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

Traffic sign recognition method based on LDCNN model and NHE algorithm

A technology of traffic sign recognition and traffic sign, which is applied in the field of traffic sign recognition, and can solve the problems that there is no original image enhancement processing, and the recognition accuracy cannot be improved.

Inactive Publication Date: 2019-03-19
GUILIN UNIV OF ELECTRONIC TECH
View PDF2 Cites 4 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

The above traffic sign recognition algorithms all achieve high recognition accuracy, but there are still two problems. One is that the accuracy of traffic sign recognition needs to be improved, and the other is that neither the traditional algorithm nor the CNN algorithm has enhanced the original image. It is also the reason why the recognition accuracy cannot be improved.

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
  • Traffic sign recognition method based on LDCNN model and NHE algorithm
  • Traffic sign recognition method based on LDCNN model and NHE algorithm
  • Traffic sign recognition method based on LDCNN model and NHE algorithm

Examples

Experimental program
Comparison scheme
Effect test

Embodiment

[0047] refer to figure 1 , a traffic sign recognition method based on the LDCNN model and the NHE algorithm, comprising the following steps:

[0048] 1) GTSRB dataset preprocessing: using data enhancement technology to amplify the number of traffic signs in the GTSRB dataset, the data enhancement techniques are horizontal flip, vertical flip, rotation and brightness adjustment;

[0049] 2) Contrast enhancement of traffic signs: use the new histogram equalization (NHE) algorithm to enhance the overall contrast of the traffic signs processed in step 1). Do linear weighting, and finally integrate the low and high frequency data to obtain the final image, such as figure 2 As shown, the process of using the NHE algorithm, where E(·) represents the frequency division filter, HE represents the histogram process; k represents the weighting coefficient, f(x,y), f a (x,y), f b (x, y) and F(x, y) represent the original input, low-frequency components, high-frequency components and th...

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 discloses a traffic sign recognition method based on an LDCNN model and an NHE algorithm. The method includes 1) GTSRB data set is preprocessed; 2) the contrast of traffic sign is enhanced; 3) the LDCNN model is constructed; 4) the LDCNN model is trained; and 5) recognition is performed. The method is convenient to use, can extract more features of traffic sign, and improves the accuracy of traffic sign recognition.

Description

technical field [0001] The invention relates to the fields of deep learning and images, in particular to a traffic sign recognition method based on an LDCNN model and an NHE algorithm. Background technique [0002] In recent years, the development of advanced driver assistance system ADAS has received extensive attention from the government and the automotive industry. Traffic sign recognition system TSRS, as one of the subsystems of ADAS, can provide drivers with warning and guidance information and improve vehicle driving safety. Therefore, research Traffic sign recognition has important theoretical significance and practical value. [0003] In the early 1990s, experts at home and abroad have proposed many TSRS based on traditional algorithms, such as: Liu H et al. proposed a TSRS model based on sparse coding features; a method based on decision tree and wavelet transform (DT-CWT) was proposed in the literature. The TSRS model. The above traditional algorithms cannot mee...

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 Applications(China)
IPC IPC(8): G06K9/00G06K9/46G06N3/04
CPCG06V20/54G06V10/50G06N3/045
Inventor 黄知超李栋
Owner GUILIN UNIV OF ELECTRONIC TECH