CN1363483A - Method and equipment for discriminating headlight of troublemaking car by computer - Google Patents
Method and equipment for discriminating headlight of troublemaking car by computer Download PDFInfo
- Publication number
- CN1363483A CN1363483A CN 01102744 CN01102744A CN1363483A CN 1363483 A CN1363483 A CN 1363483A CN 01102744 CN01102744 CN 01102744 CN 01102744 A CN01102744 A CN 01102744A CN 1363483 A CN1363483 A CN 1363483A
- Authority
- CN
- China
- Prior art keywords
- car light
- data
- car
- fragment
- computing machine
- 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
Images
Abstract
A method and device for discriminating the troublemaking car by measuring and analyzing the broken headlight is disclosed. The characteristics recognition, classifying technique, multimedia technique and relational database are used to extract the characteristics and image of headlight. A standard headlight database and an automatic inquiry and discrimination software system for broken headlight are created. It can quickly obtain the kind, name, model and manufacturer of troublemaking car.
Description
The invention belongs to the public safety traffic management technical field, relate to a kind of method and car light characteristic information that adopts computer program to carry out measurement discriminating accident automobile lamp fragment and obtain measurement mechanism.Utilize the method to carry out kind, car name and vehicle that computer search can identify the hit-and-run vehicle fast.
In general, the vehicle escape case is compared with general traffic accident, and few as the material evidence of scouting clue, the difficulty of solving a case is very big.By the investigation to accident site, behind automobile accident, pellicula can have been left in the scene, material evidences such as front and back light fixture fragment, windshield fragment, tire tread, and wherein car paint film sheet and front shroud lamp fragment can become the crucial material evidence of differentiating vehicle.For these specks check authentication methods have 1, the visual inspection method is as size, shape, color etc., take a picture then instrumentation, extraction; 2, brief test such as proportion, chromatographic analysis etc.; 3, microscopic examination method; 4, instrumental method is in specks such as pellicula and light fixture fragment, because the chemical constitution difference will use different analytic instruments to analyze its composition.Comprising infrared absorption spectrum analysis, microspectrophotometry analysis, gas chromatographic analysis and X-ray analysis etc.Utilize above-mentioned check authentication method to extract the formalness physical features and the main analytical parameters (as organic, inorganic constituents, color, thickness, homogeneity, diaphaneity, geometric configuration, size etc.) of various master samples.The formalness feature also is extremely important, industrial products particularly, and identical type is also had any different.Result of study shows, its shape of car light, size, material, floral designs, indentification mark, color and other correlation parameter of the automobile dress usefulness that orthodox car factory produces are all corresponding with car type class, car name, vehicle and manufacturer.That is to say that the car light of the variety classes of different manufacturers production, the usefulness that automobile is adorned of model has both macro and micro feature separately.Therefore, can its feature and master sample be compared by the car light fragment is differentiated, find out kind, car name, vehicle and the manufacturer of escape vehicle, and then, dwindle scope of investigation for tracking down the definite investigation of vehicle escape case direction as early as possible, dwindle the investigation time.
The investigation of accident automobile escape case at home and abroad all is thorny.More external developed countries have utilized the evaluation of the pellicula that accident automobile accident scene is carried over to determine kind, car name, vehicle and the manufacturer of hit-and-run vehicle, have set up Machine Retrieval System.Forensic science research institute of British Home Office reports the progress in analysis of automotive paints, aspect analytical technology, a kind of brand new prepares the embedding paint in gel of the method for thin layer coating agent which has to be dissolved in alcohol before use-promptly, then after solidifying, it carries out ultra-thin section, this method is not only simple and easy to do, fast, but also can test to the level of coating agent which has to be dissolved in alcohol before use with the transmitted light microscope.Compare with the bounce technique paint check of widely adopting, experiment shows that the transmission beam method check is meticulousr, and higher resolution capability is arranged.Experimental results show that: polarized light microscopy analytic method after the improvement and fluorescence microscopic analysis method are equally applicable to the check of thin layer coating agent which has to be dissolved in alcohol before use.Carry out Chemical Analysis and also study using this thin layer coating agent which has to be dissolved in alcohol before use, and the advantage of carrying out this class check in this way is described.In addition, find that also this thin layer coating agent which has to be dissolved in alcohol before use can be used in visible microspectrophotometry analysis (VisibleMicrospectrophotometry) and the Fourier transform infrared microspectrophotometry is analyzed in (FT-IRMicrospectrophotometry).Among the former, resulting transmitted spectrum selectivity absorbs and increases, and it is more easy that its resolution becomes; And among the latter, biggest advantage is exactly to need not to do further sample preparation just can obtain special quality spectrum on multilayer coating agent which has to be dissolved in alcohol before use structure at all levels.Only just can obtain high quality spectrum from quality on the coating agent which has to be dissolved in alcohol before use of nanogram (nanogram).The relevant personnel of German federal criminal investigation office work out a kind of similar methods, in the method, paint chipping are embedded in the resin, and section just can obtain spectrum at all levels in the multilayer automotive paints coating agent which has to be dissolved in alcohol before use with the FT-IR microspectrophotometry at last then.This spectral quality is high, is suitable for doing quantitative analysis and does spectrum stock retrieval with the introducing computing machine.Adopt infra-red sepectrometry to paint to analyze with comparison be a big distinguishing feature in the research of being reported in the past 3 years.Had and used little adamas pond and beam condenser device, the report that adopts the infrared microscopy AAS that trace paint is tested and analyzed.Equally, also there is the people just to use the spectrophotometric micro-accessory device of FT-IR that paint and fiber are carried out assay and done research.The analysis of data and arrangement aspect are exactly to improve technical method to assert vehicle so that utilize traffic accident to run away to be retained in on-the-spot paint chipping after case takes place in the important research project that Britain carries out.Developed a kind of instrument and equipment, can be used to set up a computerized data bank to write down, to store the color data of the outer paint of all automobiles.Facts have proved: but with the spectrocolorimeter at high speed and write down the color of metal or solid-state paint pigment exactly.The color data of Huo Deing can be compared well with the resulting color data of visible light microspectrophotometer measurement trace paint coating agent which has to be dissolved in alcohol before use in this way.In addition, also, study how setting up a color data best source data bank according to the purposes of the outer paint of automobile.Utilize the employed relevant sample of automotive paints industry to set up four different color reference source storehouses, and carried out examination, research with the suitability of the source database of their each self-formings.The result shows: one of them reference source storehouse is more suitable for application in forensic science than other storehouse, source.The feasibility of using this system's identification automobile check, research have been carried out.With microspectrophotometer the painting color data of extracting from 100 different automobiles are measured, then these data and source database ectomesoderm painting color data are compared.Utilize the aberration of color to calculate and the optimum matching selection, just can obtain vehicle identification result.Facts have proved: this system is obtaining very big success aspect the identification automobile, but also has found the factor that influences its success ratio.In addition, utilize the pertinent data of internal layer lacquer also to carry out similar research.Bibliographical information as long as the computing machine that has pellicula about 0.2mm just can the utilization science to search Research Institute in Japan is retrieved, just can detect the kind of vehicle in ten minutes.
Yet utilize the pellicula discrimination method to have many problems, at first staying on-the-spot pellicula mostly is small fragment, be not easy to find very much, and the preparation of sample and analysis is very complicated meticulous, need the technical personnel of higher level to finish, and very time-consuming.Analytical equipment is very expensive, identifies it is the method for very complicated precision as smear in criminal case, and general on-the-spot collection, analysis, discriminating are not that autonomous system is finished.
To the utilization of scene of the accident car light fragment, the prior art method only limits to car light fragment (must be enough big) and the limited visual contrast of car light sample that the people that handles a case collects are judged that the hit-and-run vehicle is the car of what kind vehicle.Visual contrast querying method is as looking for a needle in a haystack, and not only efficient is extremely low, and error is bigger, is difficult to differentiate for car light fragment microscopic feature.
The objective of the invention is to utilize Computer Multimedia Technology, sorting technique and RDBM Relational Data Base Management technology, set up standard car light storehouse and scene of the accident car light fragment and detect inquiry and identification system automatically, a kind of accident automobile lamp computing machine discrimination method and device are provided.
Its shape of car light, size, material, floral designs, indentification mark, color and other correlation parameter of the automobile dress usefulness that orthodox car factory produces, all corresponding with car type class, car name, vehicle and manufacturer.That is to say that the car light of the variety classes of different manufacturers production, the usefulness that automobile is adorned of model has both macro and micro feature separately.Therefore, can its feature and master sample be compared by the car light fragment is differentiated, find out kind, car name, vehicle and the manufacturer of escape vehicle, and then, dwindle scope of investigation for tracking down the definite investigation of vehicle escape case direction as early as possible, dwindle the investigation time.
The present invention is by to the extraction of car light characteristic parameter and obtaining of image, standard car light database structure design and analysis, developed the Machine Retrieval System software package, this system focuses on functions such as the foundation in car light storehouse and modification, the input of fragment characteristic data, retrieval automatically, manual retrieval, the demonstration of car light data image, identification result report output.Utilize native system finally to identify possible accident vehicle.
Main technological route of the present invention as shown in Figure 1.
The present invention collects regular producer automobile lamp, to car light characteristic parameter extraction and Image Acquisition, set up vehicle on computers, the corresponding relation of car light characteristic parameter and car light image, the car light of each vehicle can be obtained by the characteristic parameter of some, this wherein comprises comparison method and determines, the design of car light database structure, data base administrator, set up standard car light data bank, input case fragment is retrieved checking, show identification result, the identification result output print if retrieve two or more multiple vehicles, is revised the input of fragment characteristic parameter and is retrieved given tolerance value, and manually confirm by image recognition, finally identify the accident vehicle.
One, car light fragment material evidence and feature identification method thereof are found and extracted in the scene of the accident
On accident nidus road, find and extract leave over, car light fragment, glass, sand etc.The present invention adopts car light tagged computer retrieval discrimination method, extract car light macroscopic view, microscopic feature and set up standard car light characteristic parameter and image data base, car light fragment feature is imported, Machine Retrieval System by exploitation, car light fragment feature and database standard car light are compared, finally identify possible accident vehicle.The car light chip size that the method requires to differentiate vehicle reduces, as long as fragment can embody the decorative pattern of some, just can identify the relevant vehicle of may causing trouble according to several essential characteristics such as material, color, pattern type, sizes.Certainly feature is many more, and the vehicle that retrieves is few more, and approaching more possible accident vehicle as indentification mark is arranged on the fragment, will significantly reduce the search coverage, improves the detection rate.
Two, regular producer car gage car light sample collection
The present invention is a research object with various vehicles such as car, autotruck, utility van, jeeps, collects various vehicle standard car light samples, sets up the car light sample library.
Three, car light feature extraction and Image Acquisition and device
The present invention adopts the feature identification method, by various car light analyses, classification and measurement being extracted the macrofeature and the microscopic feature of car light, macrofeature comprises material, color, car light external form etc., and microscopic feature comprises local floral designs, decorative pattern size, indentification mark and local feature image etc.Adopt measurement mechanism to have beam trammel microscope eye gauge to be used for determining characteristic parameter, as decorative pattern size, car light overall dimensions, the size positions of fixed orifice.Adopt scanning, photograph and camera shooting and video reading device to car light material object and fragment make a video recording, image-capture and image processing, set up various sample car light fixture features record archives.Car light feature extraction and Image Acquisition are that relational database establishment, fragment searching and module identified framework are set up and the multimedia software design is laid a good foundation.
Four, relational Database Design
The present invention differentiates the characteristics of vehicle according to many feature identification, adopt Microsoft Access data structure opening relationships data bank, this data bank comprises information such as car name (vehicle), the place of production, color, decorative pattern, decorative pattern size, indentification mark, the permanent seat thickness of car light, setting height(from bottom), image etc., differentiate task for finishing rapid retrieval, by standardized process with data organization to the tabulation in, adopt the relational database of two-dimemsional number sheet form, the application testing means determine whether design is meaningful and whether convenient.
Defer to: each field should contain the information of unique type in 1, showing.2, every table must have unique identifier or primary key, it is made of one or more fields in the table, 3,, promptly have one a value also can only be arranged in that arbitrarily the data rows household function is relevant, and this value must be relevant with each entity in the table to each unique primary key value.4, field independence, under the situation that does not influence other means, must arbitrary fields be made amendment (non-primary key field), after the utilization said method, database design of the present invention comprises several tables, wherein has one to be master meter (CAR table), and other is a submeter, can find any one submeter by master meter, also can find master meter by any one submeter.Below narration wherein several tables and enclose content in the table
1 master meter (CAR table) sees Table 1
The recording mechanism of the suitable car of car_id in the master meter (CAR table).Car-id is the primary key of CAR table.Here it belongs to chain field between the table, also claims foreign key.And allow Access that foreign key is set up index.Variation by car-id just can be found different cars.Color-id is the primary key of color data submeter, can find the color data submeter by color-id, rec-id, splint-id, cir-id, cir-teeth-id, loaenge-id, tower-id, spuare-id, six-side-id, five-side-id, paral-id, nofigure-id is respectively the primary key of various decorative pattern data submeters, by the *-id of above various decorative patterns, just can find the submeter of various decorative pattern data, and know the feature that they are had in submeter, chara-fix-id is the primary key of lteral data submeter, can find the character data submeter by charac-fix-id, and picture-id is the primary key of view data submeter, find the view data submeter by picture-id, can obtain image.The data type of these recording mechanisms *-id is digital integer (integer), and master meter (CAR table) canned data also has car-name, car-producer, lamp-material, car-nation, represents car name, producer, car light material, the place of production respectively.Its data type is character type (text).Below in each table the data type integer of field represent the integer numeral, single represents single precision full mold numeral, text represents character type.Generally get 32 character lengths for the character type data field length.
2, color data submeter (COLOR table) sees Table 2
The COLOR table is to be used for storing the color characteristic that car light has.Color characteristic yellow, orange, white, red represent yellow, orange, white, redness respectively.In the color data submeter (COLOR table), different color recording mechanism color-id have corresponding color characteristic, and corresponding with the recording mechanism of car-id car in the table, therefore can link the master meter data from color data submeter (COLOR table) information, find vehicle.
3, lteral data submeter (CHARA-FIX table) sees Table 3
Lteral data submeter (CHARA-FIX table) primary key is written record chara-fix-id, in this table, equally also have car-id number, can find corresponding lteral data submeter by chara-fix-id, obtain car recording mechanism car-id then, therefore can find vehicle.The characteristic information of lteral data submeter storage has indentification mark (chara-sign), fixed form (fix-mode), fixed orifice feature (fix-figure), the setting height(from bottom) of car light (side-high), car light edge thickness (side-thick), fixed orifice internal diameter (in-diameter), fixed hole seat external diameter (out-diameter) waits some data.
4, the submeter of various decorative patterns sees Table 4
According to the decorative pattern feature that various car light had, construct various decorative pattern data submeters, rectangle decorative pattern data submeter (RECTANGLE table) for example, bar shaped decorative pattern data submeters (SPLINT), annulus decorative pattern data submeter (RING table), knuckle-tooth decorative pattern data submeter (CIRTEETH table), diamond tread design data submeter (LOZENGE table), steeple shape decorative pattern data submeter (TOWER table), pentagon decorative pattern data submeters (FIVE-SIDE), hexagon decorative pattern data submeters (SIX-SIDE), square line data submeter (SQUARE), parallelogram decorative pattern data submeters (PARAL), no car light decorative pattern data submeter (NOFIGURE table).
Decorative pattern is one of principal character of car light retrieval discriminating, and a kind of car light may have several types decorative pattern feature simultaneously, as has rectangle decorative pattern and bar shaped decorative pattern feature.In various decorative pattern feature submeters, recording mechanism *-id and car-id are arranged all, can corresponding car-id by *-id and obtain vehicle.Have floral designs information *-type in the decorative pattern data sheet, as the rectangle pattern type flat, concave, convex is arranged, that bar shaped decorative pattern class shape has is flat, concave, convex etc.Every kind of decorative pattern also has the decorative pattern size, as rectangle long length is arranged, wide with, thick thick, and bar shaped has gap-type at interval, thick thick, other pattern type has different dimensional parameters according to the decorative pattern shape.No car light decorative pattern data submeter (NOFIGURE table) is a special case.
5, view data submeter (PICTURE table) sees Table 5
The PICTURE table includes coloured images such as front elevation (front), rear view (RV) (back), character figure (letter) and decorative pattern characteristic pattern.Picture-id is the recording mechanism of view data submeter (PICTURE table), also is the primary key of image index, and it is corresponding with the car-id in this table, is linked with master meter (CAR table).By different car recording mechanism car-id, can obtain the Picture-id recording mechanism and find view data submeter (PICTURE table), access the car light image, further differentiate vehicle.
By using above designing technique, the relational database that can efficiently be linked.On this basis, write and build storehouse and data base administrator and search program, make things convenient for automatically to finish rapidly and build storehouse and retrieval tasks.
Five, build storehouse and database management module
The present invention is on car light feature extraction and Image Acquisition and relational Database Design basis, utilize Visual Basic3.0 (VB) to carry out program design, by direct design user interface, the design forms, place control, call the intrinsic function of VB, the data management system (Date Manager) that provides comprising VB, realize the relational database of the efficient link of establishment Microsoft Access form, standard car light database management module (diagram of circuit is seen Fig. 2) has sets up new storehouse, the car light quantitative data input, the car light data modification, car light browsing data and function such as withdraw from, can be input in the computing machine easily with collecting the standard car light characteristic parameter that extracts, automatically form the various relation tables of front design-calculated, set up standard car light data bank, standard car light database management module master sees Fig. 3 in the interface.
(1) data bank is set up and database update
New model constantly produces, and we just need to bring in constant renewal in data bank, with convenient inquiry, set up new storehouse, will deposit the original content in storehouse, by certain regular expansion base scope and increase the new feature project.
(2) add new record
At first pointer will point to data Kuku tail, and then every car light feature (decorative pattern, color, literal, material etc.) input database, program will guarantee that data pointer dynamically moves.
(3) amendment record
When the record input error, just need amendment record.At first made a window, be used for importing the car name, this has just increased the safety in storehouse, enter the next stage forms, all characteristic datas of car light can see, want to revise a certain, as long as with mouse point twice fast on this, just can eject this modification window, very simple, convenient and practical.
(4) browse record
When selecting to browse writing function, just can see the characteristic data of first car, just can select in the car light storehouse any car light to browse by " last car light " and " next car light ".
Six. the foundation of standard car light data bank
Taking microcomputer is hardware platform, be sampled as the basis with tens kinds of common vehicle standard car lights, find out characteristic parameter, by car light being made a video recording or scanning the car light photo and measure the characteristic dimension data and set up image and property data base, adopting gage measuring or low power microscope eye gauge that method such as take pictures is amplified in the car light part to the car light standard jig in the property data base and car light fragment characteristic dimension obtains, utilization is write builds the storehouse and data base administrator carries out characteristic parameter and image input, has set up the standard car light relational database of efficient link.
Seven, car light fragment searching identification module
This module is the core of system software.Fragment searching identification module flow process is seen Fig. 5
According to car light feature identification method, by traffic accident's scene car light fragment material evidence is extracted, through analytical investigation to a large amount of car light samples, sum up with the car light fragment and differentiate that the principal character of vehicle is a decorative pattern, color, material, make necessary condition with these three features, can compare with the car light characteristic data in the standard car light data bank, identify one or more possible vehicles, for retrieving multiple vehicle more than two, can increase supplemental characteristic, wherein the most important thing is text symbol information, also has fixed class shape in addition, edge thickness, setting height(from bottom) waits dwindles range of search, utilize VB to develop the computer search identification module and can realize the discriminating of car light fragment, this module is provided with (1) crumb data input function, comprising principal data (decorative pattern, color, material, literal) and auxiliary data (fixed class shape, the permanent seat size, edge thickness, setting height(from bottom)); (2) search function is divided into automatic retrieval (comprising the error limit input) and manual retrieval (license number inquiry, the inquiry of car name) again; Whether (3) image comparison function watches fragment to coincide with car light by image.
Fig. 4 is seen at fragment searching identification interface.
(1) fragment characteristic data input
The crumb data input is divided into main information and supplementary two big classes.In main information, decorative pattern divide again ten types selective, color is divided four types again, material divides two types of plastic and glasses, and literal is on the fragment that extracts the indentification mark feature to be arranged for a kind of situation of indentification mark also in two kinds of situation, this feature participates in retrieval as necessary condition, another kind is not have the indentification mark feature on the fragment, is not equal to fragment cooresponding with not having indentification mark on a kind of car light, so this situation does not participate in retrieval as necessary condition.The main messagewindow of fragment such as Fig. 6.Supplementary similarly
(2) retrieval and inquisition
Automatically retrieval
Behind quantitative data input, just can find the car name by automatic inquiry.At first, fragment principal character quantitative data input (material, color, decorative pattern) must be arranged, if do not have input, system can give prompting; Secondly, the input of error limit, for decorative pattern etc. type (protruding as rectangle) is arranged not only, also comparison of sizes must be arranged, this just exists standard car light characteristic dimension measured error in light fixture fragment and the storehouse, so in software when design, considered the specification error scope, | standard car light characteristic dimension among chip size-Ku |≤tolerance value promptly meets the condition of consistent size.Import corresponding tolerance value during retrieval as the case may be, the general tolerance value is big more, and the vehicle that retrieves is many more, and after tested, the general tolerance value is got between 0～1, and default value is 0.5; If there are several cars eligible, will watch in output and see result for retrieval in the result window, result window such as Figure 12.In writing this program, note the various relations of forgiving, for example: if the bar shaped decorative pattern is arranged in the fragment, do not represent car light to have only the bar shaped decorative pattern, may also have other decorative pattern, import all decorative pattern features that fragment has simultaneously and retrieve.
(3) manual retrieval
When we can directly suspect a few cars by fragment, can inquire about automatically again, and just can finish by artificial enquiry.Usually our retrieval earlier automatically when retrieving more than two kinds vehicle, can advance to go on foot one by one artificial enquiry to the vehicle that retrieves, and the manual retrieval is divided into 3 kinds of modes and inquires about, and by the recording mechanism inquiry, press the car name and inquires about and press image inquiry.Forms as shown in figure 14.Further differentiate vehicle by the manual retrieval
(4) artificial affirmation
On automatic retrieval and manual retrieval's basis, further manually confirm, mainly determine vehicle by image recognition.Image is differentiated more directly perceived.The retrieval identification module can conveniently access standard car light database images, divides front elevation, back view, local feature figure such as decorative pattern etc. and indentification mark displayed map.With mouse a bit, this view will dynamically amplify in a certain picture frame, and by recovery button, image just returns to initial condition again.Forms show that as shown in figure 15 forms such as Figure 16, Figure 17 institute were not after a certain image amplified.
(5) identification result and output
By retrieval and inquisition and artificial affirmation, can identify a kind of or two to three kinds of vehicle.Alternate manner is auxiliary to be differentiated if add, auxiliary discriminating comprises traffic accident's scene tire cut, as tread is arranged on the cut, also can measure tire cut distance, or the car light setting height(from bottom) obtains the kind of car and the composition check identification result of other material evidence coat of paint, finally can determine the vehicle of causing trouble, and with printout as a result.
Eight, working process of the present invention and example
The running environment of system
The hardware requirement of car light fragment recognition system is:
(1) one is hardware platform with 486 above microcomputers, supports Chinese Windows3.2 and Window95 to Window98.
(2) need the memory headroom of 32MB at least, memory headroom is to be the bigger the better certainly.
(3) need the hard disk of a 540MB at least, and software package is moved required system files such as dynamic threaded file copy under the window system catalogue.
The startup of system
Before the operation procedure, to comprise SHARE/L:500/F:5100 in the AUTOEXEC.BAT file, under the dos operating system prompt, as operating to the car light storehouse, can key in D: Lpic_lib Lampdb, so just can enter the car light database management module, operation such as under this menu, just can build the storehouse, dose, revise and browse, as discerning to fragment, can key in D: Lpic_lib Lamps, in case the recognition system of entering just can be carried out operations such as crumb data input, retrieval and inquisition.
Fragment computer search identification embodiment
With Isuzu Motors car light fragment is that example illustrates working process of the present invention
(1) the fragment characteristic data extracts
The material of this fragment is plastics, and color has orange and two kinds of whites, and decorative pattern is rectangle (protruding), records the long 5.4mm of decorative pattern size with the oil ga(u)ge slide calliper rule, and wide 4.2mm uses micro-meter Thickness Measurement by Microwave: 2.2mm at the lamp outside face indentification mark E6 numeral 4233 to be arranged
(2) fragment characteristic data input
Double-click D: Lamps.exe file under the Lpic_lib catalogue, enter retrieval identification module (Fig. 4 is seen at main interface), select main menu " input ", drop-down menu appears, select " input of fragment numbering ", eject " input fragment numbering " window, input numbering " 1 ", select " ok ", close this window.Select " main information " in " input " drop-down menu then, eject " the main information of car light fragment " window (see figure 6).Select decorative pattern, eject " decorative pattern messagewindow " (seeing Figure 11), in, select in " rectangle ", eject by " rectangle data window ", select classification " protruding ", the input size is long: 5.4, wide by 4.2, thick 2.2, select " determining ", close the rectangle data window, " decorative pattern messagewindow " do not close at this moment, as decorative pattern feature is in fact arranged simultaneously, can operate in a similar manner.This routine fragment has only a kind of decorative pattern feature of rectangle, so close this window, have only " the main information of car light fragment " window not close this moment, the decorative pattern button color of selecting shoals, and can not select again, also surplus color, material and literal are available, select color button, eject by " color window ", select orange and white two kinds of color (see figure 7)s, choosing determines to close this window, only remaining below material and literal are selected and are imported with similar approach, and material selects plastics, the literal choosing has at the lamp outside face, and input E6 4233 (see figure 10)s, after having imported data, have an acknowledgement window (seeing Figure 11).
(3) retrieval
Behind quantitative data input, just can find the car name by automatic retrieval.Select main window top main menu " retrieval ", drop-down menu occurs, be selected from moving retrieval, eject " input of error limit " window, according to survey precision and retrieval test case, adjusting tolerance value is 0.4; Choosing determines that control ejects " result for retrieval window " (seeing Figure 12), sees that result for retrieval is 1 for the car recording mechanism, and car is called Isuzu Motors, and place of production Japan and car recording mechanism are 17, and car is called blue arrow 3F, the two kinds of vehicles in Kunming, the place of production.Enter relatively interface (seeing Figure 13) of feature then
(4) artificial affirmation
Select main window top main menu " image ", eject " watching image window ", access Isuzu Motors car light image (Fig. 9), and enlarged character view (Figure 17), take fragment and its comparison, in full accord on sign digit on the discovery fragment and the image, and font is too.And accessing blue arrow 3F car light image, numeral and fragment on the image are inconsistent, get rid of the car light of blue arrow 3F vehicle.With the car light that fragment conforms to fully is Isuzu Motors.
(3) identification result and output
On automatic retrieval and manual retrieval's basis, further manually confirm, assert that finally the vehicle of causing trouble is an Isuzu Motors, with identification result and image print output.
The present invention adopts characteristic recognition method, by to car light feature extraction and image acquiring device, setting up standard car light storehouse and scene of the accident car light fragment detects inquiry automatically and differentiates software system, having solved prior art compares by visual limited standard car light material object to enough big car light fragment and collection only according to people's experience, not only efficient is extremely low, and error is bigger, is difficult to the problem of differentiating for car light fragment microscopic feature.Computer media discrimination method provided by the invention and device, convenient and practical, greatly improved car light fragment identification efficiency and accuracy, be kind, car name, vehicle and the manufacturer that finds out the hit-and-run vehicle as early as possible, the detection rate that improves accident automobile escape case provides effective means.
The present invention is with a wide range of applications, and this is not only because the demand of society promotes, and makes this system be easier to promote because of it has following characteristics.
1. pull-down striation menu and Pop-up control key action pane are adopted in the interface, and be both attractive in appearance, convenient again.
2. the identification of car light library management and car light fragment is separate, in identifying operation, can not make amendment to the storehouse, makes the car light storehouse have more safety.
3. each chained list also is not only to connect each other but also separate in the car light storehouse, being connected between convenient inquiry recognition system and the storehouse.
4. multiple inquiry mode is arranged, make retrieval convenient.
5. in operating process, also have prompting and online help function.
The explanation of accompanying drawing table
Table 1 relational database master meter (CAR table)
Table 3 lteral data and fixed form submeter (CHARA-FIX table)
Table 4 rectangle decorative pattern submeter (RECTANGLE table)
Table 5 view data submeter (PICTURE table)
Fig. 1 main technical flows figure of the present invention
Fig. 2 builds storehouse and database management module diagram of circuit
Fig. 3 builds storehouse and database management module master interface
Fig. 4 fragment searching is discerned main interface
Fig. 5 fragment searching identification module diagram of circuit
The main information input window of Fig. 6 fragment
Fig. 7 colouring information input window
Fig. 8 decorative pattern information input window
Fig. 9 text symbol information window
Figure 10 text symbol information input window
The main information browse window of Figure 11 fragment
Figure 12 result for retrieval window
Figure 13 result for retrieval and fragment feature comparison window
Figure 14 artificial enquiry window
Figure 15 watches image window
Figure 16 Isuzu Motors car light image
Figure 17 Isuzu Motors car light character view window
Table 1
| Field name | Data type | Field length |
| ??car-id | ??integer | |
| ??color-id | ??integer | |
| ??car-name | ??text | ??32 |
| ??car-producer | ??text | ??32 |
| ??lamp-material | ??text | ??32 |
| ??car-nation | ??text | ??32 |
| ??rec-id | ??integer | |
| ??splint-id | ??integer | |
| ??cir-id | ??integer | |
| ??tower-id | ??integer | |
| ??Spuare-id | ??integer | |
| ??Six-side-uid | ??integer | |
| ??Five-side-id | ??integer | |
| ??Paral-id | ??integer | |
| ??Nofigure-id | ??integer | |
| ??Picture-id | ??integer | |
| ??Chara-fix-id | ??integer | |
| ??Cir-teeth-id | ??integer | |
| ??Loaenge-id | ??integer |
Table 2
| Field name | Data type | Field length |
| ??color-id | ?integer | |
| ??car-id | ?integer | |
| ??yellow | ?integer | |
| ??orange | ?integer | |
| ??white | ?integer | |
| ??red | ?integer |
Table 3
| Field name | Data type | Field length |
| ??chara-fix-id | ?integer | |
| ??car-id | ?integer | |
| ??chara-sign | ?integer | |
| ??fix-mode | ?text | ?32 |
| ??fix-figure | ?text | ?32 |
| ??side-high | ?single | |
| ??side-thick | ?single | |
| ??in-diameter | ?single | |
| ??ort-diameter | ?single |
Table 4
| Field name | Data type | Field length |
| ????rec-id | ?integer | |
| ????car-id | ?integer | |
| ????rec-type | ?text | ?32 |
| ????length | ?single | |
| ????width | ?single | |
| ????thick | ?single |
Table 5
| Field name | Data type | Field length |
| ??picture-id | ?integer | |
| ??car-id | ?integer | |
| ??front | ?text | ??32 |
| ??back | ?text | ??32 |
| ??lether | ?text | ??32 |
| ??pattern | ?text | ??32 |
Claims (9)
1, a kind of accident automobile lamp computing machine discrimination method, it is characterized in that regular producer automobile lamp is collected, to car light characteristic parameter extraction and Image Acquisition, set up the corresponding relation of vehicle, car light characteristic parameter and car light image on computers, then the car light fragment feature of the scene of the accident is imported, by automatic retrieval or manual retrieval, carry out the car light data image and show, identification result is exported; Used car light characteristic parameter is divided into main feature, comprises color, material, pattern type and size, and supplemental characteristic, mainly comprises indentification mark.
2, accident automobile lamp computing machine discrimination method according to claim 1, it is characterized in that by main feature and supplemental characteristic opening relationships data bank on computers, information in the data bank has car name, the place of production, color, decorative pattern, decorative pattern size, indentification mark, car light permanent seat type and size, edge thickness, setting height(from bottom), image, by standardized process above-mentioned information data is formed tabulation, adopting the pass of two-dimemsional number sheet form is data bank.
3, accident automobile lamp computing machine discrimination method according to claim 2 is characterized in that two-dimentional numerical table includes master meter and submeter, and submeter has color data submeter, lteral data submeter, various decorative pattern submeter, view data submeter.
4, accident automobile lamp computing machine discrimination method according to claim 3, it is characterized in that utilizing Visual Basic3.0 to carry out program design, by direct design user interface, the design forms, place control, call the intrinsic function of Visual Basic, realize the relational database of the efficient link of establishment Microsoft Access form, the standard lamp data management module can be set up new storehouse, car light quantitative data input, car light data modification, car light browsing data and withdraw from.
5, accident automobile lamp computing machine discrimination method according to claim 1 is characterized in that utilizing decorative pattern, color, the material of accident car light fragment to compare with the car light characteristic data in the standard car light data bank, identifies the accident vehicle.
6, accident automobile lamp computing machine discrimination method according to claim 5 is characterized in that utilizing the supplemental characteristic of car light, and supplemental characteristic also includes fixed type, permanent seat size, edge thickness, setting height(from bottom).
7, accident automobile lamp computing machine discrimination method according to claim 6 is characterized in that utilizing Visual Basic at the computer search identification module, and this module is provided with (1) crumb data input function, comprising principal data and auxiliary data; (2) search function comprises automatic retrieval and manual retrieval; (3) image comparison function.
8, the used device of accident automobile lamp computing machine discrimination method according to claim 1 is characterized in that including measurement mechanism, scanning, photograph and camera shooting and video reading device, computing machine; Measurement mechanism has beam trammel, micro-, eye gauge, scanning, photograph and camera shooting and video reading device to car light material object and fragment make a video recording, image-capture and image processing, include relational database, fragment searching and identification module in the computing machine, build storehouse and database management module.
9, the used device of accident automobile lamp computing machine discrimination method according to claim 8, it is characterized in that the used relational database of computing machine adopts Microsoft Access data structure, building storehouse and data management module is to utilize Visual Basic3.0 to carry out program design, comprise the data management system that Visual Basic provides, car light fragment searching and identification module include the input of fragment characteristic data, retrieval and inquisition, automatic retrieval and manual retrieval are arranged, artificial affirmation, identification result and output.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CNB011027444A CN1195281C (en) | 2001-01-04 | 2001-01-04 | Method and equipment for discriminating headlight of troublemaking car by computer |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CNB011027444A CN1195281C (en) | 2001-01-04 | 2001-01-04 | Method and equipment for discriminating headlight of troublemaking car by computer |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN1363483A true CN1363483A (en) | 2002-08-14 |
| CN1195281C CN1195281C (en) | 2005-03-30 |
Family
ID=4652958
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CNB011027444A Expired - Fee Related CN1195281C (en) | 2001-01-04 | 2001-01-04 | Method and equipment for discriminating headlight of troublemaking car by computer |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN1195281C (en) |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103530366A (en) * | 2013-10-12 | 2014-01-22 | 湖北微模式科技发展有限公司 | Vehicle searching method and system based on user-defined features |
| CN104097467A (en) * | 2014-07-08 | 2014-10-15 | 深圳市道通科技股份有限公司 | Tire pressure monitoring and diagnosing equipment and applied hub parameter query method of tire pressure monitoring and diagnosing equipment |
| CN105528617A (en) * | 2015-12-03 | 2016-04-27 | 公安部物证鉴定中心 | Method for carrying out car model retrieval based on headlight shade shard morphological characteristic |
| CN111238370A (en) * | 2020-02-20 | 2020-06-05 | 中国科学院声学研究所东海研究站 | Intelligent detection method and device for KIT board |
| CN113570001A (en) * | 2021-09-22 | 2021-10-29 | 深圳市信润富联数字科技有限公司 | Classification identification positioning method, device, equipment and computer readable storage medium |
-
2001
- 2001-01-04 CN CNB011027444A patent/CN1195281C/en not_active Expired - Fee Related
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103530366A (en) * | 2013-10-12 | 2014-01-22 | 湖北微模式科技发展有限公司 | Vehicle searching method and system based on user-defined features |
| CN103530366B (en) * | 2013-10-12 | 2016-11-23 | 湖北微模式科技发展有限公司 | A kind of vehicle search method and system based on user-defined feature |
| CN104097467A (en) * | 2014-07-08 | 2014-10-15 | 深圳市道通科技股份有限公司 | Tire pressure monitoring and diagnosing equipment and applied hub parameter query method of tire pressure monitoring and diagnosing equipment |
| CN105528617A (en) * | 2015-12-03 | 2016-04-27 | 公安部物证鉴定中心 | Method for carrying out car model retrieval based on headlight shade shard morphological characteristic |
| CN111238370A (en) * | 2020-02-20 | 2020-06-05 | 中国科学院声学研究所东海研究站 | Intelligent detection method and device for KIT board |
| CN113570001A (en) * | 2021-09-22 | 2021-10-29 | 深圳市信润富联数字科技有限公司 | Classification identification positioning method, device, equipment and computer readable storage medium |
| CN113570001B (en) * | 2021-09-22 | 2022-02-15 | 深圳市信润富联数字科技有限公司 | Classification identification positioning method, device, equipment and computer readable storage medium |
Also Published As
| Publication number | Publication date |
|---|---|
| CN1195281C (en) | 2005-03-30 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US9014430B2 (en) | Image analysis platform for identifying artifacts in samples and laboratory consumables | |
| Flynn et al. | Forensic applications of infrared chemical imaging: multi-layered paint chips | |
| CN103530366B (en) | A kind of vehicle search method and system based on user-defined feature | |
| CN1656371A (en) | Surface defect judging method | |
| CN105486684A (en) | Non-destructive testing method for mahogany furniture wood varieties based on network platform | |
| CN1848162A (en) | Method, system and program for evaluating reliability on component | |
| EP2963569A2 (en) | Information handling based on trust | |
| CN110879219B (en) | Water mineral oil detection method based on three-dimensional fluorescence spectrum | |
| CN112730326A (en) | Intelligent identification device and method for rock slices | |
| CN106227788A (en) | Database query method based on Lucene | |
| CN101079124A (en) | Method for converting enterprise information to electronic media and sequencing according to trade | |
| CN1195281C (en) | Method and equipment for discriminating headlight of troublemaking car by computer | |
| CN109342712A (en) | A kind of information automatic recognition method of fluorescence immunity analyzer | |
| Kruglak et al. | The evidentiary significance of automotive paint from the Northeast: a study of red paint | |
| US20230393066A1 (en) | Information processing system and information processing method | |
| Kukkola et al. | Easy and accessible way to calibrate a fluorescence microscope and to create a microplastic identification key | |
| Cortea et al. | INFRA-ART: An open access spectral library of art-related materials as a digital support tool for cultural heritage science | |
| Sherma | Advances in the thin-layer chromatographic forensic analysis of inks | |
| CN1881211A (en) | Graphic retrieve method | |
| Coli et al. | Microcomputed tomography for discriminating between different forming techniques in ancient pottery: New segmentation method and pore distribution recognition | |
| CN105677723A (en) | Method for establishing and searching data labels for industrial signal source | |
| CN110377768B (en) | Intelligent graph recognition system and method | |
| CN106323937B (en) | A kind of the Pubei oilfield spectrum structure and discrimination method of high sense | |
| CN1684071A (en) | Sequence indexing method and system | |
| CN111795919A (en) | Bone marrow cell morphology automatic detection system and working method thereof |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C06 | Publication | ||
| PB01 | Publication | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| C19 | Lapse of patent right due to non-payment of the annual fee | ||
| CF01 | Termination of patent right due to non-payment of annual fee |