Traffic sign target detection method based on multilevel divide-and-conquer network

A traffic sign and target detection technology, applied in the field of image processing, can solve the problems of lack of category samples, low traffic sign detection accuracy and recall rate, and unbalanced number of category samples.

Pending Publication Date: 2021-09-14
XIDIAN UNIV
View PDF2 Cites 4 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0005] The purpose of the present invention is to address the deficiencies of the above-mentioned prior art, and propose a traffic sign target detection method based on a multi-level divide-and-conquer network, which is used to solve the shortage of some category samples in the traff

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 target detection method based on multilevel divide-and-conquer network
  • Traffic sign target detection method based on multilevel divide-and-conquer network

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0035] The present invention will be further described below in conjunction with the accompanying drawings and embodiments.

[0036] refer to figure 1 The implementation steps of the present invention are further described.

[0037] Step 1, generate training set and test set.

[0038] In the embodiment of the present invention, all the pictures are selected from the public TT100K2021 data set, and the images whose categories are red, blue, and yellow and whose size is W×H are used as the target picture data set I={i 1 ,i 2 ,...,i x ,...,i X}, label the coordinates of each vertex of each circumscribed rectangular frame where the traffic sign target in each picture is located and the color category it represents, and generate a label file M containing the coordinates of each vertex of each target and its category information after labeling, and label each Each vertex coordinate of each circumscribed rectangular frame where the traffic sign target in the picture is located a...

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 target detection method based on a multilevel divide-and-conquer network, which is used for solving the technical problems of low traffic sign detection precision and low recall rate in the prior art. The method comprises the following specific steps: generating a training set and a test set; training a target detection network; extracting a background category of a sample of a traffic sign-free target in the test set; enhancing the data in the training set; generating a training set of label categories and a training set of label and background categories; training a classification network; positioning and roughly classifying a to-be-detected target; and carrying out fine classification on the pictures after rough classification. The multi-level divide-and-conquer network constructed by the invention overcomes the defect that excellent results cannot be obtained on the aspects of positioning and classification of the traffic sign targets in the prior art, so the positioning and classification accuracy of the traffic sign targets is effectively improved by the multi-level divide-and-conquer network.

Description

technical field [0001] The invention belongs to the technical field of image processing, and further relates to a traffic sign target detection method based on a multi-level divide-and-conquer network in the technical field of image target detection. The invention can be used to detect objects of stationary traffic signs in natural images. Background technique [0002] Traffic sign detection refers to locating the position of traffic signs according to the color, shape and other characteristics of traffic signs in the image, and classifying and recognizing the traffic signs in the image, obtaining the meaning of traffic signs, and regulating the driving of vehicles. In recent years, a large number of target detection methods have begun to be applied to traffic sign detection, but there are still many challenges in target traffic sign detection in real tasks, such as poor classification of traffic sign targets. This is due to the uneven distribution of traffic signs caused b...

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
IPC IPC(8): G06K9/62G06N3/04G06N3/08
CPCG06N3/08G06N3/045G06F18/24G06F18/214
Inventor 冯婕要泉赫梁宇平张向荣尚荣华焦李成王蓉芳古晶
Owner XIDIAN UNIV
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Try Eureka
PatSnap group products