Global characterization and optimization of stereoscopic displays with a perceptual simulating camera

Low crosstalk between left- and right-eye images and high luminance homogeneity are two essential factors for achieving a comfortable autostereoscopic display viewing experience. In this work, we propose a backlight parameter optimization method based on a perceptual light field model for directional backlight autostereoscopic displays. A calibrated camera-based measurement system is developed to capture the full-screen distributions of crosstalk and luminance homogeneity within a few seconds. The measured parameters are incorporated into the perceptual light field model, enabling in-the-loop camera optimization. This approach ranks candidate backlight strategies for each target viewpoint and selects the optimal one. Experimental results demonstrate that the proposed method achieves a balanced design of low crosstalk and high luminance homogeneity across different viewpoints. Furthermore, a minimal forced-ranking user test conducted with representative scenes yields rankings consistent with the perceptual light field model predictions, confirming that the model aligns well with user comfort requirements.