Common Self-polar Triangle of Concentric Conics for Light Field Camera Calibration

Accurate light field camera calibration plays an important role in various applications. Instead of a planar checkerboard, we propose to calibrate light field camera using a concentric conics pattern. In this paper, we explore the property and reconstruction of common self-polar triangle with respect to concentric circle and ellipse. A light field projection model is formulated to compute out an effective linear initial solution for both intrinsic and extrinsic parameters. In addition, a 4-parameter radial distortion model is presented considering different view points in light field. Finally, we establish a cost function based on Sampson error for non-linear optimization. Experimental results on both synthetic data and real light field have verified the effectiveness and robustness of the proposed algorithm.