Looking for breakthrough ideas for innovation challenges? Try Patsnap Eureka!

Residual SSD model-based traffic sign detection and recognition method

A recognition method and technology for traffic signs, applied in character and pattern recognition, biological neural network models, instruments, etc., can solve the problems of large proportion of traffic signs, complex traffic, and inability to meet the needs of traffic scene detection, and improve the possibility of and speed effects

Inactive Publication Date: 2018-12-07
TIANJIN UNIV
View PDF2 Cites 30 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

However, due to the large proportion of traffic signs in these images, the serious lack of background information, and the traffic in our country is often more complicated than that in foreign countries, these studies cannot meet the actual detection needs of traffic scenes in our country.

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
  • Residual SSD model-based traffic sign detection and recognition method
  • Residual SSD model-based traffic sign detection and recognition method
  • Residual SSD model-based traffic sign detection and recognition method

Examples

Experimental program
Comparison scheme
Effect test

Embodiment approach

[0088] Embodiments of the present invention are described as follows:

[0089] 1. Data preprocessing

[0090] The ratio of the number of training and test pictures in the Tsinghua-Tencent 100K dataset is roughly 2:1. Select 45 signs with more than 100 occurrences as targets for detection and recognition, and use the bounding box coordinates and categories as labels for detection and recognition, respectively.

[0091] 2. Experimental environment

[0092] The training and testing of the model are carried out on the Linux PC side, including an Intel i7-7700K CPU with 32GB memory and two NVIDIA GeForce GTX 1080Ti GPUs with 11G memory.

[0093] 3. Learning parameter configuration

[0094] The initial learning rate of the training is set to 0.001, which drops to 0.0001 after 40,000 iterations, and then continues to iterate for 40,000 iterations with a learning rate of 0.0001 to stop. Use a momentum of 0.9 and a decay rate of 0.0005. Experiments are performed on the CAFFE archi...

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 residual SSD model-based traffic sign detection and recognition method. The method includes the following steps that: first step, multi-scale segmentation is performed on animage; second step, a residual SSD model is constructed with a residual network ResNet101 adopted as the basic network of the SSD; third step, network training is carried out; and fourth step, detection and identification with generalization capacity are completed. The invention aims to improve the accuracy of the detection of small targets by an existing SSD network, and realize the effective detection and recognition of a plurality of types of signs of different sizes in the real traffic scenes in China.

Description

technical field [0001] The invention relates to the field of image detection and recognition and the field of deep learning application technology, in particular to a traffic sign image detection and recognition method. Background technique [0002] Intelligent traffic sign detection and recognition is an important technology of advanced driver assistance system (ADAS), which has made a significant contribution to the realization of road condition warning, collision avoidance and obstacle avoidance. Traditional machine learning methods generally extract image features by segmenting regions of interest, and use single or several classification operators to identify targets. The main features are the shallow information of the image such as color, shape, etc., the visual salience information such as the feature map of multiple features such as mixed color, brightness, orientation, etc., and the local invariant feature information such as gradient histogram. Although the detec...

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): G06K9/00G06N3/04
CPCG06V20/582G06N3/045
Inventor 张淑芳朱彤
Owner TIANJIN UNIV
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Patsnap Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Patsnap Eureka Blog
Learn More
PatSnap group products

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic, Popular Technical Reports.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com