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

Smart car positioning device and method based on scene fingerprint

A positioning method and positioning device technology, applied in the field of intelligent vehicle positioning, can solve problems such as low efficiency and large data storage space, and achieve the effects of low cost, improved positioning efficiency, and high precision

Active Publication Date: 2021-10-15
WUHAN UNIV OF TECH
View PDF5 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

However, map data contains multi-dimensional local features, which requires a huge data storage space and is inefficient.

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
  • Smart car positioning device and method based on scene fingerprint
  • Smart car positioning device and method based on scene fingerprint
  • Smart car positioning device and method based on scene fingerprint

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0034] The present invention will be further described below in conjunction with specific examples and accompanying drawings.

[0035] The present invention provides a smart car positioning method based on scene fingerprints, such as figure 1 As shown, it includes the following steps:

[0036] S1. Construct the scene fingerprint map:

[0037] Use a test vehicle equipped with a GPS system and a camera to collect data on the driving route, set a node at a fixed distance, record the GPS information of each node and the front-view image information taken at the node position; the collected front-view image Input the information into the trained neural network model, extract the probability values ​​of various targets in the image, and construct a one-dimensional feature matrix according to certain rules. The formula of the feature matrix is ​​as follows:

[0038] P=[P 1 P 2 ···P N ]

[0039] where P 1 P 2 ···P N are the probabilities of the 1st to Nth types of targets...

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 present invention provides a smart car positioning method based on scene fingerprints, using a test car to collect data on the driving route, setting a node at a fixed distance, recording the GPS information of each node and the front-view image information taken at the node position; Input the collected front-view image information into the trained neural network model, extract the probability values ​​of various targets in the image, and construct a one-dimensional feature matrix; fuse the GPS information of the node with the feature matrix corresponding to the front-view image information to form a Scene fingerprint map; select the nodes around the point to be located from the scene fingerprint map as candidate nodes; construct the feature matrix of the front-view image of the point to be located; calculate the similarity between the point to be located and the feature matrix of the front-view image of the candidate node, and select The candidate node with the largest similarity is used as the final positioning node. The invention can improve the accuracy and efficiency of vehicle positioning and reduce the cost.

Description

technical field [0001] The invention belongs to the technical field of smart car positioning, and in particular relates to a scene fingerprint-based smart car positioning device and method. Background technique [0002] The current methods for realizing vehicle positioning are: 1) Vehicle positioning based on Global Positioning System (GPS) 2) Vehicle positioning based on Light Detection and Ranging (LiDAR) 3) Vehicle positioning based on vision. Among them, the GPS system is widely used for vehicle positioning due to its low cost and strong robustness, but it cannot meet the requirements of high-precision positioning, and often the GPS signal is blocked; while the positioning method based on LiDAR has expensive sensors and poor adaptability. Difference. [0003] Vision-based localization includes two main approaches: VSLAM (Visual SLAM) and map-based localization. Among them, VSLAM has defects such as closed-loop detection problems. Map-based vehicle localization adopts ...

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): G01C21/30G01S19/48
CPCG01C21/30G01S19/48
Inventor 胡钊政张帆王相龙陶倩文蔡浩
Owner WUHAN UNIV OF TECH