Click on “Download PDF” for the PDF version or on the title for the HTML version.

If you are not an ASABE member or if your employer has not arranged for access to the full-text, Click here for options.

Geometrical Calibration and Integration of Laser 3D and X-ray Dual Systems

Published by the American Society of Agricultural and Biological Engineers, St. Joseph, Michigan

Citation:  Paper number  033085,  2003 ASAE Annual Meeting . (doi: 10.13031/2013.15005) @2003
Keywords:   3D range image, camera, laser, geometrical features, x-ray image, dual system, calibration

In the bone fragment detection system currently being developed by the Bioimaging and Machine Vision Lab at UMD, there are three sensors and two light sources involved. To obtain the best performance of the system, the geometrical relation of these sensors and light sources need to be calculated. Among these parameters are the camera's intrinsic, extrinsic matrix, the geometrical relation between the cameras and laser beam, the relation between the laser beam and the x-ray beam, and the x-ray image distortion due to the fan-beam effect of x-rays. To accurately calibrate all these parameters, three steps were used. First, the cameras and laser were calibrated. The system was designed that all the sensors were placed in an enclosed chamber. Due to the limited size and low illumination in the chamber, normal calibration method was difficult to be adopted. To resolve this issue, calibration models and their 3D information were used to adjust the laser beam geometrical positions. By using laser beams and known geometrical constraints, the cameras were calibrated and the relation between the cameras and laser beams were calculated. In the second step, geometrical features of both x-ray image and 3D image were compared, the relation of x-ray source and laser were computed. At last, the image distortion caused by x-ray projection was corrected. The experiment was conducted to verify the calibration result. The calibration of these parameters was critical in obtaining true geometrical description of the measured object from the data through x-ray and cameras, thus improved the accuracy of the bone-fragment detection system.

(Download PDF)    (Export to EndNotes)