CN106096144A - A kind of automobile brand gene analysis method based on front face moulding - Google Patents
A kind of automobile brand gene analysis method based on front face moulding Download PDFInfo
- Publication number
- CN106096144A CN106096144A CN201610420677.XA CN201610420677A CN106096144A CN 106096144 A CN106096144 A CN 106096144A CN 201610420677 A CN201610420677 A CN 201610420677A CN 106096144 A CN106096144 A CN 106096144A
- Authority
- CN
- China
- Prior art keywords
- brand
- automobile
- front face
- gene
- automobile front
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/10—Geometric CAD
- G06F30/15—Vehicle, aircraft or watercraft design
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
Abstract
The invention belongs to vehicle configuration analysis field, it is provided that a kind of automobile brand gene analysis method based on front face moulding.This automobile brand gene analysis method includes creating automobile front face training storehouse, carrying out brand Classification and Identification and carry out brand gene analysis;Brand gene analysis farther includes in brand form gene alternate analysis and brand shape-designing trend analysis between form gene analysis, brand;This automobile brand gene analysis method can analyze the core form gene in automobile brand, it is possible to analyzes the similarity of form gene between brand, and also can be analyzed the differentiation of brand shape-designing and trend.
Description
Technical field
The invention belongs to vehicle configuration analysis field.The present invention relates to a kind of automobile brand gene based on front face moulding divide
Analysis method, is specifically related to a kind of method realizing brand gene analysis according to automobile brand sorting technique.
Background technology
Current Automobile Enterprises increasingly payes attention to brand building, and the familial design of the shape of automobile face has become automobile
One of core of brand building.By the sculpted zone that will include headlight, air-inlet grille and bumper as before automobile
Face, it appeared that the automobile front face of same brand different model has some similar shape-designings, referred to as brand moulding base
Cause.By the analysis to brand form gene, the design feature of face moulding before brand can be obtained, can be according to sales volume factorial analysis
Consumer favorite front face moulding.But the analysis of current brand form gene is mostly based in automobile front face the spy extracted
Levy line, and only single brand is analyzed, form gene can not be embodied, the most not in the way of constructing complete automobile front face
The similarity of the form gene between brand can be analyzed.
Summary of the invention
The technical problem to be solved in the present invention is to provide a kind of automobile brand gene analysis method based on front face moulding, should
Method can analyze the core form gene in automobile brand, it is possible to analyzes the similarity of form gene between brand, and also can
So that the differentiation of brand shape-designing and trend are analyzed.
Technical scheme:
A kind of automobile brand gene analysis method based on front face moulding, including creating automobile front face training storehouse, carrying out product
Board Classification and Identification and carry out brand gene analysis;
(1) automobile front face training storehouse is created: this automobile front face trains the front face picture that storehouse comprises multiple domestic automobile brand,
According to brand to automobile front face training storehouse picture classification, and set up brand label and time label;Add up each vehicle of each brand
Year sales volume, all years in the time sales volume sum calculating that each vehicle added up, the year sales volume obtaining each vehicle under same brand divides
Cloth, then quantifies the year sales volume distribution of each vehicle and as shared by vehicle picture number in each brand in automobile front face training storehouse
Proportion parameter, carries out the establishment in automobile front face training storehouse, completes the genes amplification of some brand and vehicle according to proportion parameter;Right
In every pictures, image content for only to comprise headlight, air-inlet grille, the sculpted zone of bumper, referred to as automobile front face figure
Sheet.All automobile front face pictures will be through background removal, and car plate removal processes, and every pictures to be adjusted to unified high wide
Ratio.
(2) brand classification is with identification: first, and all images in automobile front face training storehouse are carried out gray proces and by all
Image Adjusting is to Suitable pixel sizes scope 40~60 × 100~150 (high × wide).Then with PCANet character representation and SVM
The automobile front face training storehouse that (1) is obtained by the automobile brand sorting technique that grader combines carries out leaving-one method (leave-one-
Out) test, it may be assumed that every time extract an automobile front face picture as test sample, residual graph from automobile front face training storehouse during test
Sheet is as training sample;Training sample carried out after classification based training obtains sorter model by PCANet/SVM method, to test
Sample is identified;According to this process, successively all automobile front face pictures in training storehouse are tested.
(3) brand gene analysis.The test result obtaining (2) is analyzed, and analysing content is as follows:
Form gene analysis in (a) brand.Obtain the test result of certain brand in (2), before automobiles all in this brand
The decision content of face picture recognition result sorts in descending order, selects ten automobile front face picture samples corresponding to front ten values as product
The representative of board gene.Then " face before average vehicle " figure of this automobile brand is generated by " average face " technology calculated in photography
Picture." face before average vehicle " will be somebody's turn to do and be considered as the core gene of automobile brand.
Form gene alternate analysis between (b) brand.Obtain the test result of all brands in (2), record in certain brand and know
Not wrong picture sample and corresponding recognition result, record the picture sample being identified as this brand in other brands, so simultaneously
After these two groups of result visualizations are processed, set up the recognition result cross chart of certain brand and other brands, referred to as form gene
Cross chart.In like manner, the form gene cross chart of all brands and other brands can be set up.
(c) brand shape-designing trend analysis.Obtain the test result of certain brand in (2), according to time label, the most right
In certain time, the decision content of auto graph recognition result sorts in descending order, selects ten automobile front face pictures that front ten values are corresponding
Sample is as the representative of this time brand gene.Then this automobile brand is generated by " average face " technology in calculating photography
" face before average vehicle ", is considered as the automobile brand core gene by this time by this " face before average vehicle ".In like manner, it is thus achieved that should
" face before average vehicle " in brand all times, it can be observed that their variation tendency.Further, in conjunction with (b) module, root
Form gene cross chart dynamic evolution figure in time between automobile brand can be obtained according to time label, thus obtain certain automobile product
The development trend of face shape-designing before board.
Beneficial effects of the present invention:
(1) method that automobile brand core gene based on front face moulding is analyzed is provided.
(2) method of gene similarity analysis between automobile brand based on front face moulding is provided.
(3) method providing automobile brand gene trend analysis based on front face moulding.
Accompanying drawing explanation
Fig. 1 is the techniqueflow chart of patent of the present invention.
Fig. 2 be patent of the present invention brand between form gene cross chart.
In figure: be 1. the results area that certain brand recognition is correct, 2.-7. it is the knot that each self-identifying of other different brands is correct
Really region, 8. pointed circle represents 1. brand recognition and becomes other brand results, 9.-Pointed symbology other
Brand recognition becomes the result of 1. brand.
Detailed description of the invention
Below in conjunction with accompanying drawing and technical scheme, further illustrate the detailed description of the invention of the present invention.
Automobile brand identification based on automobile front face, it specifically comprises the following steps that
(1) automobile front face training storehouse is created: this automobile front face trains the front face picture that storehouse comprises multiple domestic automobile brand,
Automobile front face training storehouse picture is classified according to brand, and sets up brand label and time label;Each brand comprise multiple
Vehicle, each vehicle includes again the automobile of multiple different colours and style.Add up each vehicle of each brand from 2000 so far
Year sales volume, calculates year in each vehicle all times sales volume sum, obtains the year sales volume distribution of each vehicle under same brand, then quantifies
The year sales volume of each vehicle is distributed and as the proportion parameter shared by vehicle picture number in each brand in automobile front face training storehouse, presses
Carry out the establishment in automobile front face training storehouse according to proportion parameter, thus complete the genes amplification of some brand and vehicle.For every
Picture, image content for only to comprise headlight, air-inlet grille, the sculpted zone of bumper, referred to as automobile front face picture.All
Automobile front face picture will be through background removal, and car plate removal processes, and every pictures to be adjusted to unified depth-width ratio, as high wide
Ratio is 1:2.5.
(2) brand classification is with identification: first, and all images in automobile front face training storehouse are carried out gray proces and by all
Image Adjusting is to Suitable pixel sizes scope 40~60 × 100~150 (high × wide).Then with PCANet character representation and SVM
The automobile front face training storehouse that (1) is obtained by the automobile brand sorting technique that grader combines carries out leaving-one method (leave-one-
Out) test, it may be assumed that every time extract an automobile front face picture as test sample, residual graph from automobile front face training storehouse during test
Sheet is as training sample;Training sample carried out after classification based training obtains sorter model by PCANet/SVM method, to test
Sample is identified;According to this process, successively all automobile front face pictures in training storehouse are tested.
Further, this PCANet feature extracting method is made up of three basic data processing modules: principal component analysis
(PCA), binary system Hash (Binary Hashing) and blocked histogram (Block-wise Histogram).Wherein twice master
Component analysis constitutes the first two stage of PCANet feature extraction, is main characteristic extraction part;And binary system Hash and point
Block rectangular histogram is then the output stage of PCANet, the output result of twice principal component analysis is mapped as final output characteristic to
Amount.
Further, in PCANet character representation method, the optimized parameter scope of PCANet is: PCA moves in processing
The Pixel Dimensions of window is 3 × 3 or 5 × 5 or 7 × 7 (high × wide);Wave filter quantity is 8;Moving Window in the histogram treatment stage
The Pixel Dimensions scope of mouth is 7~10 × 7~10 (high × wide), and window coincidence factor scope is 0.5~0.7.
Further, the training process of SVM classifier is: PCANet is extracted the characteristic vector of training sample obtained and
Automobile front face training storehouse brand label information inputs to SVM classifier, SVM classifier train sorter model.
(3) brand gene analysis.The test result obtaining (2) is analyzed, and analysing content is as follows:
Form gene analysis in (a) brand.Obtain the test result of certain brand in (2), before automobiles all in this brand
The decision content of face picture recognition result sorts in descending order, selects ten automobile front face picture samples corresponding to front ten values as product
The representative of board gene.Then " face before average vehicle " figure of this automobile brand is generated by " average face " technology calculated in photography
Picture." face before average vehicle " will be somebody's turn to do and be considered as the core gene of automobile brand.
Form gene alternate analysis between (b) brand.Obtain the test result of all brands in (2), record in certain brand and know
Not wrong picture sample and corresponding recognition result, record the picture sample being identified as this brand in other brands, so simultaneously
After these two groups of result visualizations are processed, set up the recognition result cross chart of certain brand and other brands, referred to as form gene
Cross chart.In like manner, the form gene cross chart of all brands and other brands can be set up.
As shown in Figure 2,1.-represent the results area that each self-identifying of different automobile brands is correct, 1. brand the most respectively
Identification error sample 8. represented by circle, be distributed in 2.-7. in the recognition result 8. represented 1. in different brand;With
Reason, other brand recognitions become picture sample 1. by be distributed in 1. brand 9.-Represent.Thus, can set up 1. brand with
The form gene cross chart of other brands.
(c) brand shape-designing trend analysis.Obtain the test result of certain brand in (2), according to time label, the most right
In certain time, the decision content of auto graph recognition result sorts in descending order, selects ten automobile front face pictures that front ten values are corresponding
Sample is as the representative of this time brand gene.Then this automobile brand is generated by " average face " technology in calculating photography
" face before average vehicle ", is considered as the automobile brand core gene by this time by this " face before average vehicle ".In like manner, it is thus achieved that should
" face before average vehicle " in brand all times, it can be observed that their variation tendency.Further, integrating step (b) mould
Block, can obtain form gene cross chart dynamic evolution figure in time between automobile brand according to time label, thus obtain certain
The development trend of face shape-designing before automobile brand.
Above in association with accompanying drawing, the detailed description of the invention of the present invention is described, but these explanations can not be understood to limit
The scope of the present invention, protection scope of the present invention is limited by appended claims, any at the claims in the present invention base
Change on plinth is all protection scope of the present invention.
Claims (1)
1. an automobile brand gene analysis method based on front face moulding, it is characterised in that this automobile brand gene analysis side
Method includes creating automobile front face training storehouse, carrying out brand Classification and Identification and carry out brand gene analysis;
(1) automobile front face training storehouse is created: this automobile front face trains the front face picture that storehouse comprises multiple domestic automobile brand, according to
Brand is to automobile front face training storehouse picture classification, and sets up brand label and time label;Add up the year of each vehicle of each brand
Sales volume, all years in the time sales volume sum calculating that each vehicle added up, obtain the year sales volume distribution of each vehicle under same brand, so
The year sales volume of each vehicle of rear quantization is distributed and as the proportion shared by vehicle picture number in each brand in automobile front face training storehouse
Parameter, carries out the establishment in automobile front face training storehouse, completes the genes amplification of some brand and vehicle according to proportion parameter;For often
Pictures, image content for only to comprise headlight, air-inlet grille, the sculpted zone of bumper, referred to as automobile front face picture;Institute
Having the automobile front face picture will be through background removal, car plate removal processes, and every pictures to be adjusted to unified depth-width ratio;
(2) brand classification is with identification: first, and all images in automobile front face training storehouse are carried out gray proces and by all images
It is adjusted to Suitable pixel sizes scope height × a width of 40~60 × 100~150;Then with PCANet character representation and svm classifier
The automobile front face training storehouse that step (1) is obtained by the automobile brand sorting technique that device combines carries out leaving-one method test, it may be assumed that survey every time
From automobile front face training one automobile front face picture of storehouse extraction as test sample during examination, residue picture is as training sample;With
Training sample is carried out after classification based training obtains sorter model, being identified test sample by PCANet/SVM method;According to
All automobile front face pictures in training storehouse are tested by this process successively;
(3) brand gene analysis;The test result obtaining step (2) is analyzed, and analysing content is as follows: make in (a) brand
Type gene analysis;The test result of certain brand in obtaining step (2), to automobile front face picture recognition results all in this brand
Decision content sorts in descending order, selects the representative as brand gene of ten automobile front face picture samples corresponding to front ten values;So
" face before average vehicle " image of this automobile brand is generated afterwards by " average face " technology calculated in photography;Should " average vapour
Front face " it is considered as the core gene of automobile brand;
Form gene alternate analysis between (b) brand;The test result of all brands in obtaining step (2), records in certain brand and knows
Not wrong picture sample and corresponding recognition result, record the picture sample being identified as this brand in other brands, so simultaneously
After these two groups of result visualizations are processed, set up the recognition result cross chart of certain brand and other brands, referred to as form gene
Cross chart;In like manner, the form gene cross chart of all brands and other brands is set up;
(c) brand shape-designing trend analysis;In obtaining step (2), the test result of certain brand, according to time label, the most right
In certain time, the decision content of auto graph recognition result sorts in descending order, selects ten automobile front face pictures that front ten values are corresponding
Sample is as the representative of this time brand gene;Then this automobile brand is generated by " average face " technology in calculating photography
" face before average vehicle ", is considered as face shaped core gene before the automobile brand in this time by this " face before average vehicle ";With
Reason, it is thus achieved that " face before average vehicle " in this brand all times, observes their variation tendency;Further, integrating step (b)
Module, can obtain form gene cross chart dynamic evolution figure in time between automobile brand according to time label, thus obtain
The development trend of face shape-designing before certain automobile brand.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610420677.XA CN106096144B (en) | 2016-06-13 | 2016-06-13 | A kind of automobile brand genetic analysis method based on preceding face moulding |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610420677.XA CN106096144B (en) | 2016-06-13 | 2016-06-13 | A kind of automobile brand genetic analysis method based on preceding face moulding |
Publications (2)
Publication Number | Publication Date |
---|---|
CN106096144A true CN106096144A (en) | 2016-11-09 |
CN106096144B CN106096144B (en) | 2019-04-09 |
Family
ID=57846980
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201610420677.XA Active CN106096144B (en) | 2016-06-13 | 2016-06-13 | A kind of automobile brand genetic analysis method based on preceding face moulding |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN106096144B (en) |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN106874418A (en) * | 2017-01-24 | 2017-06-20 | 成都容豪电子信息科技有限公司 | A kind of automobile model data set method for building up for serving deep learning |
CN112258472A (en) * | 2020-10-20 | 2021-01-22 | 大连理工大学 | Automatic scoring method for automobile exterior shape |
Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20030086593A1 (en) * | 2001-05-31 | 2003-05-08 | Chengjun Liu | Feature based classification |
CN103295063A (en) * | 2013-05-17 | 2013-09-11 | 浙江大学 | Method for SUV (Sport Utility Vehicle) product family gene evolution based on genetic algorithm |
CN104156692A (en) * | 2014-07-07 | 2014-11-19 | 叶茂 | Automobile logo sample training and recognition method based on air-inlet grille positioning |
-
2016
- 2016-06-13 CN CN201610420677.XA patent/CN106096144B/en active Active
Patent Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20030086593A1 (en) * | 2001-05-31 | 2003-05-08 | Chengjun Liu | Feature based classification |
CN103295063A (en) * | 2013-05-17 | 2013-09-11 | 浙江大学 | Method for SUV (Sport Utility Vehicle) product family gene evolution based on genetic algorithm |
CN104156692A (en) * | 2014-07-07 | 2014-11-19 | 叶茂 | Automobile logo sample training and recognition method based on air-inlet grille positioning |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN106874418A (en) * | 2017-01-24 | 2017-06-20 | 成都容豪电子信息科技有限公司 | A kind of automobile model data set method for building up for serving deep learning |
CN112258472A (en) * | 2020-10-20 | 2021-01-22 | 大连理工大学 | Automatic scoring method for automobile exterior shape |
Also Published As
Publication number | Publication date |
---|---|
CN106096144B (en) | 2019-04-09 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN105046196B (en) | Front truck information of vehicles structuring output method based on concatenated convolutional neutral net | |
CN101329734B (en) | License plate character recognition method based on K-L transform and LS-SVM | |
CN104463195A (en) | Printing style digital recognition method based on template matching | |
CN101872416B (en) | Vehicle license plate recognition method and system of road image | |
CN112686812B (en) | Bank card inclination correction detection method and device, readable storage medium and terminal | |
CN111860499A (en) | Feature grouping-based bilinear convolutional neural network automobile brand identification method | |
CN111126224A (en) | Vehicle detection method and classification recognition model training method | |
CN111310850B (en) | License plate detection model construction method and system, license plate detection method and system | |
CN103903005A (en) | License plate image identification system and method | |
CN106610969A (en) | Multimodal information-based video content auditing system and method | |
CN107729812B (en) | Method suitable for recognizing vehicle color in monitoring scene | |
CN102663401B (en) | Image characteristic extracting and describing method | |
CN108197644A (en) | A kind of image-recognizing method and device | |
CN103366367A (en) | Pixel number clustering-based fuzzy C-average value gray level image splitting method | |
CN107704797B (en) | Real-time detection method, system and equipment based on pedestrians and vehicles in security video | |
Shah et al. | OCR-based chassis-number recognition using artificial neural networks | |
CN105975949A (en) | Visual-information-based automobile identification method | |
CN114444566A (en) | Image counterfeiting detection method and device and computer storage medium | |
JP2008251029A (en) | Character recognition device and license plate recognition system | |
CN115620075A (en) | Method, system and equipment for generating data set for leukocyte classification model | |
CN106096144A (en) | A kind of automobile brand gene analysis method based on front face moulding | |
Bolotova et al. | License plate recognition with hierarchical temporal memory model | |
CN111461002B (en) | Sample processing method for thermal imaging pedestrian detection | |
Brillantes et al. | Detection of fonts and characters with hybrid graphic-text plate numbers | |
CN108122000B (en) | Vehicle logo recognition method based on feature learning |
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 |