Monocular slam method and device combined with deep learning

A deep learning and single-purpose technology, applied in image analysis, instrumentation, computing, etc., can solve problems such as easy failure, slow initialization process of depth information, and failure to use prior knowledge of the scene, so as to improve robustness, speed, and speed up Effects of Speed ​​and Accuracy

Active Publication Date: 2022-07-05
TSINGHUA UNIV
View PDF12 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

The initialization process only uses certain geometric constraints, and does not use the prior knowledge of the scene depth, resulting in a slow and easy failure of the depth information initialization process.
The subsequent optimization process is only based on the previous initialization, without using the prior knowledge contained in the scene

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
  • Monocular slam method and device combined with deep learning
  • Monocular slam method and device combined with deep learning
  • Monocular slam method and device combined with deep learning

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0037] The following describes in detail the embodiments of the present invention, examples of which are illustrated in the accompanying drawings, wherein the same or similar reference numerals refer to the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary, and are intended to explain the present invention and should not be construed as limiting the present invention.

[0038] The monocular SLAM method and device combined with deep learning proposed according to the embodiments of the present invention will be described below with reference to the accompanying drawings. First, the monocular SLAM method combined with deep learning proposed according to the embodiments of the present invention will be described with reference to the accompanying drawings.

[0039] figure 1 It is a flow chart of the monocular SLAM method combined with deep learning accordi...

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 SLAM method and device combined with deep learning, wherein the method includes the following steps: using CNN to process an input image to obtain initial depth information; using the depth map as an initial value to initialize monocular simultaneous positioning and map construction SLAM system, combined with the optimized depth map to obtain a high-precision depth map; select and optimize key frames based on the high-precision depth map, and use the optimized key frames to optimize positioning and mapping, and obtain the final simultaneous positioning and map construction results. . This method uses the depth information obtained by the CNN network as the initial value of the image depth, which greatly improves the initialization speed of the SLAM system and the accuracy of subsequent joint optimization.

Description

technical field [0001] The invention relates to the technical field of simultaneous positioning and mapping of cameras, in particular to a monocular SLAM method and device combined with deep learning. Background technique [0002] SLAM (simultaneous localization and map construction) studies how to recover the three-dimensional information of the scene from the image sequence and locate its own pose at the same time. Great help for the SLAM process. [0003] Monocular SLAM technology can be divided into RGB-D and ordinary monocular vision SLAM technology according to different sources of depth information. The advantages and disadvantages of RGB-D technology are limited by the device. RGB-D cameras such as Microsoft's Kinect camera, Intel's RealSense, etc. can obtain image data and depth data at the same time, but they are easily affected by light, and the ranging distance is short. Taking Microsoft's Kinect camera as an example, the farthest only Can detect 5 meters. Or...

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): G06T7/55G06T7/73
CPCG06T7/55G06T7/73
Inventor 李一鹏郝敏升戴琼海
Owner TSINGHUA 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

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