研究目的
To present a method for modeling the intrinsic and extrinsic parameters of the infrared and colour cameras, and more importantly the distortions in the depth image of the Kinect system, and to demonstrate its effectiveness through calibration.
研究成果
The presented self-calibration method significantly improves the geometric accuracy and reduces noise in the Kinect point cloud. It effectively models the depth systematic errors as a function of lens distortion and relative orientation parameters, showing improvements in accuracy up to 53% and a 17% reduction in noise.
研究不足
The study is limited to the calibration of the Kinect system and does not address long-term stability of the calibration parameters. Additionally, the recovery of distortions in the projector could not be performed reliably using the proposed method.
1：Experimental Design and Method Selection:
The study employs a photogrammetric bundle adjustment with self-calibration model to correct systematic errors in the Kinect system. It includes modeling the intrinsic and extrinsic parameters of the infrared and colour cameras, and distortions in the depth image.
2：Sample Selection and Data Sources:
Two Kinect for Xbox sensors were used. Images were captured using the standard VGA resolution to ensure that the IR images are calibrated at the same image resolution as the depth images.
3：List of Experimental Equipment and Materials:
Microsoft Kinect sensors, a checkerboard target for calibration, and software tools including the MATLAB Camera Calibration Toolbox and the Microsoft Kinect SDK.
4：Experimental Procedures and Operational Workflow:
Following a two-hour warm-up period, depth images, IR images, and RGB images of a checkerboard target were acquired from various positions and orientations. The calibration involved capturing images, measuring corners of the checkerboard pattern, and selecting pixels in the depth image for depth measurements.
5：Data Analysis Methods:
The study uses the Gauss-Helmert least-square model for minimizing the summation of the weighted residuals. Baarda’s data snooping with a 5% level of significance was used to minimize the possibility of outliers in the adjustment.
独家科研数据包，助您复现前沿成果，加速创新突破
获取完整内容
