SSD-based license plate detection and correction recognition method and recognition system
A license plate detection and recognition system technology, applied in the field of computer vision, can solve the problems of reducing the recognition accuracy and achieve the effect of improving the accuracy
Pending Publication Date: 2020-09-22
南京博雅集智智能技术有限公司
View PDF4 Cites 1 Cited by
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
[0003] At present, vehicle detection and recognition mainly adopts the method of first extracting the license plate information in the vehicle, and then recognizing the license plate characters. When the traditional license plate recognition performs license plate detection, the output result after detection is directly sent to the subsequent recognition. , in the face of regular scenes, it can get good results, and can solve the problem well in general scenes, but in the case of serious license plate tilt, for example, the license plate may be detected, but the detected license plate is not is inclined, which will greatly reduce the accuracy of subsequent recognition
Method used
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View moreImage
Smart Image Click on the blue labels to locate them in the text.
Smart ImageViewing Examples
Examples
Experimental program
Comparison scheme
Effect test
Embodiment Construction
[0040] The present invention will be further described below in conjunction with the accompanying drawings.
[0041] The SSD (Single Shot MultiBox Detector) algorithm belongs to the one-stage method of multi-frame prediction. It uses CNN to directly detect the target, which solves the problem of difficult detection of small targets and inaccurate positioning to a certain extent. The core idea of the SSD algorithm is as follows:
[0042] 1. Using multi-scale feature maps for detection
[0043] Multi-scale refers to the use of feature maps of different sizes. Generally, the front feature map of the CNN network is relatively large, and the convolution or pool of stride=2 will be gradually used later to reduce the size of the feature map. For example, a relatively large feature map and a relatively small one. Feature maps can be used for detection. The advantage of multi-scale is that relatively large feature maps can be used to detect relatively small objects, while small fea...
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More PUM
Login to View More Abstract
The invention discloses an SSD-based license plate detection and correction recognition method and recognition system, and the method comprises the steps: inputting an image, and detecting whether a license plate, the position of the license plate and the position of a key point on the license plate exist in the image or not; judging whether the license plate is inclined or not, and correcting theposition of the license plate through affine transformation of the key points if the license plate is inclined; recognizing the license plate image which is not inclined or the corrected license plate image, detecting whether characters exist in the license plate image or not and the positions of the characters, and outputting a recognition result. The system comprises a license plate and licenseplate key point detection module, a correction module and a license plate character detection module which are connected in sequence; a to-be-detected image is input into the license plate and license plate key point detection module, and the license plate character detection module outputs character information on the license plate. The invention provides an end-to-end license plate detection, correction and license plate character recognition method on the basis of an SSD algorithm, and the accuracy of license plate recognition can be improved.
Description
technical field [0001] The invention belongs to the technical field of computer vision, and in particular relates to an SSD-based license plate detection and correction recognition method and recognition system. Background technique [0002] Vehicle detection and recognition is an important part of modern intelligent transportation system, a good vehicle detection and recognition system can greatly alleviate the increasingly severe traffic pressure. [0003] At present, vehicle detection and recognition mainly adopts the method of first extracting the license plate information in the vehicle, and then recognizing the license plate characters. When the traditional license plate recognition performs license plate detection, the output result after detection is directly sent to the subsequent recognition. , in the face of regular scenes, it can get good results, and can solve the problem well in general scenes, but in the case of serious license plate tilt, for example, the lic...
Claims
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More Application Information
Patent Timeline
Login to View More Patent Type & Authority Applications(China)
IPC IPC(8): G06K9/32G06K9/34G06N3/04G06N3/08
CPCG06N3/08G06V30/1478G06V30/153G06V20/625G06N3/045
Inventor 孙超邢卫国施远银鞠蓉
Owner 南京博雅集智智能技术有限公司