3D Data Acquisition and Registration Using Two Opposing Kinects

Abstract

General rightsThis document is made available in accordance with publisher policies. Please cite only the published version using the reference above. AbstractWe present an automatic, open source data acquisition and calibration approach using two opposing RGBD sensors (Kinect V2) and demonstrate its efficacy for dynamic object reconstruction in the context of monitoring for remote lung function assessment. First, the relative pose of the two RGBD sensors is estimated through a calibration stage and rigid transformation parameters are computed. These are then used to align and register point clouds obtained from the sensors at frame level. We validated the proposed system by performing experiments on known-size box objects with the results demonstrating accurate measurements. We also report on dynamic object reconstruction by way of human subjects undergoing respiratory functional assessment.

Omnidirectional 3D Point Clouds Using Dual Kinect Sensors

“…Soleimani et al [7] used two identical checker patterns on a double-sided board for the calibration of two Kinect V2. Here, the same checker pattern should be placed at the same position on both sides of the flat board.…”

“…In this case, however, the overlapped scanned data required for the correspondence-based registration methods may not be sufficient, which makes it difficult to employ the conventional calibration approaches. To solve this problem, a specially designed calibration device such as a spherical object with a rod [6] or a double-sided checker board [7] can be used. In this paper, a novel calibration method without using any calibration-assisting device is proposed.…”

Depth-Based Whole Body Photoplethysmography in Remote Pulmonary Function Testing

“…dual-Kinect 3-D data acquisition and registration system [27] reconstructs an almost complete 3-D model of a subject, performing the routine PFT in a sitting position, at full frame rate (30fps). Deploying only two Kinect sensors, (a) minimises the temporal and spatial alignment error, (b) reduces the system setup and calibration effort, (c) keeps system costs low, and (d) minimises the overall operation space.…”

“…The average error range across 3 boxes at 3 different placements was 0.21− 0.84 cm. The synchronisation and registration methodology, and registration accuracy, is comprehensively reported in [27] and the source code is publicly available. 1…”

“…1. Each of the Kinects was placed at the distance of ∼1.5 m away from the subject to minimise the noise [6], [27] and at a height of 0.6 m. Each subject was asked to wear a reasonably tight T-shirt and sit up straight on a backless chair. Participants were neither restricted nor advised to be stationary during the PFTs, and the tests were performed as routine spirometry.…”