CN103065143A - Automobile logo identification method and device - Google Patents
Automobile logo identification method and device Download PDFInfo
- Publication number
- CN103065143A CN103065143A CN2012105923108A CN201210592310A CN103065143A CN 103065143 A CN103065143 A CN 103065143A CN 2012105923108 A CN2012105923108 A CN 2012105923108A CN 201210592310 A CN201210592310 A CN 201210592310A CN 103065143 A CN103065143 A CN 103065143A
- Authority
- CN
- China
- Prior art keywords
- car
- map
- picture
- proper vector
- marking
- 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
Images
Landscapes
- Image Analysis (AREA)
Abstract
The invention discloses an automobile logo identification method and a device. The method includes the steps of obtaining automobile logo images, carrying out feature transformation process on the automobile logo images for obtaining feature vector collection of the automobile logo images, matching feature points in the feature vector collection of the automobile logo images with the feature points in preset feature vector collection for obtaining matched feature point collection, and obtaining corresponding vehicle types according to the matched feature point collection. According to the automobile logo identification method and the device, the problem that identification of vehicles is influenced due to the fact that interference objects of automobile licenses influence identification of the automobile logos in the prior art is solved, and accurate identification results of the automobile logos are obtained.
Description
Technical field
The present invention relates to image processing field, identify other method and device in particular to a kind of car.
Background technology
Along with socioeconomic development, increasing of vehicle becomes inevitable by computer information, intelligentized management vehicle.License plate recognition technology is widely used in traffic flow monitoring, and the charge of highway bayonet socket is in red light violation vehicle monitoring and the residential quarter automatic fare collection system.Vehicle recongnition technique is one of gordian technique of intelligent transportation field, it is the new research direction of vehicle recongnition technique that car identifies other technology, the car target type of vehicle can be divided into circle by shape, oval and square, present treatment technology can only be identified, but can not identify concrete vehicle car plate and large-scale, medium-sized, dilly.The identification of car target is divided into car mark coarse positioning substantially, and car mark fine positioning, car identify other three steps, utilize in the prior art car plate and car target topological relation, determine car target approximate region; Extract accurately again the car mark according to the rough position that extracts, but the car mark is being positioned, when identifying, cutting apart because the chaff interference in some background and the place ahead can have influence on the car target, and then have influence on the identification of car target.
Affect the car sign not for the chaff interference owing to car plate in the prior art, thereby the problem of vehicle is distinguished in impact, not yet proposes at present effective solution.
Summary of the invention
For correlation technique because the chaff interference of car plate affect the car sign not, thereby affect the problem of distinguishing vehicle, effective solution is not yet proposed at present, for this reason, fundamental purpose of the present invention is to provide a kind of car to identify other method and device, to address the above problem.
To achieve these goals, according to an aspect of the present invention, provide a kind of car to identify other method, the method comprises: obtain the car picture of marking on a map; The car picture of marking on a map is carried out eigentransformation and processes, to obtain mark on a map the proper vector set of picture of car; The unique point that car is marked on a map in the proper vector set of picture is carried out matching treatment with the unique point in the set of default proper vector, gather to obtain matching characteristic point; Corresponding car mark type is obtained in set according to matching characteristic point.
Further, the car picture of marking on a map is carried out eigentransformation and processes, comprise to obtain mark on a map the step of proper vector set of picture of car: the car picture of marking on a map is carried out the location, edge, to obtain edge image; The edge image carries out eigentransformation to be processed, to obtain the set of corresponding every breadths edge Characteristic of Image vector; Cluster analysis is carried out in proper vector set, with obtain a plurality of class data and with each class data characteristic of correspondence average; Each characteristic mean is carried out the label setting, to obtain the characteristic mean collection that label is set, the characteristic mean collection is gathered as proper vector.
Further, obtaining mark on a map the step of picture of car comprises: gather original image; Original image is carried out car demarcate the position and calculate, to obtain initial car in the original image picture of marking on a map; The initial car picture of marking on a map is intercepted processing, to obtain the car picture of marking on a map.
Further, unique point in the proper vector set of picture that car is marked on a map and the unique point in the set of default proper vector are carried out matching treatment, to obtain before the matching characteristic point set, method comprises: inspection vehicle mark on a map whether exist in the proper vector set of picture with default proper vector set in the unique point that is complementary of unique point, and in car is marked the proper vector set of picture on a map, exist with default proper vector set in the situation of the unique point that is complementary of unique point under, obtain matching characteristic point and gather.
Further, unique point in the proper vector set of picture that car is marked on a map and the unique point in the set of default proper vector are carried out matching treatment, to obtain before the matching characteristic point set, method comprises: obtain the default car image set of marking on a map, wherein, the default car image set of marking on a map comprises one or more default cars picture of marking on a map; Default car any one default car in image set picture of marking on a map of marking on a map is carried out eigentransformation and processes, obtaining default proper vector, and obtain default proper vector set.
To achieve these goals, according to an aspect of the present invention, provide not device of a kind of car sign, this device comprises: comprising: the first acquisition module, for obtaining the car picture of marking on a map; The first processing module is used for the car picture of marking on a map is carried out eigentransformation and processes, to obtain mark on a map the proper vector set of picture of car; The second processing module is used for mark on a map the unique point of proper vector set of picture of car is carried out matching treatment with the unique point in the set of default proper vector, gathers to obtain matching characteristic point; The 3rd processing module is used for gathering the car mark type of obtaining correspondence according to matching characteristic point.
Further, the first processing module comprises: the first sub-processing module is used for car is marked on a map as carrying out the location, edge, to obtain edge image; The second sub-processing module is used for the edge image and carries out the eigentransformation processing, to obtain the set of corresponding every breadths edge Characteristic of Image vector; The first analysis module is used for to proper vector set and carries out cluster analysis, with obtain a plurality of class data and with each class data characteristic of correspondence average; First arranges module, is used for each characteristic mean is carried out the label setting, to obtain the characteristic mean collection that label is set, the characteristic mean collection is gathered as proper vector.
Further, the first acquisition module comprises: acquisition module is used for gathering original image; The first computing module is used for that original image is carried out car and demarcates the position and calculate, to obtain initial car in the original image picture of marking on a map; The 3rd sub-processing module is used for the initial car picture of marking on a map is intercepted processing, to obtain the car picture of marking on a map.
Further, device comprises: the 4th sub-processing module, for detection of car mark on a map whether exist in the proper vector set of picture with default proper vector set in the unique point that is complementary of unique point, and in car is marked the proper vector set of picture on a map, exist with default proper vector set in the situation of the unique point that is complementary of unique point under, obtain matching characteristic point and gather.
Further, unique point in the proper vector set of picture that car is marked on a map and the unique point in the set of default proper vector are carried out matching treatment, to obtain before the matching characteristic point set, method comprises: the second acquisition module, be used for obtaining the default car image set of marking on a map, wherein, the default car image set of marking on a map comprises one or more default cars picture of marking on a map; The 5th sub-processing module is used for default car any one default car of image set picture of marking on a map of marking on a map is carried out eigentransformation and processes, and obtaining default proper vector, and obtains default proper vector set.
By the present invention, mark on a map and use the SIFT algorithm that car to be identified is marked on a map after the picture to look like to process obtaining car, obtain mark on a map the SIFT proper vector set of picture of car, then mark on a map the unique point of picture of the unique point in the default proper vector set and car to be identified is carried out matching operation, realized that the car sign in the image to be identified is other, solved in the prior art because the chaff interference of car plate affects the car sign not, thereby vehicle is distinguished in impact, realized Obtaining Accurate car target recognition result, thus can be to complex background, tilt, deformation, dirty, partial occlusion, the effect that the car mark that light changes is effectively identified.
Description of drawings
Accompanying drawing described herein is used to provide a further understanding of the present invention, consists of the application's a part, and illustrative examples of the present invention and explanation thereof are used for explaining the present invention, do not consist of improper restriction of the present invention.In the accompanying drawings:
Fig. 1 be according to the car of embodiment of the invention sign not Zhuan Zhi structural representation;
Fig. 2 is the process flow diagram that identifies other method according to the car of the embodiment of the invention; And
Fig. 3 is the process flow diagram that identifies other method according to car embodiment illustrated in fig. 2.
Embodiment
Need to prove that in the situation of not conflicting, embodiment and the feature among the embodiment among the application can make up mutually.Describe below with reference to the accompanying drawings and in conjunction with the embodiments the present invention in detail.
Fig. 1 be according to the car of embodiment of the invention sign not Zhuan Zhi structural representation.As shown in Figure 1, this device can comprise: the first acquisition module 10 is used for obtaining the car picture of marking on a map; The first processing module 30 is used for the car picture of marking on a map is carried out eigentransformation and processes, to obtain mark on a map the proper vector set of picture of car; The second processing module 50 is used for mark on a map the unique point of proper vector set of picture of car is carried out matching treatment with the unique point in the set of default proper vector, gathers to obtain matching characteristic point; The 3rd processing module 70 is used for gathering the car mark type of obtaining correspondence according to matching characteristic point.
Adopt the present invention, obtain the car picture of marking on a map by the first acquisition module, then by the first processing module car is marked on a map and process as carrying out eigentransformation, to obtain mark on a map the proper vector set of picture of car, unique point during unique point in the second processing module is marked the proper vector set of picture on a map to car is gathered with default proper vector is carried out matching treatment, obtaining after the matching characteristic point set, by the 3rd processing module according to matching characteristic point set obtain the car mark type of correspondence.By the present invention, mark on a map and use the SIFT algorithm that car to be identified is marked on a map after the picture to look like to process obtaining car, obtain mark on a map the SIFT proper vector set of picture of car, then mark on a map the unique point of picture of the unique point in the default proper vector set and car to be identified is carried out matching operation, realized the identification of the car mark type in the image to be identified, solved in the prior art because the chaff interference of car plate affects the car sign not, thereby vehicle is distinguished in impact, realized Obtaining Accurate car target recognition result, thus can be to complex background, tilt, deformation, dirty, partial occlusion, the effect that the car mark that light changes is effectively identified.
In the above embodiment of the present invention, the first processing module 30 can comprise: the first sub-processing module is used for car is marked on a map as carrying out the location, edge, to obtain edge image; The second sub-processing module is used for the edge image and carries out the eigentransformation processing, to obtain the set of corresponding every breadths edge Characteristic of Image vector; The first analysis module is used for to proper vector set and carries out cluster analysis, with obtain a plurality of class data and with each class data characteristic of correspondence average; First arranges module, is used for each characteristic mean is carried out the label setting, to obtain the characteristic mean collection that label is set, the characteristic mean collection is gathered as proper vector.
According to the abovementioned embodiments of the present invention, the first acquisition module 10 can comprise: acquisition module is used for gathering original image; The first computing module is used for that original image is carried out car and demarcates the position and calculate, to obtain initial car in the original image picture of marking on a map; The 3rd sub-processing module is used for the initial car picture of marking on a map is intercepted processing, to obtain the car picture of marking on a map.
In the above embodiment of the present invention, device can comprise: the 4th sub-processing module, for detection of car mark on a map whether exist in the proper vector set of picture with default proper vector set in the unique point that is complementary of unique point, and in car is marked the proper vector set of picture on a map, exist with default proper vector set in the situation of the unique point that is complementary of unique point under, obtain matching characteristic point and gather.
According to the abovementioned embodiments of the present invention, unique point in the proper vector set of picture that car is marked on a map and the unique point in the set of default proper vector are carried out matching treatment, to obtain before the matching characteristic point set, method can also comprise: the second acquisition module, be used for obtaining the default car image set of marking on a map, wherein, the default car image set of marking on a map comprises one or more default cars picture of marking on a map; The 5th sub-processing module is used for default car any one default car of image set picture of marking on a map of marking on a map is carried out eigentransformation and processes, and obtaining default proper vector, and obtains default proper vector set.
Fig. 2 is the process flow diagram that identifies other method according to the car of the embodiment of the invention.Fig. 3 is the process flow diagram that identifies other method according to car embodiment illustrated in fig. 2.
As shown in Figures 2 and 3, the method comprises the steps:
Step S102 obtains the car picture of marking on a map.
Step S104 carries out eigentransformation and processes the car picture of marking on a map, to obtain mark on a map the proper vector set of picture of car.
Step S106 carries out matching treatment to the unique point that car is marked on a map in the proper vector set of picture with the unique point in the set of default proper vector, gathers to obtain matching characteristic point.
Step S108, corresponding car mark type is obtained in set according to matching characteristic point.
Adopt the present invention, obtain car mark on a map the picture after, car is marked on a map as carrying out the eigentransformation processing, to obtain mark on a map the proper vector set of picture of car, unique point in the proper vector set of picture that car is marked on a map and the unique point in the set of default proper vector are carried out matching treatment, to obtain after the matching characteristic point set, the car mark type of correspondence is obtained in set according to matching characteristic point.By the present invention, mark on a map and use the SIFT algorithm that car to be identified is marked on a map after the picture to look like to process obtaining car, obtain mark on a map the SIFT proper vector set of picture of car, then mark on a map the unique point of picture of the unique point in the default proper vector set and car to be identified is carried out matching operation, realized that the car sign in the image to be identified is other, solved in the prior art because the chaff interference of car plate affects the car sign not, thereby vehicle is distinguished in impact, realized the recognition result of Obtaining Accurate car mark type, thus can be to complex background, tilt, deformation, dirty, partial occlusion, the effect that the car mark that light changes is effectively identified.
In the above embodiment of the present invention, the car picture of marking on a map is carried out eigentransformation and processes, comprise to obtain mark on a map the step of proper vector set of picture of car: the car picture of marking on a map is carried out the location, edge, to obtain edge image; The edge image carries out eigentransformation to be processed, to obtain the set of corresponding every breadths edge Characteristic of Image vector; Cluster analysis is carried out in proper vector set, with obtain a plurality of class data and with each class data characteristic of correspondence average; Each characteristic mean is carried out the label setting, to obtain the characteristic mean collection that label is set, the characteristic mean collection is gathered as proper vector.
Particularly, after the acquisition car is marked picture on a map, each width of cloth car is marked on a map as carrying out pre-service, particularly car is marked on a map as carrying out the location, edge, to obtain edge image, utilize the SIFT operator to obtain the SIFT proper vector set of each breadths edge image, with the KNN algorithm SIFT proper vector is gathered afterwards and carried out cluster analysis, obtain K class and characteristic of correspondence average, and give a label for each characteristic mean, identifying the class of this characteristic mean representative, and obtain the preferred SIFT proper vector that the characteristic mean by K tape label consists of and gather.
Wherein, the SIFT unique point has the local feature of image, and rotation, yardstick convergent-divergent, brightness variation are maintained the invariance, visual angle change, affined transformation, noise are also kept to a certain degree stability, thereby can use the unique point in the SIFT proper vector set to mate fast and accurately with the unique point in the set of default proper vector, thereby can effectively identify the car mark of complex background, inclination, deformation, dirt, partial occlusion, light variation.
Step S208 as shown in Figure 3: obtain the initial car picture of marking on a map.
Step S210: the initial car picture of marking on a map is intercepted processing, to obtain the car picture of marking on a map.
Step S212: the car picture of marking on a map is carried out eigentransformation and processes, to obtain mark on a map the proper vector set of picture of car.
Use the SIFT algorithm that the initial car that obtains in the above-mentioned steps picture of marking on a map is processed, obtain mark on a map the SIFT unique point vector set of picture of car to be identified.
Particularly, above-described embodiment kind in the application, obtain mark on a map the SIFT proper vector set of picture of car, utilize the SIFT algorithm to obtain each width of cloth car car target SIFT unique point in the picture of marking on a map, then calculate car to be identified mark on a map the histogram distance of picture of picture and default car of marking on a map; Mark on a map as corresponding object as recognition result apart from the target carriage of minimum with histogram.
According to the abovementioned embodiments of the present invention, obtaining mark on a map the step of picture of car comprises: gather original image; Original image is carried out car demarcate the position and calculate, to obtain initial car in the original image picture of marking on a map; The initial car picture of marking on a map is intercepted processing, to obtain the car picture of marking on a map.
Particularly, the original image that collects is carried out pre-service, and original image carried out car target location Calculation, obtain one or more initial cars that may comprise car mark candidate region picture of marking on a map, and intercept initial car according to actual conditions and mark a part in the picture on a map as the car to be identified picture of marking on a map, then the picture of car to be identified being marked on a map uses the SIFT algorithm to carry out eigentransformation and processes, obtain the set of SIFT proper vector, then according to this proper vector set and the unique point of presetting set of eigenvectors total calculation coupling, obtain the matching characteristic set, obtain the result of identification car mark type.
Particularly, as shown in Figure 3, execution in step S202: gather original image, particularly, filming apparatus is installed on the position of highway crossing, charge station, parking lot or other requirement, vehicle is carried out image acquisition, obtain containing mark on a map the original image of picture of car, then carry out car and demarcate the position and calculate containing mark on a map the original image of picture of car, obtain one or morely may comprising the initial car of the car target picture of marking on a map.
In the above embodiment of the present invention, unique point in the proper vector set of picture that car is marked on a map and the unique point in the set of default proper vector are carried out matching treatment, to obtain before the matching characteristic point set, method can also comprise: inspection vehicle mark on a map whether exist in the proper vector set of picture with default proper vector set in the unique point that is complementary of unique point, and in car is marked the proper vector set of picture on a map, exist with default proper vector set in the situation of the unique point that is complementary of unique point under, obtain matching characteristic point and gather.
Particularly, carry out step S214 shown in Figure 3: inspection vehicle mark on a map whether exist in the proper vector set of picture with default proper vector set in the unique point that is complementary of unique point, and after the unique point of searching mutual coupling execution in step S216, obtain matching characteristic point and gather; Not have the unique point of mutually mating if default proper vector set and car are marked on a map in the proper vector set of picture, judge that then image to be identified does not comprise the car mark, then return execution in step S208 until handle all candidates' the car picture of marking on a map.
According to the abovementioned embodiments of the present invention, unique point in the proper vector set of picture that car is marked on a map and the unique point in the set of default proper vector are carried out matching treatment, to obtain before the matching characteristic point set, method comprises: obtain the default car image set of marking on a map, wherein, the default car image set of marking on a map comprises one or more default cars picture of marking on a map; Default car any one default car in image set picture of marking on a map of marking on a map is carried out eigentransformation and processes, obtaining default proper vector, and obtain default proper vector set.
Particularly, as shown in Figure 3, execution in step S204: obtain the default car image set of marking on a map.Particularly, the car mark that contains whole vehicles is carried out high definition take, the car mark part in the cut-away view picture, and car mark type arranged forms the default car image set of marking on a map.
Execution in step S206: each default car picture of marking on a map is carried out eigentransformation and processes, and obtain default proper vector set.Particularly, use the SIFT algorithm to obtain each default car proper vector of picture of marking on a map, and obtain mark on a map the default proper vector set of image set of default car.
From above description, can find out, the present invention has realized following technique effect: by the present invention, mark on a map and use the SIFT algorithm that car to be identified is marked on a map after the picture to look like to process obtaining car, obtain mark on a map the SIFT proper vector set of picture of car, then mark on a map the unique point of picture of the unique point in the default proper vector set and car to be identified is carried out matching operation, realized that the car sign in the image to be identified is other, solved in the prior art because the chaff interference of car plate affects the car sign not, thereby vehicle is distinguished in impact, realized Obtaining Accurate car target recognition result, thus can be to complex background, tilt, deformation, dirty, partial occlusion, the effect that the car mark that light changes is effectively identified.
Through the above description of the embodiments, those skilled in the art can be well understood to the present invention and can realize by the mode that software adds essential general hardware platform, can certainly pass through hardware, but the former is better embodiment in a lot of situation.Based on such understanding, the part that technical scheme of the present invention contributes to prior art in essence in other words can embody with the form of software product, this computer software product can be stored in the storage medium, such as ROM/RAM, magnetic disc, CD etc., comprise that some instructions are with so that a computer equipment (can be personal computer, server, the perhaps network equipment etc.) carry out the described method of some part of each embodiment of the present invention or embodiment.
The above only is preferred implementation of the present invention; should be pointed out that for those skilled in the art, under the prerequisite that does not break away from the principle of the invention; can also make some improvements and modifications, these improvements and modifications also should be considered as protection scope of the present invention.
Claims (10)
1. a car identifies other method, it is characterized in that, comprising:
Obtain the car picture of marking on a map;
The described car picture of marking on a map is carried out eigentransformation and processes, to obtain mark on a map the proper vector set of picture of described car;
The unique point that described car is marked on a map in the proper vector set of picture is carried out matching treatment with the unique point in the set of default proper vector, gather to obtain matching characteristic point;
Obtain corresponding car mark type according to described matching characteristic point set.
2. method according to claim 1 is characterized in that, the described car picture of marking on a map is carried out eigentransformation and processes, and comprises to obtain mark on a map the step of proper vector set of picture of described car:
Described car is marked on a map as carrying out the location, edge, to obtain edge image;
Described edge image is carried out eigentransformation process, to obtain the proper vector set of corresponding every described edge image;
Cluster analysis is carried out in set to described proper vector, with obtain a plurality of class data and with each described class data characteristic of correspondence average;
Each described characteristic mean is carried out the label setting, to obtain the characteristic mean collection that label is set, described characteristic mean collection is gathered as described proper vector.
3. method according to claim 1 is characterized in that, obtains mark on a map the step of picture of car and comprises:
Gather original image;
Described original image is carried out car demarcate the position and calculate, to obtain initial car in the described original image picture of marking on a map;
The described initial car picture of marking on a map is intercepted processing, to obtain the described car picture of marking on a map.
4. method according to claim 3 is characterized in that, the unique point in the set of the proper vector of picture that described car is marked on a map is carried out matching treatment with the unique point in the set of default proper vector, and to obtain before matching characteristic point gathers, described method comprises:
Detect described car mark on a map whether exist in the proper vector set of picture with described default proper vector set in the unique point that is complementary of unique point;
In described car is marked the proper vector set of picture on a map, exist with default proper vector set in the situation of the unique point that is complementary of unique point under, obtain described matching characteristic point and gather.
5. method according to claim 1 is characterized in that, the unique point in the set of the proper vector of picture that described car is marked on a map is carried out matching treatment with the unique point in the set of default proper vector, and to obtain before matching characteristic point gathers, described method comprises:
Obtain the default car image set of marking on a map, wherein, the described default car image set of marking on a map comprises one or more default cars picture of marking on a map;
Described default car any one described default car in image set picture of marking on a map of marking on a map is carried out eigentransformation and processes, obtaining default proper vector, and obtain described default proper vector set.
Car sign not Zhuan Zhi, it is characterized in that, comprising:
The first acquisition module is used for obtaining the car picture of marking on a map;
The first processing module is used for the described car picture of marking on a map is carried out eigentransformation and processes, to obtain mark on a map the proper vector set of picture of described car;
The second processing module is used for mark on a map the unique point of proper vector set of picture of described car is carried out matching treatment with the unique point in the set of default proper vector, gathers to obtain matching characteristic point;
The 3rd processing module is used for obtaining corresponding car mark type according to described matching characteristic point set.
7. device according to claim 6 is characterized in that, described the first processing module comprises:
The first sub-processing module is used for described car is marked on a map as carrying out the location, edge, to obtain edge image;
The second sub-processing module is used for that described edge image is carried out eigentransformation and processes, to obtain the proper vector set of corresponding every described edge image;
The first analysis module is used for cluster analysis is carried out in described proper vector set, with obtain a plurality of class data and with each described class data characteristic of correspondence average;
First arranges module, is used for each described characteristic mean is carried out the label setting, to obtain the characteristic mean collection that label is set, described characteristic mean collection is gathered as described proper vector.
8. method according to claim 6 is characterized in that, described the first acquisition module comprises:
Acquisition module is used for gathering original image;
The first computing module is used for that described original image is carried out car and demarcates the position and calculate, to obtain initial car in the described original image picture of marking on a map;
The 3rd sub-processing module is used for the described initial car picture of marking on a map is intercepted processing, to obtain the described car picture of marking on a map.
9. device according to claim 8 is characterized in that, described device comprises:
The 4th sub-processing module, for detection of described car mark on a map whether exist in the proper vector set of picture with described default proper vector set in the unique point that is complementary of unique point, and in described car is marked the proper vector set of picture on a map, exist with default proper vector set in the situation of the unique point that is complementary of unique point under, obtain described matching characteristic point and gather.
10. device according to claim 6 is characterized in that, the unique point in the set of the proper vector of picture that described car is marked on a map is carried out matching treatment with the unique point in the set of default proper vector, and to obtain before matching characteristic point gathers, described method comprises:
The second acquisition module is used for obtaining the default car image set of marking on a map, and wherein, the described default car image set of marking on a map comprises one or more default cars picture of marking on a map;
The 5th sub-processing module is used for described default car any one described default car of image set picture of marking on a map of marking on a map is carried out eigentransformation and processes, and obtaining default proper vector, and obtains described default proper vector set.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN2012105923108A CN103065143A (en) | 2012-12-30 | 2012-12-30 | Automobile logo identification method and device |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN2012105923108A CN103065143A (en) | 2012-12-30 | 2012-12-30 | Automobile logo identification method and device |
Publications (1)
Publication Number | Publication Date |
---|---|
CN103065143A true CN103065143A (en) | 2013-04-24 |
Family
ID=48107766
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN2012105923108A Pending CN103065143A (en) | 2012-12-30 | 2012-12-30 | Automobile logo identification method and device |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN103065143A (en) |
Cited By (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103488973A (en) * | 2013-09-12 | 2014-01-01 | 上海依图网络科技有限公司 | Method and system for recognizing vehicle brand based on image |
CN105096308A (en) * | 2014-05-13 | 2015-11-25 | 斯耐尔有限公司 | Logo detection by edge matching |
CN105528610A (en) * | 2014-09-30 | 2016-04-27 | 阿里巴巴集团控股有限公司 | Character recognition method and device |
CN106599850A (en) * | 2016-12-16 | 2017-04-26 | 合肥寰景信息技术有限公司 | Traffic analysis method and traffic analysis system based on vehicle logo recognition |
CN107944352A (en) * | 2017-11-09 | 2018-04-20 | 石数字技术成都有限公司 | A kind of Vehicle License Plate Recognition System and method based on picture retrieval method |
Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101196979A (en) * | 2006-12-22 | 2008-06-11 | 四川川大智胜软件股份有限公司 | Method for recognizing vehicle type by digital picture processing technology |
CN102497539A (en) * | 2011-12-15 | 2012-06-13 | 航天科工哈尔滨风华有限公司 | Panoramic monitoring system and monitoring method of the same based on improved SIFT feature matching |
-
2012
- 2012-12-30 CN CN2012105923108A patent/CN103065143A/en active Pending
Patent Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101196979A (en) * | 2006-12-22 | 2008-06-11 | 四川川大智胜软件股份有限公司 | Method for recognizing vehicle type by digital picture processing technology |
CN102497539A (en) * | 2011-12-15 | 2012-06-13 | 航天科工哈尔滨风华有限公司 | Panoramic monitoring system and monitoring method of the same based on improved SIFT feature matching |
Non-Patent Citations (2)
Title |
---|
路昆: "运动车辆标牌识别算法的研究", 《中国优秀硕士学位论文全文数据库》 * |
钱志伟: "智能交通系统中车型识别的研究与应用", 《中国优秀硕士学位论文全文数据库》 * |
Cited By (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103488973A (en) * | 2013-09-12 | 2014-01-01 | 上海依图网络科技有限公司 | Method and system for recognizing vehicle brand based on image |
CN103488973B (en) * | 2013-09-12 | 2017-10-20 | 上海依图网络科技有限公司 | Vehicle brand recognition methods and system based on image |
CN105096308A (en) * | 2014-05-13 | 2015-11-25 | 斯耐尔有限公司 | Logo detection by edge matching |
CN105528610A (en) * | 2014-09-30 | 2016-04-27 | 阿里巴巴集团控股有限公司 | Character recognition method and device |
CN105528610B (en) * | 2014-09-30 | 2019-05-07 | 阿里巴巴集团控股有限公司 | Character recognition method and device |
CN106599850A (en) * | 2016-12-16 | 2017-04-26 | 合肥寰景信息技术有限公司 | Traffic analysis method and traffic analysis system based on vehicle logo recognition |
CN107944352A (en) * | 2017-11-09 | 2018-04-20 | 石数字技术成都有限公司 | A kind of Vehicle License Plate Recognition System and method based on picture retrieval method |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
Holgado‐Barco et al. | Automatic inventory of road cross‐sections from mobile laser scanning system | |
Mammeri et al. | Lane detection and tracking system based on the MSER algorithm, hough transform and kalman filter | |
Luvizon et al. | Vehicle speed estimation by license plate detection and tracking | |
Lee et al. | Directional-DBSCAN: Parking-slot detection using a clustering method in around-view monitoring system | |
CN101339601B (en) | License plate Chinese character recognition method based on SIFT algorithm | |
Danti et al. | An image processing approach to detect lanes, pot holes and recognize road signs in Indian roads | |
CN103065143A (en) | Automobile logo identification method and device | |
EP2813973B1 (en) | Method and system for processing video image | |
Li et al. | A robust lane detection method based on hyperbolic model | |
Abou Chacra et al. | Fully automated road defect detection using street view images | |
CN103065144A (en) | Automobile logo identification method and device | |
Chen et al. | Front vehicle detection and distance estimation using single-lens video camera | |
Du et al. | Dynamic pavement distress image stitching based on fine‐grained feature matching | |
CN112699267A (en) | Vehicle type recognition method | |
Satzoda et al. | Robust extraction of lane markings using gradient angle histograms and directional signed edges | |
Chen et al. | A precise information extraction algorithm for lane lines | |
Srivastava et al. | Video-based real-time surveillance of vehicles | |
CN104408431A (en) | Vehicle model identification method under traffic monitoring | |
Xuan et al. | Robust lane-mark extraction for autonomous driving under complex real conditions | |
CN103093205B (en) | Identify car calibration method | |
CN103559492A (en) | Car logo recognition device and method | |
Vuong et al. | Toward planet-wide traffic camera calibration | |
JP4784932B2 (en) | Vehicle discrimination device and program thereof | |
Deb et al. | Automatic vehicle identification by plate recognition for intelligent transportation system applications | |
Ding et al. | A comprehensive approach for road marking detection and recognition |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20130424 |
|
RJ01 | Rejection of invention patent application after publication |