Robust camera calibration using neural network

Accurate camera calibration is required for achieving accurate visual measurements. In this paper, we propose a simple and flexible camera calibration using neural network that doesn't require a specialized knowledge of 3D geometry and computer vision. There are some applications, which are not in need of the values of the internal and external parameters. The proposed method is very useful to these applications. Also, the proposed camera calibration has advantage that resolves the ill-conditioned calibration in which the object plane is nearly parallel to the image plane. For a little more accurate calibration, the acquired image is divided into two regions according to radial distortion of lens and the neural network is applied to each region. Experimental results and comparison with Tsai's (1987) algorithm prove the validity of the proposed camera calibration.