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

Multi-ToF camera system real-time registration method

A camera and registration technology, applied in the field of computer vision, can solve the problem of high camera cost, achieve the effect of compact structure, reduce hardware cost, and reduce timing synchronization modules

Active Publication Date: 2021-07-13
SUN YAT SEN UNIV
View PDF6 Cites 2 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

Because it needs to calibrate the external parameters of the camera, it is not suitable for the relative movement of multiple camera positions or the movement of the shooting target, and the timing of each camera is required to be completely synchronized. The cost of such a camera is too high

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
  • Multi-ToF camera system real-time registration method
  • Multi-ToF camera system real-time registration method
  • Multi-ToF camera system real-time registration method

Examples

Experimental program
Comparison scheme
Effect test

Embodiment

[0071] like Figure 1 to Figure 3 Shown is a real-time registration method for a multi-ToF camera system, including:

[0072] S1: Multiple cameras obtain phase and amplitude information in real time, and multiple cameras have overlapping fields of view;

[0073] S2: converting the phase and amplitude information into a depth image and an infrared image respectively;

[0074] S3: Calculate the real-time pose relationship of multiple cameras based on the infrared image;

[0075] S4: Get the 3D point cloud information in the camera coordinate system according to the depth image;

[0076] S5: According to the real-time relative pose and 3D point cloud information, complete the real-time registration of the 3D information of multiple cameras.

[0077] In step S1 in this embodiment, the light source frequencies and demodulation frequencies of multiple cameras are modulated to be orthogonal to each other, so as to avoid mutual interference of light sources when multiple cameras wo...

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 belongs to the technical field of computer vision, and more particularly relates to a multi-ToF camera system real-time registration method, which comprises: respectively obtaining phase and amplitude information in real time by a plurality of cameras, wherein the plurality of cameras have overlapped visual fields; respectively converting the phase and amplitude information into a depth image and an infrared image; calculating according to the infrared image and the depth image to obtain a real-time pose relationship of the plurality of cameras; obtaining three-dimensional point cloud information under a camera coordinate system according to the depth image; and completing three-dimensional information real-time registration of the plurality of cameras according to the real-time relative poses and the three-dimensional point cloud information. According to the method, the pose relation between the cameras is solved through feature point matching, calibration of the multiple cameras is not needed, the method is suitable for the condition that relative movement exists between the multiple cameras, strict synchronization of the time sequences of the multiple cameras does not need to be guaranteed, the system cost is reduced, and high practicability is achieved.

Description

technical field [0001] The invention belongs to the technical field of computer vision, and more specifically relates to a real-time registration method for a multi-ToF camera system. Background technique [0002] In recent years, with the development of fields such as human-computer interaction, autonomous driving, medical imaging, augmented reality, and robot perception, ToF (time-of-flight) technology has become more and more widely used. Because the depth image output by a single ToF camera may have a lack of depth and a limited viewing angle, these problems can be solved by using two or more ToF cameras to work together in practical applications. At present, when matching the depth information collected by two or more ToF cameras, the method of calibrating the heterodyne between the cameras in advance is used. This method uses the heterodyne to determine the pose relationship between the cameras, and then matches different cameras. depth information, since the heterody...

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 Applications(China)
IPC IPC(8): G06T7/33G06T19/20
CPCG06T7/33G06T19/20G06T2207/10028G06T2207/10048G06T2207/30244
Inventor 刘立林胡泽天
Owner SUN YAT SEN UNIV