CN110705548A - Coarse-to-fine license plate detection algorithm and system thereof - Google Patents
Coarse-to-fine license plate detection algorithm and system thereof Download PDFInfo
- Publication number
- CN110705548A CN110705548A CN201910848256.0A CN201910848256A CN110705548A CN 110705548 A CN110705548 A CN 110705548A CN 201910848256 A CN201910848256 A CN 201910848256A CN 110705548 A CN110705548 A CN 110705548A
- Authority
- CN
- China
- Prior art keywords
- license plate
- detected
- module
- detecting
- region
- 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
- 238000001514 detection method Methods 0.000 title claims abstract description 52
- 238000005070 sampling Methods 0.000 claims abstract description 19
- 238000012549 training Methods 0.000 claims abstract description 10
- 238000012544 monitoring process Methods 0.000 claims description 6
- 238000000034 method Methods 0.000 claims description 4
- 238000012545 processing Methods 0.000 claims description 2
- 238000004364 calculation method Methods 0.000 abstract description 4
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000011897 real-time detection Methods 0.000 description 1
- 230000004044 response Effects 0.000 description 1
Images
Classifications
-
- 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
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T3/00—Geometric image transformations in the plane of the image
- G06T3/40—Scaling of whole images or parts thereof, e.g. expanding or contracting
- G06T3/4023—Scaling of whole images or parts thereof, e.g. expanding or contracting based on decimating pixels or lines of pixels; based on inserting pixels or lines of pixels
-
- 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)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- Artificial Intelligence (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Multimedia (AREA)
- Traffic Control Systems (AREA)
Abstract
The invention discloses a license plate detection algorithm from coarse to fine and a system thereof in the field of computer vision, wherein the algorithm comprises the following specific steps: s1: down-sampling the input image based on a minimum vehicle size to be detected; s2: detecting a vehicle region for the down-sampled image using a convolution network one; s3: based on the size of the minimum license plate to be detected, down-sampling the detected vehicle area image; s4: detecting a candidate license plate region of the downsampled vehicle region image by using a convolution network II; s5: for the candidate license plate area detected in the step S4, the resolution is increased in proportion, a convolutional network is used for detecting the license plate, and coordinates of 4 corner points of the license plate are output, wherein the system comprises the following steps: the system comprises a data acquisition module, a license plate detection module, an output module and a model training module, wherein the used convolution network structure is relatively simple, the required calculation amount is far less than that of the existing license plate detection algorithm, and the system can be effectively applied to scenes with high requirements on the real-time performance of license plate detection, such as high-speed parking-free charging and the like.
Description
Technical Field
The invention relates to the technical field of computer vision, in particular to a license plate detection algorithm from coarse to fine and a system thereof.
Background
At present, license plate detection and recognition are widely applied to the fields of parking lots, traffic crossing violation detection and the like. With the gradual development and application of smart cities, freeway toll collection has become a trend. The application scene has high requirements on the real-time performance of license plate detection and recognition. If the detection time is too long, the automobile still needs to wait for the system response, the aim of quick passing without stopping can not be achieved, and traffic jam is easily caused.
At present, most of the mainstream license plate detection algorithms are used for detecting license plates by using a deep convolution network on an input image, in order to ensure the detection accuracy, the network is generally complex, the calculation time required by detection is more, and the real-time requirement under the condition of quick passing is difficult to achieve.
Based on the above problems, the present invention provides a license plate detection algorithm from coarse to fine and a system thereof, so as to solve the problems in the background art.
Disclosure of Invention
The invention aims to provide a coarse-to-fine license plate detection algorithm and a system thereof, the method respectively detects vehicles and license plates based on different resolutions, and uses a complex network only in a smaller image range, so that the calculated amount can be effectively reduced, the real-time detection effect is achieved, and the problems provided in the background technology are solved.
In order to achieve the purpose, the invention provides the following technical scheme: a coarse-to-fine license plate detection algorithm comprises the following specific steps:
s1: down-sampling the input image based on a minimum vehicle size to be detected;
s2: detecting a vehicle region for the down-sampled image using a convolution network one;
s3: based on the size of the minimum license plate to be detected, down-sampling the detected vehicle area image;
s4: detecting a candidate license plate region of the downsampled vehicle region image by using a convolution network II;
s5: and for the candidate license plate region detected in the step S4, the resolution is increased in proportion, a convolutional network is used for detecting the license plate, whether the candidate region contains the license plate or not is determined, and if yes, coordinates of 4 corner points of the license plate are output.
Preferably, the first convolutional network, the second convolutional network and the third convolutional network are all selected as shallow convolutional networks.
Preferably, the resolution of the down-sampling of the input image in step S1 is determined according to the minimum vehicle size that needs to be detected.
Preferably, in step S3, the resolution of the down-sampling of the detected vehicle region image is determined according to the minimum license plate size that needs to be detected.
A coarse-to-fine license plate detection system comprises
The data acquisition module acquires a monitoring video and inputs the monitoring video into the license plate detection module;
the license plate detection module is used for processing the image by using the trained model to obtain a detection result;
the output module outputs the detected license plate and coordinates of 4 corner points corresponding to the license plate as input of subsequent application;
and the model training module is used for training the model used in the license plate detection module by using the training set data.
Preferably, the license plate detection module comprises a vehicle detection sub-module, a license plate pre-detection sub-module and a license plate detection sub-module, and detects the region, the candidate license plate region and the final license plate in the resolution from low to high respectively.
Compared with the prior art, the invention has the beneficial effects that: the method comprises the steps of firstly detecting a vehicle area on a low-resolution image, and then gradually increasing the resolution in the vehicle area to detect the license plate and the position of the license plate.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the description of the embodiments will be briefly introduced below, and it is obvious that the drawings in the following description are only some embodiments of the present invention, and it is obvious for those skilled in the art that other drawings can be obtained according to the drawings without creative efforts.
FIG. 1 is a flow chart of the detection method of the present invention;
FIG. 2 is a schematic block diagram of the detection system of the present invention.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
Referring to fig. 1-2, the present invention provides a technical solution: a coarse-to-fine license plate detection algorithm comprises the following specific steps:
s1: the input image is firstly down-sampled to a lower resolution ratio through a first down-sampling module, and the down-sampling ratio is determined according to the minimum vehicle size required to be detected;
s2: detecting a vehicle region for the down-sampled image using a convolution network one;
s3: for the image of the vehicle area, the second down-sampling module is used for down-sampling, wherein the down-sampling proportion is determined according to the minimum license plate size needing to be detected and is specified by a user, such as
Recording the input data size of the convolution network I as s multiplied by s, and the minimum vehicle size to be detected as minisize multiplied by minisize, then the input image is proportioned
S4: inputting the low-resolution vehicle area image into a convolution network II to detect a candidate license plate area;
s5: and for the candidate license plate region detected in the step S4, the resolution is increased in proportion, a convolutional network is used for detecting the license plate, whether the candidate region contains the license plate or not is determined, and if yes, coordinates of 4 corner points of the license plate are output.
In the algorithm, the first down-sampling module and the second down-sampling module can adopt simple down-sampling or learning down-sampling based on a network.
In the algorithm, a network structure of a shallow layer can be selected and used for the first convolutional network, the second convolutional network and the third convolutional network, and compared with a common target detection network, the network structure is simpler and the required calculation amount is less. And because the downsampling and detection range reduction are used, the input images of the networks are small, and the whole calculation amount is greatly reduced compared with that of a conventional detection network.
A coarse-to-fine license plate detection system, the system comprising:
and the data acquisition module acquires the monitoring video and inputs the monitoring video into the license plate detection module.
And the license plate detection module processes the image by using the trained model to obtain a detection result. The license plate detection module of the system comprises three submodules of vehicle detection, license plate pre-detection and license plate detection, and is used for detecting a vehicle region, a candidate license plate region and finishing final license plate detection on the resolution from low to high respectively.
And the output module outputs the detected license plate and the coordinates of the 4 corner points corresponding to the license plate as the input of subsequent application.
And the model training module is used for training the model used in the license plate detection module by using the training set data.
In the description herein, references to the description of "one embodiment," "an example," "a specific example" or the like are intended to mean that a particular feature, structure, material, or characteristic described in connection with the embodiment or example is included in at least one embodiment or example of the invention. In this specification, the schematic representations of the terms used above do not necessarily refer to the same embodiment or example. Furthermore, the particular features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments or examples.
The preferred embodiments of the invention disclosed above are intended to be illustrative only. The preferred embodiments are not intended to be exhaustive or to limit the invention to the precise embodiments disclosed. Obviously, many modifications and variations are possible in light of the above teaching. The embodiments were chosen and described in order to best explain the principles of the invention and the practical application, to thereby enable others skilled in the art to best utilize the invention. The invention is limited only by the claims and their full scope and equivalents.
Claims (6)
1. A coarse-to-fine license plate detection algorithm is characterized in that: the method comprises the following specific steps:
s1: down-sampling the input image based on a minimum vehicle size to be detected;
s2: detecting a vehicle region for the down-sampled image using a convolution network one;
s3: based on the size of the minimum license plate to be detected, down-sampling the detected vehicle area image;
s4: detecting a candidate license plate region of the downsampled vehicle region image by using a convolution network II;
s5: and for the candidate license plate region detected in the step S4, the resolution is increased in proportion, a convolutional network is used for detecting the license plate, whether the candidate region contains the license plate or not is determined, and if yes, coordinates of 4 corner points of the license plate are output.
2. The algorithm for detecting license plates from coarse to fine according to claim 1, wherein: and the first convolution network, the second convolution network and the third convolution network are all selected as shallow layer convolution networks.
3. The algorithm for detecting license plates from coarse to fine according to claim 1, wherein: the resolution of the down-sampling of the input image in step S1 is determined according to the minimum vehicle size that needs to be detected.
4. The algorithm for detecting license plates from coarse to fine according to claim 1, wherein: in step S3, the resolution of the down-sampling of the detected vehicle region image is determined according to the minimum license plate size that needs to be detected.
5. A thick-to-fine license plate detection system is characterized in that: the system comprises
The data acquisition module acquires a monitoring video and inputs the monitoring video into the license plate detection module;
the license plate detection module is used for processing the image by using the trained model to obtain a detection result;
the output module outputs the detected license plate and coordinates of 4 corner points corresponding to the license plate as input of subsequent application;
and the model training module is used for training the model used in the license plate detection module by using the training set data.
6. The system of claim 5, wherein the system is configured to perform the following steps: the license plate detection module comprises a vehicle detection sub-module, a license plate pre-detection sub-module and a license plate detection sub-module, and is used for detecting the region, the candidate license plate region and the final license plate in the resolution from low to high respectively.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910848256.0A CN110705548A (en) | 2019-09-09 | 2019-09-09 | Coarse-to-fine license plate detection algorithm and system thereof |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910848256.0A CN110705548A (en) | 2019-09-09 | 2019-09-09 | Coarse-to-fine license plate detection algorithm and system thereof |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN110705548A true CN110705548A (en) | 2020-01-17 |
Family
ID=69194813
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910848256.0A Pending CN110705548A (en) | 2019-09-09 | 2019-09-09 | Coarse-to-fine license plate detection algorithm and system thereof |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110705548A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113409502A (en) * | 2021-06-28 | 2021-09-17 | 杭州海康威视数字技术股份有限公司 | Method, device and system for controlling vehicle to pass and electronic equipment |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107506763A (en) * | 2017-09-05 | 2017-12-22 | 武汉大学 | A kind of multiple dimensioned car plate precise positioning method based on convolutional neural networks |
| CN108428248A (en) * | 2018-03-14 | 2018-08-21 | 苏州科达科技股份有限公司 | Vehicle window localization method, system, equipment and storage medium |
| CN109255350A (en) * | 2018-08-29 | 2019-01-22 | 南京邮电大学 | A kind of new energy detection method of license plate based on video monitoring |
| CN109815906A (en) * | 2019-01-25 | 2019-05-28 | 华中科技大学 | Method for traffic sign detection and system based on substep deep learning |
| CN110020651A (en) * | 2019-04-19 | 2019-07-16 | 福州大学 | Car plate detection localization method based on deep learning network |
-
2019
- 2019-09-09 CN CN201910848256.0A patent/CN110705548A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107506763A (en) * | 2017-09-05 | 2017-12-22 | 武汉大学 | A kind of multiple dimensioned car plate precise positioning method based on convolutional neural networks |
| CN108428248A (en) * | 2018-03-14 | 2018-08-21 | 苏州科达科技股份有限公司 | Vehicle window localization method, system, equipment and storage medium |
| CN109255350A (en) * | 2018-08-29 | 2019-01-22 | 南京邮电大学 | A kind of new energy detection method of license plate based on video monitoring |
| CN109815906A (en) * | 2019-01-25 | 2019-05-28 | 华中科技大学 | Method for traffic sign detection and system based on substep deep learning |
| CN110020651A (en) * | 2019-04-19 | 2019-07-16 | 福州大学 | Car plate detection localization method based on deep learning network |
Non-Patent Citations (3)
| Title |
|---|
| SÉRGIO MONTAZZOLLI，ET AL.: "License Plate Detection and Recognition in Unconstrained Scenarios", 《ECCV 2018》 * |
| SÉRGIO MONTAZZOLLI，ET AL.: "Real-Time Brazilian License Plate Detection and Recognition Using Deep Convolutional Neural Networks", 《2017 30TH SIBGRAPI CONFERENCE ON GRAPHICS, PATTERNS AND IMAGES (SIBGRAPI)》 * |
| 孙康，等.: "基于级联卷积神经网络的人脸检测算法", 《南京理工大学学报》 * |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113409502A (en) * | 2021-06-28 | 2021-09-17 | 杭州海康威视数字技术股份有限公司 | Method, device and system for controlling vehicle to pass and electronic equipment |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN112329776B (en) | License plate detection method and device based on improved CenterNet network | |
| CN102902983B (en) | A kind of taxi identification method based on support vector machine | |
| KR101163042B1 (en) | Apparatus and Method for Extracting Vehicle | |
| CN101714213B (en) | Vehicle-mounted pedestrian detection method and system | |
| CN109583442B (en) | False license plate detection method and device based on line segment detection | |
| Xiang et al. | Lightweight fully convolutional network for license plate detection | |
| CN112991349B (en) | Image processing method, device, equipment and storage medium | |
| CN111127516A (en) | Target detection and tracking method and system without search box | |
| CN115376089A (en) | Deep learning-based lane line detection method | |
| CN111382722A (en) | License plate image optimization method, image processing device and device with storage function | |
| US10043067B2 (en) | System and method for detecting pedestrians using a single normal camera | |
| Liang et al. | Car detection and classification using cascade model | |
| Kumar et al. | An efficient approach for automatic number plate recognition for low resolution images | |
| Santos et al. | Car recognition based on back lights and rear view features | |
| CN110705548A (en) | Coarse-to-fine license plate detection algorithm and system thereof | |
| Yang | Estimation of vehicle's lateral position via the Lucas-Kanade optical flow method | |
| CN117115690A (en) | Unmanned aerial vehicle traffic target detection method and system based on deep learning and shallow feature enhancement | |
| CN112241963A (en) | Lane line identification method and system based on vehicle-mounted video and electronic equipment | |
| CN116958811A (en) | Road ponding area detection method, system, equipment and medium | |
| CN114882393B (en) | Road reverse running and traffic accident event detection method based on target detection | |
| CN101577052B (en) | Device and method for detecting vehicles by overlooking | |
| CN111985267B (en) | Multi-target detection method, device, computer equipment, vehicle and storage medium | |
| CN109614841B (en) | Rapid face detection method in embedded system | |
| Jiao et al. | A robust multi-class traffic sign detection and classification system using asymmetric and symmetric features | |
| CN117789144B (en) | Cross network lane line detection method and device based on weight fusion |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20200117 |