Geometric Calibration of IR Camera Using Trinocular Vision

A three-dimensional (3-D) thermogram can provide spatial information; however, it is rarely applied because it lacks an accurate method in obtaining the intrinsic and extrinsic parameters of an infrared (IR) camera. Conventional methods cannot be used for such calibration because an IR camera cannot capture visible calibration patterns. Therefore, in the current study, a trinocular vision system composed of two visible cameras and an IR camera is constructed and a calibration board with miniature bulbs is designed. The two visible cameras compose a binocular vision system that obtains 3-D information from the miniature bulbs while the IR camera captures the calibration board to obtain the two dimensional subpixel coordinates of miniature bulbs. The corresponding algorithm is proposed to calibrate the IR camera based on the gathered information. Experimental results show that the proposed calibration can accurately obtain the intrinsic and extrinsic parameters of the IR camera, and meet the requirements of its application.