Zhang´s one-dimensional calibration revisited with the heteroscedastic error-in-variables model

Camera calibration is a necessary step to extract 3D information from 2D images. Since the ID object is easy to construct and without self-occlusion, the ID calibration proposed by Zhang has received many attentions. However, the progress in ID calibration mainly focuses on reducing restrictions on the ID object´s movements. The calibration accuracy still demands improvements. In this paper, the computational model of ID calibration is reformulated, noises in ID calibration are analyzed with this model, and an heteroscedastic error-in- variables model-based ID calibration algorithm is proposed. In comparison with exiting algorithms, the proposed algorithm has advantages of high accuracy with a small number of measurements, rapid convergence and weak insensitivity to initial conditions. Experiments with both synthetic and real image data validate the proposed algorithm.