Our goal is to validate the out-of-focus camera calibration methods proposed by previous authors, and to disentangle variables such as the effects of image spatial resolution, pattern array size, and noise that affect camera calibration performance. Using synthetic and real-world experiments, we compare Phase-shifted Circular Gradient (PCG) patterns and circle grids to examine their relative calibration accuracy under out-of-focus blurred conditions. Real-world experiments are performed using a color E-Ink display as the calibration target, and target positioning is facilitated using a robotic arm. Initial results show that there is no relative advantage to the phase-shifted approaches over using a grid of small but still detectable circles, however, the difference is very small, and it is hypothesized that the E-Ink display might not work favourably for PCGs due to its inability to render grayscale patterns without dithering. This work provides a foundation for further investigation to compare real-world and synthetic cases which could offer additional clues about the effectiveness of each method and, could highlight how further improvements to accuracy can be found.
PDF