CN110032991B - Car logo detection and identification method based on car logo relocation - Google Patents
Car logo detection and identification method based on car logo relocation Download PDFInfo
- Publication number
- CN110032991B CN110032991B CN201910331226.2A CN201910331226A CN110032991B CN 110032991 B CN110032991 B CN 110032991B CN 201910331226 A CN201910331226 A CN 201910331226A CN 110032991 B CN110032991 B CN 110032991B
- Authority
- CN
- China
- Prior art keywords
- logo
- car
- license plate
- car logo
- image
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
- G06F18/2411—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on the proximity to a decision surface, e.g. support vector machines
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/40—Extraction of image or video features
- G06V10/50—Extraction of image or video features by performing operations within image blocks; by using histograms, e.g. histogram of oriented gradients [HoG]; by summing image-intensity values; Projection analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/50—Context or environment of the image
- G06V20/56—Context or environment of the image exterior to a vehicle by using sensors mounted on the vehicle
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/60—Type of objects
- G06V20/62—Text, e.g. of license plates, overlay texts or captions on TV images
- G06V20/625—License plates
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- General Physics & Mathematics (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- Bioinformatics & Computational Biology (AREA)
- General Engineering & Computer Science (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- Multimedia (AREA)
- Image Analysis (AREA)
- Image Processing (AREA)
Abstract
The invention relates to a car logo detection and identification method based on car logo relocation. Firstly, the color characteristics are adopted as the positioning characteristics of the license plate to position the license plate, and the vehicle logo can be roughly positioned through the positioned license plate position information. After the rough location range of the car logo is selected, the car logo is further located through a novel Canny operator-based car logo extraction method. After positioning is finished, because the shapes of the vehicle radiating nets are different, accurate positioning cannot be carried out on some conditions, the invention provides a concept based on vehicle logo repositioning, and the classification range of the vehicle logo is repositioned. Since the logo edge features are distinct, the present invention uses a histogram of gradient directions (HOG) feature as the classification feature. And calculating the gradient direction Histogram (HOG) features of the extracted car logo range, and sending the features into a Support Vector Machine (SVM) for training and classification. The invention can effectively detect the car logo image.
Description
Technical Field
The invention relates to the field of pattern recognition and computer vision, in particular to a car logo detection and recognition method based on car logo relocation.
Background
The intelligent traffic has the core idea that the automobile data is collected by using a sensor, a camera and the like, the traditional manual monitoring is replaced by using a computer-aided management mode, the data sharing and retrieval can be rapidly completed, and the effect of integrating traffic management is achieved. Among these, computer vision techniques play a crucial role. The task assigned to intelligent transportation is based on the acquisition of automobile data. For a car, the license plate and the logo are two important signs. Various researches on license plate recognition are carried out at home and abroad, and a large number of mature and stable systems are developed and widely applied to actual production scenes. But the research on the identification of the car logo is not focused, and the research is a loop in intelligent transportation. Therefore, the demand for identification of the emblem will be increasing in the future.
Disclosure of Invention
In view of this, the present invention is directed to a method for detecting and identifying a vehicle logo based on vehicle logo relocation, which can effectively locate and detect the vehicle logo.
The invention is realized by adopting the following scheme: a car logo detection and identification method based on car logo relocation comprises the following steps:
step S1: acquiring front images of different types of vehicles, renaming the front images to mark the categories of the vehicles, and dividing the front images into two parts according to a proportion, wherein 2/3 is used as a training set, and the rest 1/3 is used as a test set;
step S2: extracting license plate regions existing in the front images of the test set by adopting an HSV color space method, and detecting and screening the license plate regions by utilizing a morphological transformation algorithm;
step S3: determining a vehicle logo range through the license plate region screened in the step S2, and then acquiring a relocated vehicle logo region by using a relocation method;
step S4: and training a support vector machine, extracting a direction gradient histogram from the relocated car logo area, and classifying by using the trained support vector machine.
Further, the step S2 specifically includes the following steps:
step S21: converting the color space of the front image into HSV color space from RGB color space through a cvtColor function in an openCV image visual library;
step S22: the searching step S21 converts each pixel point in the image of the HSV color space, and if three channels of the current searched pixel point respectively satisfy: the hue channel, namely H, ranges from 90 to 150, the purity channel, namely S, ranges from 55 to 255, and the brightness channel, namely V, ranges from 30 to 255, the current pixel point is taken as 1, otherwise, the current pixel point is taken as 0, and finally a black-and-white image with the pixel values only being 0 and 1 is obtained;
step S23: the operator with the size of [4,4] is used for carrying out open operation on the black-and-white image obtained in the step S22, so that noise can be removed;
step S24: performing closed operation on the image with the noise removed in the step S23 by using an operator with the size of [20, 2], and connecting unfilled regions to obtain a possible region with a license plate;
step S25: performing an opening operation on the license plate possibility image obtained in the step S24 by using an operator with the size of [1, 5], so that the edge frizz of the license plate possibility area obtained in the step S24 can be removed;
step S26: processing the black-and-white image obtained in step S22 in step S23, step S24, and step S25 to finally obtain each connected region, that is, a license plate possibility region, and calculating an aspect ratio and an area of each connected region; according to the formula:
wherein WHi is the aspect ratio of the ith connected region and Area is the Area of the connected region; if the formula is satisfied, the license plate area is judged.
Further, the step S3 specifically includes the following steps:
step S31: and carrying out coarse positioning on the car logo, wherein the positioning relation is obtained by the following formula:
wherein lb, rb, tb and bb respectively correspond to the left, right, upper and lower boundaries of the vehicle logo coarse range; lp, rp, bp and tp respectively correspond to the left, right, upper and lower boundaries of the license plate;
step S32: processing the vehicle logo coarse range by adopting a Canny edge detection algorithm to obtain a vehicle logo coarse range edge image;
step S33: performing opening operation on the edge image of the vehicle logo coarse range obtained in the step S32 by using an operator with the size of [1, 2] so as to keep pixel points at the edge of the vehicle logo;
step S34: because some pixel points in the edges of the car logo reserved in the step S33 do not belong to the car logo, whether the pixel points reserved in the step S33 are car logo pixel points is determined by using a method for counting other pixel points in the neighborhood of the pixel; whether the number of reserved pixels in the neighborhood of 7 × 7 of each pixel point is more than 30 is counted; if the number is less than 30, the noise pixels are regarded as noise pixels and discarded; otherwise, reserving;
step S35: putting the pixel points finally reserved in the step S34 into the original positions of the pixel points to obtain images finally reserved, performing closing operation on the images finally reserved by using operators with the size of [15, 1], and combining the pixel points reserved in the step S34;
step S36: and repositioning the car logo area, wherein the formula is as follows:
wherein, nlb, nrb, ntb and nbb respectively correspond to the left and right upper and lower boundaries of the vehicle logo relocation range, and lp, rp, bp and tp respectively correspond to the left and right upper and lower boundaries of the license plate.
Further, the step S4 specifically includes the following steps:
step S41: carrying out size normalization on the relocated car logo range, and uniformly scaling to 64 × 64;
step S42: extracting a direction gradient histogram from the normalized relocated automobile logo range image by using an HOGDescriptor class of an openCV image visual library; wherein the block size is 16 × 16, the cell size is 8 × 8, the block sliding increment is 8 × 8, the gradient direction number is 9, and the finally extracted feature dimension is 1764 dimensions;
step S43: detecting and extracting car logos from the training sample images in the training set in the step S1, filtering the correctly extracted car logos to mark the car logos as the categories, and setting a negative category of a non-car logo type; the vector machine type is a C-type support vector machine, and a kernel function is set to be a POLY function; training a classifier by using the marked car logo;
step S44: and (4) reclassifying the training set in the step (S1) by using a trained classifier, marking the non-vehicle logo image of the false detection positive sample as a negative class, and feeding the negative class back to the classifier for reclassification so as to improve the classification effect.
Compared with the prior art, the invention has the following beneficial effects:
1. the invention can effectively detect the car logo image.
2. The Canny operator-based car logo background suppression method can effectively filter background information and extract car logo main bodies.
3. The invention provides a method for relocating a car logo to select the range of the car logo to be classified, which not only utilizes the information of the car logo, but also utilizes the background information of a vehicle cooling net.
4. Aiming at the characteristic of concentrated energy at the edge of the car logo, the invention adopts a histogram of gradient directions (HOG) algorithm to extract features. The HOG is an algorithm based on edge features, and the car logo usually has very obvious edge features, so that a histogram of gradient directions (HOG) algorithm is robust to a car logo identification task and is very suitable for use. The Support Vector Machine (SVM) classifier is adopted to match with the SVM classifier, and the achieved effect can exceed all similar algorithms.
Drawings
FIG. 1 is a flow chart of an embodiment of the present invention.
Detailed Description
The invention is further explained below with reference to the drawings and the embodiments.
As shown in fig. 1, the embodiment provides a car logo detection and identification method based on car logo relocation, which specifically includes the following steps:
step S1: acquiring front images of different types of vehicles, renaming the front images to mark the categories of the vehicles, and dividing the front images into two parts according to a proportion, wherein 2/3 is used as a training set, and the rest 1/3 is used as a test set;
step S2: extracting license plate regions existing in the front images of the test set by adopting an HSV color space method, and detecting and screening the license plate regions by utilizing a morphological transformation algorithm;
step S3: determining a vehicle logo range through the license plate region screened in the step S2, and then acquiring a relocated vehicle logo region by using a relocation method;
step S4: and training a support vector machine, extracting a direction gradient histogram from the relocated car logo area, and classifying by using the trained support vector machine.
In this embodiment, the step S1 specifically includes the following steps:
constructing a website crawler program, acquiring a large amount of vehicle data, and labeling different categories by image naming; the method comprises the steps of crawling front images of 7 types of vehicles from a website, renaming the images, wherein the file name of the ith type of image starts with a number i, so that subsequent classification is facilitated;
in this embodiment, step S2 specifically includes the following steps:
step S21: converting the color space of the front image containing the car logo from the RGB color space into HSV color space through a cvtColor function in an openCV image visual library;
step S22: the searching step S21 converts each pixel point in the image of the HSV color space, and if three channels of the current searched pixel point respectively satisfy: the hue channel, namely H, ranges from 90 to 150, the purity channel, namely S, ranges from 55 to 255, and the brightness channel, namely V, ranges from 30 to 255, the current pixel point is taken as 1, otherwise, the current pixel point is taken as 0, and finally a black-and-white image with the pixel values only being 0 and 1 is obtained;
step S23: the operator with the size of [4,4] is used for carrying out open operation on the black-and-white image obtained in the step S22, so that noise can be removed;
step S24: performing closed operation on the image with the noise removed in the step S23 by using an operator with the size of [20, 2], and connecting unfilled regions to obtain a region with possibility of existence of a license plate;
step S25: performing an opening operation on the license plate possibility image obtained in the step S24 by using an operator with the size of [1, 5], so that the edge frizz of the license plate possibility area obtained in the step S24 can be removed;
step S26: processing the black-and-white image obtained in step S22 in step S23, step S24, and step S25 to finally obtain each connected region, that is, a license plate possibility region, and calculating an aspect ratio and an area of each connected region; according to the formula:
wherein WHi is the aspect ratio of the ith connected region and Area is the Area of the connected region; if the formula is satisfied, the license plate area is judged.
In this embodiment, step S3 specifically includes the following steps:
step S31: the car logo is roughly positioned, and the positioning relation is obtained by the following formula:
wherein lb, rb, tb and bb respectively correspond to the left, right, upper and lower boundaries of the vehicle logo coarse range; lp, rp, bp and tp respectively correspond to the left, right, upper and lower boundaries of the license plate;
step S32: processing the vehicle logo coarse range by adopting a Canny edge detection algorithm to obtain a vehicle logo coarse range edge image;
step S33: performing opening operation on the edge image of the vehicle logo coarse range obtained in the step S32 by using an operator with the size of [1, 2] so as to keep pixel points at the edge of the vehicle logo;
step S34: because some pixel points in the edges of the car logo reserved in the step S33 do not belong to the car logo, the invention determines whether the pixel points reserved in the step S33 are car logo pixel points by using a method for counting other pixel points in the neighborhood of the pixel; whether the number of reserved pixels in the neighborhood of 7 × 7 of each pixel point is more than 30 is counted; if the number is less than 30, the noise pixels are regarded as noise pixels and discarded; otherwise, reserving;
step S35: placing the pixel points finally reserved in the step S34 into the original positions of the pixel points to obtain images finally reserved, performing closed operation on the images finally reserved by using operators with the size of [15, 1], and combining the pixel points reserved in the step S34;
step S36: and repositioning the car logo area, wherein the formula is as follows:
wherein, nlb, nrb, ntb and nbb respectively correspond to the left and right upper and lower boundaries of the vehicle logo relocation range, and lp, rp, bp and tp respectively correspond to the left and right upper and lower boundaries of the license plate.
In this embodiment, step S4 specifically includes the following steps:
step S41: normalizing the size of the relocated car logo range, and uniformly scaling the size to 64 x 64;
step S42: extracting a direction gradient histogram from the normalized relocated automobile logo range image by using an HOGDescriptor class of an openCV image visual library; wherein the block size is 16 × 16, the cell size is 8 × 8, the block sliding increment is 8 × 8, the gradient direction number is 9, and the finally extracted feature dimension is 1764 dimensions;
step S43: detecting and extracting car logos from the training sample images in the training set in the step S1, filtering the correctly extracted car logos to mark the car logos as the categories, and setting a negative category of a non-car logo type; the type of the used vector machine is a C-type support vector machine, and a kernel function is set to be a POLY function; training a classifier by using the marked car logo;
step S44: and (4) reclassifying the training set in the step (S1) by using a trained support vector machine classifier, marking the non-vehicle logo images of the false detection positive samples as negative classes, and feeding the non-vehicle logo images back to the classifier for reclassification so as to improve the classification effect.
The embodiment can effectively position and detect the car logo. Firstly, an HSV color space is used for filtering a license plate region, and the model accords with the intuitive understanding of human colors. By using the model, the color extraction can be conveniently carried out. After the license plate range is extracted, coarse range positioning is carried out by using an empirical formula. Then, within the extracted coarse range, an algorithm based on Canny operator is used for further positioning. After the positioning is completed, the embodiment provides a method for selecting the range of the car logo to be classified based on the car logo relocation, and the method not only utilizes the information of the car logo, but also utilizes the background information of the vehicle cooling net. And finally, extracting a histogram of gradient directions (HOG), and classifying by a Support Vector Machine (SVM). In the embodiment, a feedback-based mode is adopted for optimizing the classification result, so that a certain effect is achieved.
The above description is only a preferred embodiment of the present invention, and all equivalent changes and modifications made in accordance with the claims of the present invention should be covered by the present invention.
Claims (2)
1. A car logo detection and identification method based on car logo relocation is characterized in that: the method comprises the following steps:
step S1: acquiring front images of different types of vehicles, renaming the front images to mark the categories of the vehicles, and dividing the front images into two parts according to a proportion, wherein 2/3 is used as a training set, and the rest 1/3 is used as a test set;
step S2: extracting license plate regions existing in the front images of the test set by adopting an HSV color space method, and detecting and screening the license plate regions by utilizing a morphological transformation algorithm;
step S3: determining a vehicle logo range through the license plate region screened in the step S2, and then acquiring a relocated vehicle logo region by using a relocation method;
step S4: training a support vector machine, extracting a direction gradient histogram from the relocated car logo area, and classifying by using the trained support vector machine;
the step S2 specifically includes the following steps:
step S21: converting the color space of the front image into HSV color space from RGB color space through a cvtColor function in an openCV image visual library;
step S22: the searching step S21 converts each pixel point in the image of the HSV color space, and if three channels of the current searched pixel point respectively satisfy: the hue channel, namely H, ranges from 90 to 150, the purity channel, namely S, ranges from 55 to 255, and the brightness channel, namely V, ranges from 30 to 255, the current pixel point is taken as 1, otherwise, the current pixel point is taken as 0, and finally a black-and-white image with the pixel values only being 0 and 1 is obtained;
step S23: opening operation is carried out on the black-white image obtained in the step S22 by using an operator with the size of [4,4], and noise is removed;
step S24: performing closed operation on the image with the noise removed in the step S23 by using an operator with the size of [20, 2], and connecting unfilled areas to obtain a possibility area with a license plate;
step S25: performing opening operation on the license plate possibility image obtained in the step S24 by using an operator with the size of [1, 5], and removing the edge frizzy of the license plate possibility area obtained in the step S24;
step S26: processing the black-and-white image obtained in step S22 in step S23, step S24, and step S25 to finally obtain each connected region, that is, a license plate possibility region, and calculating an aspect ratio and an area of each connected region; according to the formula:
wherein WHi is the aspect ratio of the ith connected region and Area is the Area of the connected region; if the formula is satisfied, judging the license plate area;
the step S3 specifically includes the following steps:
step S31: the car logo is roughly positioned, and the positioning relation is obtained by the following formula:
wherein lb, rb, tb, bb respectively corresponds to the left, right, upper and lower boundaries of the vehicle logo thick range; lp, rp, bp and tp respectively correspond to the left, right, upper and lower boundaries of the license plate;
step S32: processing the vehicle logo coarse range by adopting a Canny edge detection algorithm to obtain a vehicle logo coarse range edge image;
step S33: performing opening operation on the edge image of the vehicle logo coarse range obtained in the step S32 by using an operator with the size of [1, 2] so as to keep pixel points of the vehicle logo edge;
step S34: because some pixel points in the edges of the car logo reserved in the step S33 do not belong to the car logo, whether the pixel points reserved in the step S33 are car logo pixel points is determined by using a method for counting other pixel points in the neighborhood of the pixel; whether the number of reserved pixels in the neighborhood of 7 × 7 of each pixel point is more than 30 is counted; if the number is less than 30, the noise pixels are regarded as noise pixels and discarded; otherwise, reserving;
step S35: putting the pixel points finally reserved in the step S34 into original positions of the pixel points to obtain images finally reserved, performing closing operation on the images finally reserved by using operators with the size of [15, 1], and combining the pixel points reserved in the step S34;
step S36: and repositioning the car logo area, wherein the formula is as follows:
wherein, nlb, nrb, ntb and nbb respectively correspond to the left and right upper and lower boundaries of the vehicle logo relocation range, and lp, rp, bp and tp respectively correspond to the left and right upper and lower boundaries of the license plate.
2. The emblem detection and identification method based on emblem relocation of claim 1, wherein: the step S4 specifically includes the following steps:
step S41: carrying out size normalization on the relocated car logo range, and uniformly scaling to 64 × 64;
step S42: extracting a direction gradient histogram from the normalized relocated automobile logo range image by using an HOGDescriptor class of an openCV image visual library; wherein the block size is 16 × 16, the cell size is 8 × 8, the block sliding increment is 8 × 8, the gradient direction number is 9, and the finally extracted feature dimension is 1764 dimensions;
step S43: detecting and extracting car logos from the training sample images in the training set in the step S1, filtering the correctly extracted car logos to mark the car logos as the categories, and setting a negative category of a non-car logo type; the vector machine type is a C-type support vector machine, and a kernel function is set to be a POLY function; training a classifier by using the marked car logo;
step S44: and (4) reclassifying the training set in the step (S1) by using a trained classifier, marking the non-vehicle logo images of the false detection positive samples as negative classes, and feeding the negative classes back to the classifier for reclassification.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910331226.2A CN110032991B (en) | 2019-04-23 | 2019-04-23 | Car logo detection and identification method based on car logo relocation |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910331226.2A CN110032991B (en) | 2019-04-23 | 2019-04-23 | Car logo detection and identification method based on car logo relocation |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110032991A CN110032991A (en) | 2019-07-19 |
| CN110032991B true CN110032991B (en) | 2022-07-08 |
Family
ID=67239950
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910331226.2A Active CN110032991B (en) | 2019-04-23 | 2019-04-23 | Car logo detection and identification method based on car logo relocation |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110032991B (en) |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105426896A (en) * | 2015-11-16 | 2016-03-23 | 成都神州数码索贝科技有限公司 | Car logo automatic identification method and system |
| CN106503710A (en) * | 2016-10-26 | 2017-03-15 | 北京邮电大学 | A kind of automobile logo identification method and device |
| CN106529424A (en) * | 2016-10-20 | 2017-03-22 | 中山大学 | Vehicle logo recognition method and system based on selective search algorithm |
| CN107590492A (en) * | 2017-08-28 | 2018-01-16 | 浙江工业大学 | A kind of vehicle-logo location and recognition methods based on convolutional neural networks |
-
2019
- 2019-04-23 CN CN201910331226.2A patent/CN110032991B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105426896A (en) * | 2015-11-16 | 2016-03-23 | 成都神州数码索贝科技有限公司 | Car logo automatic identification method and system |
| CN106529424A (en) * | 2016-10-20 | 2017-03-22 | 中山大学 | Vehicle logo recognition method and system based on selective search algorithm |
| CN106503710A (en) * | 2016-10-26 | 2017-03-15 | 北京邮电大学 | A kind of automobile logo identification method and device |
| CN107590492A (en) * | 2017-08-28 | 2018-01-16 | 浙江工业大学 | A kind of vehicle-logo location and recognition methods based on convolutional neural networks |
Non-Patent Citations (3)
| Title |
|---|
| VEHICLE LOGO RECOGNITION USING ENHANCED LEARNING : VLR Recognition;S. Gopinathan et al.;《Proceedings of the Second International conference on I-SMAC》;20190228;第707-712页 * |
| Vehicle Logo Recognition with Small Sample Problem in Complex Scene Based on Data Augmentation;Ke,Xiao et al.;《Mathematical Problems in Engineering》;20200709;第1-10页 * |
| 基于图像处理与机器学习的车标及安全带识别研究;莫文英;《中国优秀博硕士学位论文全文数据库(硕士)信息科技辑》;20190115(第01期);第1-70页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN110032991A (en) | 2019-07-19 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN105373794B (en) | A kind of licence plate recognition method | |
| WO2020173022A1 (en) | Vehicle violation identifying method, server and storage medium | |
| CN103530366B (en) | A kind of vehicle search method and system based on user-defined feature | |
| CN103034836B (en) | Road sign detection method and road sign checkout equipment | |
| CN107016362B (en) | Vehicle weight recognition method and system based on vehicle front windshield pasted mark | |
| CN110837807A (en) | Identification method and system for fake-licensed vehicle | |
| CN102799859B (en) | Method for identifying traffic sign | |
| CN106650553A (en) | License plate recognition method and system | |
| CN105740886B (en) | A kind of automobile logo identification method based on machine learning | |
| CN103544484A (en) | Traffic sign identification method and system based on SURF | |
| CN106709412B (en) | Traffic sign detection method and device | |
| CN106980855B (en) | Traffic sign rapid identification and positioning system and method | |
| CN101436252A (en) | Method and system for recognizing vehicle body color in vehicle video image | |
| CN105184291B (en) | A kind of polymorphic type detection method of license plate and system | |
| CN106203420B (en) | A kind of bayonet vehicle color identification method | |
| CN102902957A (en) | Video-stream-based automatic license plate recognition method | |
| CN104636754A (en) | Intelligent image classifying method based on tongue body partition color feature | |
| CN102542244A (en) | Face detection method and system and computer program product | |
| CN107590500A (en) | A kind of color recognizing for vehicle id method and device based on color projection classification | |
| CN104463134A (en) | License plate detection method and system | |
| CN105787475A (en) | Traffic sign detection and identification method under complex environment | |
| CN108664969A (en) | Landmark identification method based on condition random field | |
| CN104408431A (en) | Vehicle model identification method under traffic monitoring | |
| CN104268509B (en) | The method and system of dump truck car plate detection | |
| CN104008404A (en) | Pedestrian detection method and system based on significant histogram features |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |