A positioning method of monocular visual odometer

A positioning method and monocular vision technology, applied in the field of computer vision, can solve problems such as unreasonable positioning strategies and insufficient robustness, and achieve high positioning accuracy, high success rate, and high robustness

Inactive Publication Date: 2021-06-08
NANJING UNIV OF AERONAUTICS & ASTRONAUTICS
View PDF0 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

However, SVO also has the problems of the direct method, as well as unreasonable positioning strategies, resulting in insufficient robustness to lighting, fast motion, and scenes with concentrated feature distribution.

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
  • A positioning method of monocular visual odometer
  • A positioning method of monocular visual odometer
  • A positioning method of monocular visual odometer

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0016] In order to make the technical means, creative features, goals and effects achieved by the present invention easy to understand, the present invention will be further described below in conjunction with specific embodiments.

[0017] Taking the UAV public dataset EUROC as an example, its collected image resolution is 752*480, and the image sequence has illumination transformation and motion blur caused by the rapid movement of the UAV, so it is challenging.

[0018] The present invention provides a monocular visual odometer positioning method that combines feature matching and semi-direct method, such as figure 1 As shown, including the following specific processes:

[0019] Step 1), read the current frame input image.

[0020] Step 2), read the current state of the current frame image, if it is the initial state, then enter the initialization module. If the current frame image is in a normal state, directly enter the sparse image alignment step. Such as figure 2 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 monocular visual odometer positioning method. The method includes the following steps: the method includes: reading the input image of the current frame; adopting a parallax-based method to complete the initialization of the visual odometer, and starting subsequent steps after success Motion estimation step; on the current frame, based on the gray invariance, construct a nonlinear optimization problem with the pose as the optimization variable, and obtain the initial inter-frame pose; on the basis of the initial pose, construct a local map, and use the The density tracking strategy completes the feature matching and sub-pixel position optimization of the local map projection points; through the local map tracking, a more accurate constraint relationship is obtained, and the pose and map points are further optimized through this relationship, and the final positioning result is output. The application of the monocular visual odometer positioning method described in the present invention can achieve high positioning accuracy, and has the robustness and real-time performance required for unmanned aerial vehicle navigation applications.

Description

technical field [0001] The invention relates to the technical field of computer vision, in particular to a monocular vision odometer positioning method. Background technique [0002] The research on autonomous navigation in known environment has been relatively mature, but the research on navigation in unknown environment has not yet formed a unified and perfect system. Autonomous navigation tasks can be divided into three steps: positioning, mapping, and path planning. Synchronous positioning and mapping technology involves the first two elements. With the rapid development of computer vision technology, the miniaturization and low cost of visual sensors, visual SLAM is gradually It has become a hot issue in the field of robot autonomous navigation. [0003] Visual odometry (Visual Odometry, VO) is an important part of the front-end of visual SLAM. Its function is to respond to the input image frame in real time and estimate the camera pose. From the implementation method,...

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): G01C11/06G01C22/00
CPCG01C11/06G01C22/00
Inventor 程月华徐贵力杨吉多才谢瑒马栎敏
Owner NANJING UNIV OF AERONAUTICS & ASTRONAUTICS
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

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

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap