A non-intrusive volumetric camera calibration system

When acquiring quantitative data using cameras, calibration is required to establish the mapping relation between the image space and physical space. Calibration targets with known dimensions are often used, with the most popular being physical targets. In setups where physical access is a challenge, using physical targets may not be possible, and so we develop an adaptive non-intrusive calibration target capable of conducting volumetric calibrations in free space. The calibration target is formed by two intersecting laser beams traversed in the test domain. A novel algorithm is presented for accurately finding the beam intersections, even at small crossing angles. The error sources are assessed along with their scaling behavior with respect to key parameters. The performance of the calibration method is evaluated by using it to examine a test object with known dimensions.