Research on spatial resolution of a single light field camera based on forward ray tracing technique

Objetive: Since the light field camera (LFC) simultaneously samples spatial and angular information, its spatial resolution varies with the three-dimensional (3D) position, thereby exhibiting three-dimensionality, complexity, and pronounced non-uniformity. In the process of 3D scene reconstruction, the spatial resolution of the LFC affects the recoverable spatial details as well as the depth resolution, thereby influencing the accuracy of the 3D reconstruction. Therefore, calculating and analyzing the spatial resolution of the LFC is crucial for identifying the high and low resolution regions. Method: In this paper, a calculation method for the spatial resolution of an LFC is explored based on the forward ray-tracing technique, which has the advantage of high accuracy. The spatial resolutions of LFC 1.0 and LFC 2.0 under different microlens array configurations are quantitatively calculated and compared. In addition, the effects of the inverse magnification (M_l) of the main lens on the depth resolution of the LFC are investigated. Result: The results show that the LFC exhibits higher depth resolution in regions away from the intersection of the object plane and the optical axis. Conclusion: The depth resolution of LFC 2.0 near the region of (0,0,0) is better than that of LFC 1.0. For a microlens array arranged in a square pattern, the lateral resolution of LFC 2.0 shows a slight improvement over that of LFC 1.0. The depth resolution of the LFC 1.0 gradually decreases as M_l increases.