Unlock instant, AI-driven research and patent intelligence for your innovation.

A traffic sign recognition method based on multi-attribute joint features

A technology of traffic sign recognition and joint feature, applied in the field of traffic sign recognition in intelligent transportation system, can solve the problem of ignoring learning speed and other problems

Active Publication Date: 2021-07-06
南京龙盾智能科技有限公司
View PDF6 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0004] At present, most CNN research focuses on improving the classification accuracy, while ignoring the learning speed. Traffic sign recognition just needs to increase the learning speed to ensure real-time requirements.

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 more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • A traffic sign recognition method based on multi-attribute joint features
  • A traffic sign recognition method based on multi-attribute joint features
  • A traffic sign recognition method based on multi-attribute joint features

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0027] The traffic sign recognition method based on multi-attribute joint features of the present invention will be further described below in conjunction with the accompanying drawings.

[0028] Such as figure 1 As shown, the traffic sign recognition method of the present invention comprises the following steps:

[0029] Step 1: database image preprocessing;

[0030] The traffic sign images used in this embodiment mainly come from the GTSRB data set and the traffic sign images in natural scenes captured by smart cameras. There are a total of 5000 training samples and 1000 test samples, including 43 types of traffic signs.

[0031] Firstly, the image size is uniformly adjusted to 48×48 pixels, and normalized. The training samples need to carry labels, and the labeled training samples are expressed as (x i ,t i ), i=1,2,...,N, N represents the number of training samples; x i Represents the feature vector of the i-th sample, t i A vector of labels representing the i-th sam...

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

No PUM Login to View More

Abstract

The invention relates to a traffic sign recognition method based on multi-attribute joint features. First, the image is preprocessed, then a convolutional neural network structure is designed, and traffic sign features are automatically extracted through the CNN network. information, extract the feature maps of the last three layers, and use the extracted feature maps of each layer to form three-scale feature matrices through multi-scale pooling operations, expand the three-scale feature matrices by columns and concatenate them into column vectors; The obtained three column vectors are cascaded into a joint feature vector with multi-scale and multi-attribute; finally, the joint feature vector is classified by the ELM classifier, so as to efficiently complete the recognition and classification of traffic signs.

Description

technical field [0001] The invention relates to a traffic sign recognition method based on multi-attribute joint features, belonging to the field of traffic sign recognition in intelligent transportation systems. Background technique [0002] In recent years, traffic sign recognition has been widely used in driver assistance systems, driverless smart cars, and road maintenance. Traditional traffic sign recognition methods are difficult to meet the requirements of high accuracy and real-time performance. [0003] The traffic sign recognition method based on deep learning has become a research hotspot in the past two years. For example, the convolutional neural network (CNN) has been successfully applied to the traffic sign recognition system, but usually the last layer of CNN is used for classifier training. However, these features may not contain enough useful information to realize the classification of traffic signs. Therefore, if we can make full use of the features extr...

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
no application Login to View More
Patent Type & Authority Patents(China)
IPC IPC(8): G06K9/00G06K9/62G06N3/04
CPCG06V20/582G06N3/045G06F18/2155
Inventor 孙伟杜宏吉张小瑞赵玉舟施顺顺杨翠芳
Owner 南京龙盾智能科技有限公司