Three-dimensional reconstruction method and device of multi-eye camera, vr camera and panoramic camera

A multi-camera, three-dimensional reconstruction technology, applied in 3D modeling, computer parts, character and pattern recognition, etc., can solve problems such as poor robustness, and achieve increased robustness, favorable calculation accuracy, and reduced data volume. Effect

Active Publication Date: 2021-01-29
ARASHI VISION INC
View PDF7 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0003] The object of the present invention is to provide a three-dimensional reconstruction method, device, computer-readable storage medium, VR camera and panoramic camera of a multi-eye camera, aiming to solve the robustness of three-dimensional reconstruction based on images collected by a multi-eye camera with ordinary lenses poor problem

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
  • Three-dimensional reconstruction method and device of multi-eye camera, vr camera and panoramic camera
  • Three-dimensional reconstruction method and device of multi-eye camera, vr camera and panoramic camera
  • Three-dimensional reconstruction method and device of multi-eye camera, vr camera and panoramic camera

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0025] see figure 1 , the three-dimensional reconstruction method of the multi-eye camera provided in Embodiment 1 of the present invention includes the following steps: It should be noted that, if there are substantially the same results, the three-dimensional reconstruction method of the multi-eye camera of the present invention does not use figure 1 The flow sequence shown is limited.

[0026] S101. Obtain a video captured by each camera during the movement of the multi-camera.

[0027] In Embodiment 1 of the present invention, the multi-eye camera may be a VR camera or a panoramic camera including at least two cameras, such as figure 2 The multi-eye camera shown includes 6 cameras arranged at equal intervals in a circular shape, and of course, may include 3, 4, 5, 8, or even more than 8 cameras. VR cameras and panoramic cameras usually use fisheye lenses or wide-angle lenses.

[0028] In Embodiment 1 of the present invention, S101 may be specifically: Acquiring videos ...

Embodiment 2

[0061] see image 3 , the three-dimensional reconstruction device of the multi-eye camera provided in Embodiment 2 of the present invention includes:

[0062] The video acquisition module 11 is used to acquire the video taken by each camera during the moving process of the multi-eye camera;

[0063] Key frame extracting module 12, is used for extracting key frame simultaneously to the video that all camera shoots, and does distortion correction to each key frame;

[0064] The three-dimensional reconstruction module 13 is used to perform three-dimensional reconstruction according to the key frames corresponding to the videos taken by all cameras, and generate point clouds and camera poses of the cameras;

[0065] The point cloud processing module 14 is configured to acquire common point clouds generated by adjacent cameras, and determine scales of all point clouds according to the common point clouds generated by adjacent cameras.

[0066] The 3D reconstruction device for a m...

Embodiment 3

[0068] Embodiment 3 of the present invention provides a computer-readable storage medium, the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the three-dimensional Steps of the reconstruction method.

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 is applicable to the field of computer vision, and provides a three-dimensional reconstruction method and device of a multi-eye camera, a VR camera and a panoramic camera. The method includes: acquiring a video taken by each camera during the movement of a multi-camera, the multi-camera being a VR camera or a panoramic camera comprising at least two cameras; simultaneously extracting key frames from the videos taken by all cameras, and Perform distortion correction for each key frame; perform 3D reconstruction based on the key frames corresponding to the videos captured by all cameras, generate point clouds and camera poses; obtain public point clouds generated by adjacent cameras, and generate public point clouds based on adjacent cameras The point cloud determines the scale of all point clouds. The present invention can provide a specific and real physical scale for the point cloud, and the lens of the VR camera or panoramic camera can provide a wider field of view, cover a more complete angle, and increase the coincidence of two adjacent lenses at the same time. The determination of point cloud scale provides more common points and increases robustness.

Description

technical field [0001] The invention belongs to the field of computer vision, and in particular relates to a three-dimensional reconstruction method and device of a multi-eye camera, a VR camera and a panoramic camera. Background technique [0002] In the prior art, for the positioning and navigation of robots, a common lens-based multi-eye camera is usually used to collect images, and then three-dimensional reconstruction is performed through the SFM (Structure From Motion, motion recovery structure) algorithm. However, the field of view of ordinary lenses is relatively narrow, the collected images cannot cover a complete angle, and the coincidence degree of two adjacent lenses is relatively low, so the robustness of 3D reconstruction is poor. Contents of the invention [0003] The object of the present invention is to provide a three-dimensional reconstruction method, device, computer-readable storage medium, VR camera and panoramic camera of a multi-eye camera, aiming t...

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): G06T7/55G06K9/00
CPCG06T7/55G06T2207/10016G06T17/00G06T7/80G06V20/46G06T7/579G06T7/593G06T2207/10021G06T2200/08
Inventor 谢亮
Owner ARASHI VISION INC
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