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

4D millimeter wave radar point cloud-based vehicle detection method

A millimeter wave radar, vehicle detection technology, applied in neural learning methods, instruments, biological neural network models, etc., can solve the problem of inability to obtain accurate motion information of traffic targets, lidar performance easily affected by bad weather, susceptible to light and weather and other factors, to achieve good target detection effect, improve target detection probability, balance effectiveness and diversity.

Pending Publication Date: 2021-10-22
NANJING UNIV OF SCI & TECH
View PDF0 Cites 2 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

Among them, the vision-based vehicle detection method can capture real-time picture information of traffic targets, but is easily affected by factors such as light and weather, and cannot obtain accurate movement information of traffic targets; the performance of lidar is easily affected by bad weather and the cost is high

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
  • 4D millimeter wave radar point cloud-based vehicle detection method
  • 4D millimeter wave radar point cloud-based vehicle detection method
  • 4D millimeter wave radar point cloud-based vehicle detection method

Examples

Experimental program
Comparison scheme
Effect test

Embodiment

[0050] The self-made 4D radar data set in this example contains a total of 3726 frames of data. The data content includes 4D millimeter-wave radar point cloud files, label files generated based on manual annotation of 4D millimeter-wave radar point clouds, and pictures and lidar for corresponding reference. data.

[0051] The original point cloud of the 4D millimeter wave radar in this embodiment is as follows: Figure 4 As shown, the original point cloud is preprocessed: According to the fusion filtering method in step 1, the effect after filtering is as follows Figure 5 As shown in (a), a large number of noise points and outliers are filtered out, and the target point cloud information is completely preserved; then the point cloud is clustered according to the improved DBSCAN algorithm in step 2, and the clustering results are as follows Figure 5 (b) shown. Then perform step 3, input the preprocessed point cloud into the three-branch PointRCNN network for training and te...

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 4D millimeter wave radar point cloud-based vehicle detection method. The method comprises the steps of filtering outliers and noise points in a point cloud data set by using a fusion filtering algorithm; performing point cloud clustering by using an improved DBSCAN algorithm, recognizing point cloud data with similar characteristics, and completing point cloud preprocessing; and detecting the preprocessed 4D millimeter wave radar point cloud data set by using a three-branch Point RCNN network to obtain a detection result. According to the invention, high target detection probability is realized, and the problem that 4D millimeter wave radar target detection technologies are deficient is preliminarily solved.

Description

technical field [0001] The invention belongs to radar point cloud target recognition technology, in particular to a vehicle detection method based on 4D millimeter wave radar point cloud. Background technique [0002] With the rapid development of the economy and the gradual improvement of people's living standards, the penetration rate of cars has gradually increased. However, with the continuous growth of car ownership, traffic accidents have gradually occurred frequently. Improper operation of drivers is the main cause of frequent traffic accidents. reason. The development of intelligent transportation and automatic driving technology will greatly improve driving safety, and vehicle detection technology is one of the key links. [0003] At present, common vehicle detection methods include vision-based vehicle detection and lidar point cloud-based vehicle detection. Among them, the vision-based vehicle detection method can capture real-time picture information of traffic...

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): G06K9/62G06K9/40G06N3/04G06N3/08
CPCG06N3/08G06N3/045G06F18/2321G06F18/2433Y02T10/40
Inventor 吴礼仝璐李小柳胡泰洋肖泽龙彭树生
Owner NANJING UNIV OF SCI & TECH
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