary skill in the art without creative efforts, can be with
It obtains other drawings based on these drawings.
Fig. 1 shows a kind of flow chart of lacquer painting recognition methods provided by the embodiment of the present invention;
Fig. 2 shows in lacquer painting recognition methods provided by the embodiment of the present invention, the specific method of lacquer painting type is determined
Flow chart;
Fig. 3 shows the flow chart of another kind lacquer painting recognition methods provided by the embodiment of the present invention;
Fig. 4 shows a kind of structural schematic diagram of lacquer painting identification device provided by the embodiment of the present invention;
Fig. 5 is shown in lacquer painting identification device provided by the embodiment of the present invention, the concrete structure schematic diagram of categorization module;
Fig. 6 shows the structural schematic diagram of another kind lacquer painting identification device provided by the embodiment of the present invention;
Fig. 7 shows the structural schematic diagram that the electronic equipment of lacquer painting recognition methods is executed provided by the embodiment of the present invention.
Specific embodiment
In the description of the present invention, it is to be understood that, term " center ", " longitudinal direction ", " transverse direction ", " length ", " width ",
" thickness ", "upper", "lower", "front", "rear", "left", "right", "vertical", "horizontal", "top", "bottom" "inner", "outside", " up time
The orientation or positional relationship of the instructions such as needle ", " counterclockwise " is to be based on the orientation or positional relationship shown in the drawings, and is merely for convenience of
The description present invention and simplified description, rather than the device or element of indication or suggestion meaning must have a particular orientation, with spy
Fixed orientation construction and operation, therefore be not considered as limiting the invention.
In addition, term " first ", " second " are used for descriptive purposes only and cannot be understood as indicating or suggesting relative importance
Or implicitly indicate the quantity of indicated technical characteristic.Define " first " as a result, the feature of " second " can be expressed or
Implicitly include one or more of the features.In the description of the present invention, the meaning of " plurality " is two or more,
Unless otherwise specifically defined.
In the present invention unless specifically defined or limited otherwise, term " installation ", " connected ", " connection ", " fixation " etc.
Term shall be understood in a broad sense, for example, it may be being fixedly connected, may be a detachable connection, or be integrally connected;It can be machine
Tool connection, is also possible to be electrically connected;It can be directly connected, two members can also be can be indirectly connected through an intermediary
Connection inside part.For the ordinary skill in the art, above-mentioned term can be understood in this hair as the case may be
Concrete meaning in bright.
A kind of lacquer painting recognition methods provided in an embodiment of the present invention, it is shown in Figure 1, including step 101-103:
Step 101: obtaining the lacquer painting data of object to be measured, and determine lacquer painting characteristic, the lacquer painting from lacquer painting data
Characteristic is used to indicate the lacquer painting type of object to be measured.
In the embodiment of the present invention, object to be measured be need to detect the article of lacquer painting, such as new, used car, vehicle one
Partially (cabin cover, the right-rear-door of such as vehicle) gets object to be measured ready by paint film instrument, can acquire the paint of object to be measured
Face data.Wherein, which is specially the thickness data of lacquer painting, and the lacquer painting data include the multiple location points of object to be measured
Data.The lacquer painting data are pre-processed after getting lacquer painting data, count the lacquer painting characteristic of object to be measured,
The lacquer painting characteristic can serve to indicate that the lacquer painting type of object to be measured, lacquer painting characteristic can specifically include data maximum
It is value, data mean value, data variance, data minimum value, one or more in the difference of data maximums and data minimum value.For
For vehicle lacquer painting, the thickness of normal lacquer painting is minimum, followed by spray painting lacquer painting, and the thickness of metal plate lacquer painting is maximum.In lacquer painting feature
In data, the maximum value of lacquer painting data is bigger, indicates that the lacquer painting of object to be measured more may be metal plate type;Due under normal circumstances
Only a part of vehicle may require that reparation, such as spray painting reparation, metal plate reparation etc., and the position not being repaired is normal paint
The lacquer painting thickness in face, this position is minimum, and data minimum value can generally embody the normal lacquer painting thickness of the object to be measured;Data are maximum
The difference of value and data minimum value can indicate painting thickness or metal plate thickness;Data variance can generally indicate to repair the more of point
Few, under normal circumstances, variance is bigger, it is more to repair point, or be more likely to be metal plate reparation.
Step 102: lacquer painting class corresponding with lacquer painting characteristic is determined according to trained lacquer painting disaggregated model in advance
Type, wherein lacquer painting disaggregated model is used to indicate the corresponding relationship between lacquer painting characteristic and lacquer painting type.
In the embodiment of the present invention, previously according to relevant sample data training lacquer painting disaggregated model, lacquer painting classification mould is determined
The parameter of type determines the lacquer painting type of object to be measured by trained lacquer painting disaggregated model.Wherein, the lacquer painting lacquer painting type
It can be normal type, spray painting type, metal plate type.
Wherein, it needs first to train lacquer painting disaggregated model according to sample data before the lacquer painting data for obtaining object to be measured,
The training process specifically includes step A1-A2:
Step A1: obtaining the lacquer painting sample data of sample object and the lacquer painting type of sample object, and from lacquer painting sample number
According to middle determining lacquer painting sample characteristics data.
Step A2: according to the corresponding relationship between lacquer painting sample characteristics data and the lacquer painting type of sample object to initial point
Class model carries out classification based training, obtains lacquer painting disaggregated model, wherein when carrying out classification based training, the input of preliminary classification model
Parameter is lacquer painting sample characteristics data, and the output parameter of preliminary classification model is used to indicate the lacquer painting type of sample object.
In the embodiment of the present invention, sample object is similar with object to be measured, or new car, used car etc., only sample
The lacquer painting type of object is known, and the lacquer painting sample data of sample object and lacquer painting type are one-to-one.The lacquer painting
The lacquer painting type fruit of sample data specifically can be what experienced appraiser assessed according to lacquer painting sample data, alternatively,
It measures corresponding lacquer painting sample data again after the lacquer painting type for learning certain an object (such as used car), its other party can also be used
Formula obtains lacquer painting sample data, as long as can guarantee the authenticity of data, accuracy.After determining lacquer painting sample data,
It is similar with above-mentioned steps 101, corresponding lacquer painting sample characteristics data can be obtained by same preprocessing process, later with
The lacquer painting sample characteristics data are as input, lacquer painting type corresponding with lacquer painting sample data as output training lacquer painting classification
Model, and then can determine the parameter of lacquer painting disaggregated model.XGboost model, random specifically can be used in the lacquer painting disaggregated model
Forest model, GBDT (Gradient Boosting Decision Tree) model etc..
Step 103: exporting the lacquer painting type of object to be measured.
After determining lacquer painting type according to lacquer painting disaggregated model, which can be used as the lacquer painting class of object to be measured
Type exports the lacquer painting type later and (for example shows the lacquer painting type in local terminal, the lacquer painting type is sent to other equipment
Deng) which kind of is particularly belonged to the lacquer painting for informing user's object to be measured.
A kind of lacquer painting recognition methods provided in an embodiment of the present invention, extracts lacquer painting feature first after getting lacquer painting data
Data can more accurately judge lacquer painting type based on lacquer painting characteristic;It can be with based on trained lacquer painting disaggregated model
It quickly and easily determines corresponding lacquer painting type, the recognition efficiency of lacquer painting identification can be improved.
On the basis of the above embodiments, lacquer painting disaggregated model includes the first lacquer painting classification submodel and the classification of the second lacquer painting
Submodel.Specifically, " according to trained lacquer painting disaggregated model determination in advance and lacquer painting characteristic phase in above-mentioned steps 102
Corresponding lacquer painting type " includes step 1021-1025:
Step 1021: whether judging lacquer painting type corresponding with lacquer painting characteristic using the first lacquer painting classification submodel
For the first lacquer painting type;When lacquer painting type corresponding with lacquer painting characteristic is the first lacquer painting type, continue step 1022,
Otherwise continue step 1023.
Step 1022: the lacquer painting type for determining object to be measured is the first lacquer painting type.
Step 1023: whether judging lacquer painting type corresponding with lacquer painting characteristic using the second lacquer painting classification submodel
For the second lacquer painting type;When lacquer painting type corresponding with lacquer painting characteristic is the second lacquer painting type, continue step 1024,
Otherwise continue step 1025.
Step 1024: the lacquer painting type for determining object to be measured is the second lacquer painting type.
Step 1025: the lacquer painting type for determining object to be measured is third lacquer painting type.
In the embodiment of the present invention, the first lacquer painting classification submodel and the second lacquer painting classification submodel can be dividing for same type
Class model, only model parameter is different, for example be all made of XGboost model etc..The output result of first lacquer painting classification submodel
Lacquer painting for object to be measured, which belongs to the first lacquer painting type or is not belonging to the first lacquer painting type, (belongs to the second lacquer painting type or
Three lacquer painting types), the output result of the second lacquer painting classification submodel belongs to the second lacquer painting type or not for the lacquer painting of object to be measured
Belong to the second lacquer painting type (belonging to the first lacquer painting type or third lacquer painting type).
Wherein, the first lacquer painting type, the second lacquer painting type and third lacquer painting type are respectively in lacquer painting type set
One of type, wherein lacquer painting type set includes metal plate type, normal type, spray painting type.
In the embodiment of the present invention, which is a critical types, which refers in all paints
The type of first cis-position or the type of last cis-position in noodles type.Specifically, lacquer painting type include normal type, spray painting type,
Metal plate type is arranged according to the sequence of lacquer painting thickness from small to large, is put in order as normal type → spray painting type → metal plate
Type, then the type of the first cis-position is normal type, and the type of last cis-position is metal plate type;If according to lacquer painting thickness from greatly to
Small sequence arrangement, then the type of the first cis-position and last cis-position is metal plate type and normal type respectively.That is, the present invention is implemented
In example, the first lacquer painting type is metal plate type or normal type.Meanwhile the first lacquer painting type, the second lacquer painting type and third
Lacquer painting type is different.For example, the first lacquer painting type is metal plate type, the second lacquer painting type is normal type, third lacquer painting
Type is spray painting type.
Since lacquer painting disaggregated model can be divided into three classes according to the size of lacquer painting thickness, may determine that by the first subseries
Whether object to be measured belongs to critical types, passes through the lacquer painting that the second subseries determines object to be measured again when being not belonging to critical types
Type.The present embodiment is that the classification submodel of binary tree form judges the lacquer painting type of object to be measured using two output results,
Compared to directly showing that classification results have higher accuracy using a lacquer painting disaggregated model, classification results are more accurate.
Optionally, judge whether object to be measured is certain according to the output parameter of lacquer painting disaggregated model (such as predicted value etc.)
A kind of lacquer painting type.Specifically, above-mentioned steps 1021 " are judged opposite with lacquer painting characteristic using the first lacquer painting classification submodel
Whether the lacquer painting type answered is the first lacquer painting type " include:
Step B1: it is input to the first lacquer painting classification submodel using lacquer painting characteristic as input parameter, it is defeated to obtain first
Parameter out.
Step B2: judge whether the first output parameter is greater than first threshold, wherein be greater than the first threshold in the first output parameter
When value, determine that corresponding with lacquer painting characteristic lacquer painting type is the first lacquer painting type, be less than in the first output parameter or
When equal to first threshold, determine that lacquer painting type corresponding with lacquer painting characteristic is not the first lacquer painting type.
In the embodiment of the present invention, the first output parameter is the output parameter of the first lacquer painting classification submodel, first output
Parameter is specifically as follows the predicted value (value range is 0~1) of model, when first output parameter is greater than preset first threshold
When, illustrate that lacquer painting type corresponding with lacquer painting characteristic is the first lacquer painting type.For example, first threshold is 0.5, if first
The classify predicted value of submodel output of lacquer painting is greater than 0.5, then illustrates that the object to be measured belongs to the first lacquer painting type (such as metal plate class
Type), it otherwise needs further to be judged based on the second lacquer painting classification submodel.
Likewise, above-mentioned " judge lacquer painting type corresponding with lacquer painting characteristic using the second lacquer painting classification submodel
Whether it is the second lacquer painting type " include:
Step C1: it is input to the second lacquer painting classification submodel using lacquer painting characteristic as input parameter, it is defeated to obtain second
Parameter out.
Step C2: judge whether the second output parameter is greater than second threshold, wherein be greater than the second threshold in the second output parameter
When value, determine that corresponding with lacquer painting characteristic lacquer painting type is the second lacquer painting type, be less than in the second output parameter or
When equal to second threshold, determine that lacquer painting type corresponding with lacquer painting characteristic is third lacquer painting type.
Wherein, above-mentioned steps C1-C2 is similar to the principle of step B1-B2, is not repeated herein.Specifically, for example, second
Threshold value is set as 0.6, if the second output parameter is greater than 0.6, object to be measured is the second lacquer painting type (such as spray painting type), otherwise
For third lacquer painting type (such as normal type).
On the basis of the above embodiments, lacquer painting data include the lacquer painting subdata of the multiple components of object to be measured.To lacquer painting
Subdata can determine the lacquer painting subcharacter data of each component after being pre-processed, which includes: current part
Mark, the variance of the maximum value of current part data, the mean value of current part data, current part data, all parts datas
One in the maximum value of minimum value, current part data in mean value and the difference of the minimum value in all parts data mean values or
It is multinomial.Likewise, lacquer painting type includes the lacquer painting type of each component of object to be measured.
In the embodiment of the present invention, since object to be measured is when needing to repair, does not need generally to repair all surfaces, that is, only have
Part lacquer painting was repaired, and what is do not repaired is normal type.It, can be more smart by the way that object to be measured is divided into multiple components
Really identify the lacquer painting type of each component.By taking object to be measured is used car as an example, the surface of used car can be divided into cabin
Multiple components such as lid, left front door, left front fender, left A column, right front door, roof are got ready on each component using paint film instrument more
A thickness data generates the lacquer painting subdata of each component, is one group later with each component and is located in advance to lacquer painting subdata
Reason, determines the lacquer painting subcharacter data of each component, the lacquer painting subcharacter data are similar with above-mentioned lacquer painting characteristic, also wrap
Maximum value, mean value, variance containing data etc., distinguishing the minimum value being in lacquer painting subcharacter data is all parts data mean values
In minimum value.Data minimum value in the application is used to characterize the lacquer painting thickness of normal type, and for same to be measured right
As, the mean value of some parts data is smaller, illustrate that a possibility that component was repaired is lower, the mean value of the parts data with just
The lacquer painting thickness of normal type is closer, more can accurately indicate the lacquer painting thickness of normal type.Current part mark is for unique
It indicates the component, is specifically as follows the form of serial number, and also identified comprising corresponding component in lacquer painting subdata.
On the basis of the above embodiments, when lacquer painting data include the lacquer painting subdata of the multiple components of object to be measured, on
Step A2 is stated " according to the corresponding relationship between lacquer painting sample characteristics data and the lacquer painting type of sample object to preliminary classification model
Carry out classification based training " it specifically includes: using element attribute as first decision node of lacquer painting disaggregated model, and according to lacquer painting sample
Eigen data are trained initial lacquer painting disaggregated model.
In the embodiment of the present invention, when object to be measured is divided into multiple components, the lacquer painting of each component can be determined respectively
Type, at this time using element attribute as first decision node, which is used to distinguish which component the component to be, it
The lacquer painting type of the component is determined further according to the characteristic of the component (such as maximum value, mean value etc.) afterwards.For example, object to be measured
For used car, it is divided into the components such as cabin cover, left front door, left front fender, left A column, right front door, roof, then the lacquer painting is classified
Model splits data into cabin cover, left front door, left front fender etc. first, later recycle decision Tree algorithms (such as ID3 calculate
Method, C4.5 algorithm etc.) determine subsequent Attribute transposition.It, can be accurately by using element attribute as first decision node
The disaggregated model for being directed to different components is established, avoids occurring deviation in training process.
The lacquer painting recognition methods process is discussed in detail below by one embodiment.
Using used car as object to be measured in the embodiment of the present invention, the lacquer painting type of four components of used car is identified.Specifically
, which includes step 301-:
Step 301: pre-establishing trained first classification submodel and the second classification submodel.
Step 302: obtaining the lacquer painting data of object to be measured, which includes the lacquer painting subdata of each component.
In the present embodiment, lacquer painting subdata includes the mark and multiple groups lacquer painting thickness data of corresponding component, object to be measured with
Including illustrating for four components, four components are followed successively by cabin cover, right front fender, right front door, right A column, and four groups of acquisition
Lacquer painting subdata specifically:
'1000001':[185.8,136.1,167.5,226.5,212.5,211.8,137.6,169.0,140.6],
'1000002':[91.2,91.7,91.1,86.2,80.9,94.4,84.5,92.9,91.5],
'1000003':[101.5,86.9,98.9,99.3,92.6,92.8,97.2,97.4,86.2],
'1000004':[85.7,72.7,79.9,74.7,75.7,94.4].
Wherein, ' 1000001', ' 1000002' etc. indicates for component, and 185.8,136.1 etc. indicate lacquer painting thickness, first three groups
Every group includes 9 data (being got ready 9 times using paint film instrument), and the 4th group includes 6 data.
Step 303: being one group with each component and lacquer painting subdata is pre-processed, determine that lacquer painting of each component is special
Levy data.
Wherein, " minimum value in all parts data mean values " in four groups of lacquer painting subcharacter data is identical.
Step 304: using one group of lacquer painting subcharacter data as input, which being judged according to the first lacquer painting classification submodel
Whether subcharacter data match with the first lacquer painting type, continue step 305 when matching, and otherwise continue step 306.
Step 305: the determining lacquer painting type to match with lacquer painting subcharacter data is the first lacquer painting type.
Step 306: according to the second lacquer painting classify submodel judge lacquer painting subcharacter data whether with the second lacquer painting type phase
Matching continues step 307 when matching, and otherwise continues step 308.
Step 307: the determining lacquer painting type to match with lacquer painting subcharacter data is the second lacquer painting type.
Step 308: the determining lacquer painting type to match with lacquer painting subcharacter data is third lacquer painting type.
Step 309: determining the lacquer painting type of each lacquer painting subcharacter data, and export corresponding result.
In the embodiment of the present invention, can numerically lacquer painting type, for example 1 represent normal, and 2 represent spray painting, and 3 represent metal plate
Gold.A kind of form of final output result can be with are as follows:
'1000001':2,'1000002':1,'1000003':3,'1000004':1。
Lacquer painting characteristic is extracted in a kind of lacquer painting recognition methods provided in an embodiment of the present invention after getting lacquer painting data
According to can more accurately judge lacquer painting type based on lacquer painting characteristic;It can be square based on trained lacquer painting disaggregated model
Just corresponding lacquer painting type is quickly determined, the recognition efficiency of lacquer painting identification can be improved.It is two using two output results
The classification submodel for pitching tree-like formula judges the lacquer painting type of object to be measured, and classification results are more accurate;It include more in object to be measured
When a component, using the minimum value in all parts data mean values as characteristic, normal type can be more accurately indicated
Lacquer painting thickness;By can accurately establish the classification for different components using element attribute as first decision node
Model avoids occurring deviation in training process.
Describe lacquer painting recognition methods process in detail above, this method can also be realized by corresponding device, below in detail
Carefully introduce the structure and function of the device.
A kind of lacquer painting identification device provided in an embodiment of the present invention, it is shown in Figure 4, comprising:
Preprocessing module 41 determines that lacquer painting is special for obtaining the lacquer painting data of object to be measured, and from the lacquer painting data
Data are levied, the lacquer painting characteristic is used to indicate the lacquer painting type of the object to be measured;
Categorization module 42, for opposite with the lacquer painting characteristic according to trained lacquer painting disaggregated model determination in advance
The lacquer painting type answered, wherein the lacquer painting disaggregated model is used to indicate the corresponding pass between lacquer painting characteristic and lacquer painting type
System;
Output module 43, for exporting the lacquer painting type of the object to be measured.
On the basis of the above embodiments, the lacquer painting disaggregated model includes the first lacquer painting classification submodel and the second lacquer painting
Classification submodel;
Shown in Figure 5, the categorization module 42 includes:
First taxon 421, for utilizing first lacquer painting classification submodel judgement and the lacquer painting characteristic
Whether corresponding lacquer painting type is the first lacquer painting type;Judge and the paint using first lacquer painting classification submodel
When the corresponding lacquer painting type of region feature data is the first lacquer painting type, determine the lacquer painting type of the object to be measured for institute
State the first lacquer painting type;
Second taxon 422, for judging and the lacquer painting feature using first lacquer painting classification submodel
When the corresponding lacquer painting type of data is not the first lacquer painting type, second lacquer painting classification submodel judgement and institute are utilized
State whether the corresponding lacquer painting type of lacquer painting characteristic is the second lacquer painting type;Utilizing second lacquer painting classification submodel
When judging lacquer painting type corresponding with the lacquer painting characteristic as the second lacquer painting type, the object to be measured is determined
Lacquer painting type is the second lacquer painting type, otherwise determines that the lacquer painting type of the object to be measured is third lacquer painting type.
On the basis of the above embodiments, the first lacquer painting type, the second lacquer painting type and the third paint
Noodles type is respectively one of type in lacquer painting type set, wherein the lacquer painting type set includes metal plate type, just
Normal type, spray painting type;The first lacquer painting type is metal plate type or normal type, and the first lacquer painting type, described
Second lacquer painting type and the third lacquer painting type are different.
On the basis of the above embodiments, described that first taxon 421 is utilized to be used for:
It is input to the first lacquer painting classification submodel using the lacquer painting characteristic as input parameter, it is defeated to obtain first
Parameter out;
Judge whether first output parameter is greater than first threshold, wherein be greater than in first output parameter described
When first threshold, determine that corresponding with lacquer painting characteristic lacquer painting type is the first lacquer painting type, described the
When one output parameter is less than or equal to the first threshold, lacquer painting type corresponding with the lacquer painting characteristic is determined not
It is the first lacquer painting type.
On the basis of the above embodiments, second taxon 422 is used for:
It is input to the second lacquer painting classification submodel using the lacquer painting characteristic as input parameter, it is defeated to obtain second
Parameter out;
Judge whether second output parameter is greater than second threshold, wherein be greater than in second output parameter described
When second threshold, determine that corresponding with lacquer painting characteristic lacquer painting type is the second lacquer painting type, described the
When two output parameters are less than or equal to the second threshold, determine that lacquer painting type corresponding with the lacquer painting characteristic is
The third lacquer painting type.
On the basis of the above embodiments, shown in Figure 6, which further includes training module 44;
Before the preprocessing module 41 obtains the lacquer painting data of object to be measured, the training module 44 is for obtaining sample
The lacquer painting sample data of this object and the lacquer painting type of the sample object, and lacquer painting sample is determined from the lacquer painting sample data
Eigen data;According to the corresponding relationship between the lacquer painting sample characteristics data and the lacquer painting type of the sample object to first
Beginning disaggregated model carries out classification based training, obtains the lacquer painting disaggregated model, wherein described first when carrying out the classification based training
The input parameter of beginning disaggregated model is the lacquer painting sample characteristics data, and the output parameter of the preliminary classification model is used to indicate
The lacquer painting type of the sample object.
A kind of lacquer painting identification device provided in an embodiment of the present invention extracts lacquer painting characteristic after getting lacquer painting data
According to can more accurately judge lacquer painting type based on lacquer painting characteristic;It can be square based on trained lacquer painting disaggregated model
Just corresponding lacquer painting type is quickly determined, the recognition efficiency of lacquer painting identification can be improved.It is two using two output results
The classification submodel for pitching tree-like formula judges the lacquer painting type of object to be measured, and classification results are more accurate;It include more in object to be measured
When a component, using the minimum value in all parts data mean values as characteristic, normal type can be more accurately indicated
Lacquer painting thickness;By can accurately establish the classification for different components using element attribute as first decision node
Model avoids occurring deviation in training process.
The embodiment of the invention also provides a kind of storage medium, the storage medium is stored with computer executable instructions,
It includes the program for executing above-mentioned lacquer painting recognition methods, which, which can be performed above-mentioned any means, implements
Method in example.
Wherein, the storage medium can be any usable medium or data storage device that computer can access, packet
Include but be not limited to magnetic storage (such as floppy disk, hard disk, tape, magneto-optic disk (MO) etc.), optical memory (such as CD, DVD,
BD, HVD etc.) and semiconductor memory (such as it is ROM, EPROM, EEPROM, nonvolatile memory (NAND FLASH), solid
State hard disk (SSD)) etc..
Fig. 7 shows the structural block diagram of a kind of electronic equipment of another embodiment of the invention.The electronic equipment
1100 can be the host server for having computing capability, personal computer PC or portable portable computer or end
End etc..The specific embodiment of the invention does not limit the specific implementation of electronic equipment.
The electronic equipment 1100 includes at least one processor (processor) 1110, communication interface
(Communications Interface) 1120, memory (memory array) 1130 and bus 1140.Wherein, processor
1110, communication interface 1120 and memory 1130 complete mutual communication by bus 1140.
Communication interface 1120 with network element for communicating, and wherein network element includes such as Virtual Machine Manager center, shared storage.
Processor 1110 is for executing program.Processor 1110 may be a central processor CPU or dedicated collection
At circuit ASIC (Application Specific Integrated Circuit), or it is arranged to implement the present invention
One or more integrated circuits of embodiment.
Memory 1130 is for executable instruction.Memory 1130 may include high speed RAM memory, it is also possible to also wrap
Include nonvolatile memory (non-volatile memory), for example, at least a magnetic disk storage.Memory 1130 can also be with
It is memory array.Memory 1130 is also possible to by piecemeal, and described piece can be combined into virtual volume by certain rule.Storage
The instruction that device 1130 stores can be executed by processor 1110, so that processor 1110 is able to carry out in above-mentioned any means embodiment
Lacquer painting recognition methods.
The above description is merely a specific embodiment, but scope of protection of the present invention is not limited thereto, any
Those familiar with the art in the technical scope disclosed by the present invention, can easily think of the change or the replacement, and should all contain
Lid is within protection scope of the present invention.Therefore, the protection scope of the present invention shall be subject to the protection scope of the claims.
Claims (10)
1. a kind of lacquer painting recognition methods characterized by comprising
The lacquer painting data of object to be measured are obtained, and determine lacquer painting characteristic, the lacquer painting characteristic from the lacquer painting data
According to the lacquer painting type for being used to indicate the object to be measured;
Lacquer painting type corresponding with the lacquer painting characteristic is determined according to trained lacquer painting disaggregated model in advance, wherein
The lacquer painting disaggregated model is used to indicate the corresponding relationship between lacquer painting characteristic and lacquer painting type;
Export the lacquer painting type of the object to be measured.
2. the method according to claim 1, wherein the lacquer painting disaggregated model includes the first lacquer painting classification submodule
Type and the second lacquer painting classification submodel;
Determine that lacquer painting type corresponding with the lacquer painting characteristic includes: according to trained lacquer painting disaggregated model in advance
Judge whether corresponding with lacquer painting characteristic lacquer painting type is the using first lacquer painting classification submodel
One lacquer painting type;
Judging that lacquer painting type corresponding with the lacquer painting characteristic is institute using first lacquer painting classification submodel
When stating the first lacquer painting type, determine that the lacquer painting type of the object to be measured is the first lacquer painting type;
Judging that lacquer painting type corresponding with the lacquer painting characteristic is not using first lacquer painting classification submodel
When the first lacquer painting type, paint corresponding with the lacquer painting characteristic is judged using second lacquer painting classification submodel
Whether noodles type is the second lacquer painting type;
Judging that lacquer painting type corresponding with the lacquer painting characteristic is described using second lacquer painting classification submodel
When the second lacquer painting type, determines that the lacquer painting type of the object to be measured is the second lacquer painting type, otherwise determine described to be measured
The lacquer painting type of object is third lacquer painting type.
3. according to the method described in claim 2, it is characterized in that, the first lacquer painting type, the second lacquer painting type with
And the third lacquer painting type is respectively one of type in lacquer painting type set, wherein the lacquer painting type set packet
Include metal plate type, normal type, spray painting type, the first lacquer painting type, the second lacquer painting type and the third paint
Noodles type is different.
4. according to the method described in claim 2, it is characterized in that, it is described using first lacquer painting classification submodel judgement with
Whether the corresponding lacquer painting type of the lacquer painting characteristic is that the first lacquer painting type includes:
It is input to the first lacquer painting classification submodel using the lacquer painting characteristic as input parameter, obtains the first output ginseng
Number;
Judge whether first output parameter is greater than first threshold, wherein be greater than described first in first output parameter
When threshold value, determine that lacquer painting type corresponding with the lacquer painting characteristic is the first lacquer painting type, it is defeated described first
When parameter is less than or equal to the first threshold out, determine that lacquer painting type corresponding with the lacquer painting characteristic is not institute
State the first lacquer painting type.
5. according to the method described in claim 2, it is characterized in that, it is described using second lacquer painting classification submodel judgement with
Whether the corresponding lacquer painting type of the lacquer painting characteristic is that the second lacquer painting type includes:
It is input to the second lacquer painting classification submodel using the lacquer painting characteristic as input parameter, obtains the second output ginseng
Number;
Judge whether second output parameter is greater than second threshold, wherein be greater than described second in second output parameter
When threshold value, determine that lacquer painting type corresponding with the lacquer painting characteristic is the second lacquer painting type, it is defeated described second
When parameter is less than or equal to the second threshold out, determine that lacquer painting type corresponding with the lacquer painting characteristic is described
Third lacquer painting type.
6. method according to any one of claims 1 to 5, which is characterized in that in the lacquer painting number for obtaining object to be measured
According to before, the method also includes:
The lacquer painting sample data of sample object and the lacquer painting type of the sample object are obtained, and from the lacquer painting sample data
Determine lacquer painting sample characteristics data;
According to the corresponding relationship between the lacquer painting sample characteristics data and the lacquer painting type of the sample object to preliminary classification
Model carries out classification based training, obtains the lacquer painting disaggregated model, wherein when carrying out the classification based training, the preliminary classification
The input parameter of model is the lacquer painting sample characteristics data, and the output parameter of the preliminary classification model is used to indicate the sample
The lacquer painting type of this object.
7. a kind of lacquer painting identification device characterized by comprising
Preprocessing module for obtaining the lacquer painting data of object to be measured, and determines lacquer painting characteristic from the lacquer painting data,
The lacquer painting characteristic is used to indicate the lacquer painting type of the object to be measured;
Categorization module, for determining paint corresponding with the lacquer painting characteristic according to trained lacquer painting disaggregated model in advance
Noodles type, wherein the lacquer painting disaggregated model is used to indicate the corresponding relationship between lacquer painting characteristic and lacquer painting type;
Output module, for exporting the lacquer painting type of the object to be measured.
8. device according to claim 7, which is characterized in that the lacquer painting disaggregated model includes the first lacquer painting classification submodule
Type and the second lacquer painting classification submodel;
The categorization module includes:
First taxon, it is corresponding with the lacquer painting characteristic for being judged using first lacquer painting classification submodel
Whether lacquer painting type is the first lacquer painting type;Judge and the lacquer painting characteristic using first lacquer painting classification submodel
When according to corresponding lacquer painting type being the first lacquer painting type, determine that the lacquer painting type of the object to be measured is first paint
Noodles type;
Second taxon, for using first lacquer painting classification submodel judge it is opposite with the lacquer painting characteristic
It is special using second lacquer painting classification submodel judgement and the lacquer painting when lacquer painting type answered is not the first lacquer painting type
Levy whether the corresponding lacquer painting type of data is the second lacquer painting type;Utilizing second lacquer painting classification submodel judgement and institute
State the corresponding lacquer painting type of lacquer painting characteristic be the second lacquer painting type when, determine the lacquer painting type of the object to be measured
For the second lacquer painting type, otherwise determine that the lacquer painting type of the object to be measured is third lacquer painting type.
9. a kind of storage medium, which is characterized in that the storage medium is stored with computer executable instructions, and the computer can
It executes instruction and requires lacquer painting recognition methods described in 1-6 any one for perform claim.
10. a kind of electronic equipment characterized by comprising
At least one processor;And
The memory being connect at least one described processor communication;Wherein,
The memory is stored with the instruction that can be executed by least one described processor, and described instruction is by described at least one
It manages device to execute, so that at least one described processor is able to carry out lacquer painting recognition methods as claimed in any one of claims 1 to 6.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201811320189.7A CN109447174A (en) | 2018-11-07 | 2018-11-07 | A kind of lacquer painting recognition methods, device, storage medium and electronic equipment |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201811320189.7A CN109447174A (en) | 2018-11-07 | 2018-11-07 | A kind of lacquer painting recognition methods, device, storage medium and electronic equipment |
Publications (1)
Publication Number | Publication Date |
---|---|
CN109447174A true CN109447174A (en) | 2019-03-08 |
Family
ID=65550589
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201811320189.7A Pending CN109447174A (en) | 2018-11-07 | 2018-11-07 | A kind of lacquer painting recognition methods, device, storage medium and electronic equipment |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN109447174A (en) |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110108245A (en) * | 2019-05-22 | 2019-08-09 | 优信拍(北京)信息科技有限公司 | Detection method, device and the equipment of vehicle lacquer painting situation |
CN111027601A (en) * | 2019-11-25 | 2020-04-17 | 歌尔股份有限公司 | Plane detection method and device based on laser sensor |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102354422A (en) * | 2011-10-19 | 2012-02-15 | 湖南德顺电子科技有限公司 | Perimeter protection-oriented method for monitoring suspicious target |
CN103955699A (en) * | 2014-03-31 | 2014-07-30 | 北京邮电大学 | Method for detecting tumble event in real time based on surveillance videos |
CN106127747A (en) * | 2016-06-17 | 2016-11-16 | 史方 | Car surface damage classifying method and device based on degree of depth study |
CN106557939A (en) * | 2015-09-25 | 2017-04-05 | 优信拍(北京)信息科技有限公司 | A kind of motor vehicle detecting system and method |
-
2018
- 2018-11-07 CN CN201811320189.7A patent/CN109447174A/en active Pending
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102354422A (en) * | 2011-10-19 | 2012-02-15 | 湖南德顺电子科技有限公司 | Perimeter protection-oriented method for monitoring suspicious target |
CN103955699A (en) * | 2014-03-31 | 2014-07-30 | 北京邮电大学 | Method for detecting tumble event in real time based on surveillance videos |
CN106557939A (en) * | 2015-09-25 | 2017-04-05 | 优信拍(北京)信息科技有限公司 | A kind of motor vehicle detecting system and method |
CN106127747A (en) * | 2016-06-17 | 2016-11-16 | 史方 | Car surface damage classifying method and device based on degree of depth study |
Cited By (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110108245A (en) * | 2019-05-22 | 2019-08-09 | 优信拍(北京)信息科技有限公司 | Detection method, device and the equipment of vehicle lacquer painting situation |
CN110108245B (en) * | 2019-05-22 | 2021-04-13 | 优信拍(北京)信息科技有限公司 | Method, device and equipment for detecting vehicle paint surface condition |
CN111027601A (en) * | 2019-11-25 | 2020-04-17 | 歌尔股份有限公司 | Plane detection method and device based on laser sensor |
CN111027601B (en) * | 2019-11-25 | 2023-10-17 | 歌尔股份有限公司 | Plane detection method and device based on laser sensor |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN110348580B (en) | Method and device for constructing GBDT model, and prediction method and device | |
CN105426905B (en) | Robot barrier object recognition methods based on histogram of gradients and support vector machines | |
CN103258213B (en) | A kind of for the dynamic vehicle model recognizing method in intelligent transportation system | |
CN108648074A (en) | Loan valuation method, apparatus based on support vector machines and equipment | |
CN110111113B (en) | Abnormal transaction node detection method and device | |
CN107665353A (en) | Model recognizing method, device, equipment and computer-readable recording medium based on convolutional neural networks | |
CN106295541A (en) | Vehicle type recognition method and system | |
CN104680144A (en) | Lip language recognition method and device based on projection extreme learning machine | |
CN107274543B (en) | A kind of recognition methods of bank note, device, terminal device and computer storage medium | |
CN105103166A (en) | Systems and methods for texture assessment of a coating formulation | |
CN109325424A (en) | Discrimination method, device and the computer equipment of break in traffic rules and regulations case | |
CN107393118A (en) | A kind of recognition methods of bank note, device, terminal device and computer-readable storage medium | |
CN109447174A (en) | A kind of lacquer painting recognition methods, device, storage medium and electronic equipment | |
CN110346514A (en) | Mixed gas identification method, apparatus, computer equipment and storage medium | |
CN103679160A (en) | Human-face identifying method and device | |
CN110555339A (en) | target detection method, system, device and storage medium | |
CN109034086A (en) | Vehicle recognition methods, apparatus and system again | |
CN104517052A (en) | Invasion detection method and device | |
CN112613617A (en) | Uncertainty estimation method and device based on regression model | |
CN108629680A (en) | A kind of Risk Identification Method and system | |
CN113011961A (en) | Method, device and equipment for monitoring risk of company associated information and storage medium | |
CN107077617A (en) | fingerprint extraction method and device | |
CN103761538A (en) | Traffic sign recognition method based on shape feature invariant subspace | |
CN109285009B (en) | Bill brushing identification method and bill brushing identification device | |
CN116188445A (en) | Product surface defect detection and positioning method and device and terminal equipment |
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 | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20190308 |
|
RJ01 | Rejection of invention patent application after publication |