In this paper, we present a novel approach based on genetic algorithms for
performing camera calibration. Contrary to the classical nonlinear photogra
mmetric approach [1], the proposed technique can correctly find the near-op
timal solution without the need of initial guesses (with only very loose pa
rameter bounds) and with a minimum number of control points (7 points). Res
ults from our extensive study using both synthetic and real image data as w
ell as performance comparison with Tsai's procedure [2] demonstrate the exc
ellent performance of the proposed technique in terms of convergence, accur
acy, and robustness.