Skip to Main Content
An automultiscopic display system model with a real parallel camera layout is derived. And a solution to evaluate the performance of 3D system models is also discussed. According to the theory of computer vision and binocular parallax, the mathematical expressions map the real object space to the perceived image space by 4*4 homogenous matrices. The model is estimated quantitatively by considering perceived depth range, 3D depth adjusting sensitivity and the fusion limitation of human eyes. The performance of the model is also compared with other two models. The analysis shows this model has higher 3D depth adjusting sensitivity and larger fusion limitation. The parameters of shooting and display can be adjusted more precisely so that the perceived depth can be controlled more effectively. The theory and simulation results prove this model has better value of reference in real shooting condition.