PYCASO: Python module for calibration of cameras by Soloff’s method

Abstract

The PYthon module for the CAlibration of cameras by SOloff’s method (PYCASO) provides an open-source Python-based framework for stereoscopic reconstructions from pairs of 2D images. The determination of a matching function between image and physical coordinates is essential to accurately build 3D representations at any scales. Moreover, at the finest scales, this relationship is strongly non-linear, mainly due to optical distortions of the cameras (such as misalignment of the axis of the lens with that of the objective, refraction of optical windows etc.). Therefore, this paper proposes a general and accurate method based on the Soloff method through the PYCASO module. While the basic method is time-consuming, an acceleration of the procedure is proposed by relying on artificial intelligence (AI).

Keywords

• Stereoscopic calibration
• Precision measurement
• Artificial intelligence
• Open-source module
• Python