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

A real-time slam method based on visual measurement

A visual measurement and incremental technology, applied in measurement devices, photogrammetry/video measurement, surveying and navigation, etc., can solve the problems of large camera offset, poor camera tracking performance, lack of 3D deep-level information, etc. Improve processing speed, good effect

Active Publication Date: 2018-12-18
上海趣立信息科技有限公司
View PDF6 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

This rating algorithm is prone to errors in the following scenarios: large camera offset or lack of 3D depth information
For example, the performance of camera tracking will be poor when facing a wall or in the absence of 3D features

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 real-time slam method based on visual measurement
  • A real-time slam method based on visual measurement
  • A real-time slam method based on visual measurement

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0021] The present invention will be further described below in conjunction with the accompanying drawings.

[0022] A kind of real-time SLAM method based on visual measurement, this method comprises the following steps: as shown in accompanying drawing 1,

[0023] Step 1: Implement CUDA: CUDA is a parallel computing framework for solving camera transformations. CUDA needs to pair two consecutive RGB depth maps given and To calculate the camera transformation that can maximize the consistency between the two pairs, the present invention selects a four-level pyramid mode to iteratively solve the camera transformation.

[0024] Step 1.1: Image preprocessing: On the GPU, preprocess the given sequence of RGB images and depth images to obtain continuous RGB depth image pairs; the calculation process of RGB depth image pairs is: for each depth image , first convert the original sensor value into a metric depth map M, because in CUDA applications, only the grayscale value of the ...

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 provides a real-time SLAM method based on vision measurement, and particularly provides a real-time SLAM method based on vision measurement and improving multi-environmental camera tracking performance, and belongs to the field of simultaneous localization and mapping (SLAM) of a mobile robot. According to the method, based on an existing miniature GPU, ICP and FOVIS are combined to achieve an RGB-D vision measurement method. Through adoption of an improved RGB-D camera position tracking system, a high-quality dense color map is produced, and a high-quality color surface model is generated, and the system has robustness. According to the method, in a situation without using a key frame, reconstruction of the real-time color surface model is achieved. The processing speed is increased, tracking can be carried out in a challenging environment with low latency, and the tracking is good in effect specially in an indoor condition.

Description

technical field [0001] The present invention relates to a simultaneous localization and mapping (SLAM, simultaneous localization and mapping) method, in particular to a real-time SLAM method based on visual measurement that can enhance camera tracking performance in multiple environments. Background technique [0002] As technology advances and hardware prices drop, density methods become more popular and applications of dense mapping and SLAM become more widespread. The traditional Kinec fusion algorithm, one of the earliest systems, reconstructs real-time scenes with high accuracy. Although the volume representation is a useful metric for planning task volumes, this algorithm has a number of limitations. And a notable feature of this algorithm is that it utilizes deep-level information only in camera motion tracking. This rating algorithm is prone to errors in the following scenarios: large camera offset or lack of 3D depth information. For example, camera tracking will...

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): G01C11/00
CPCG01C11/00
Inventor 廖鸿宇孙放
Owner 上海趣立信息科技有限公司
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