Resolution enhanced 3D image reconstruction by use of ray tracing and auto-focus in computational integral imaging

We propose a three-dimensional (3D) image reconstruction algorithm that can computationally reconstruct the light field image with enhanced resolution. This method is based on ray tracing and can reconstruct the elemental images from arbitrary viewpoints. In the conventional integral imaging reconstruction process, a plane needs to be selected to make the reconstructed image in focus, so prior knowledge is required. Auto-focus (AF) of the digital camera is applied in the integral imaging reconstruction process and the optimal reconstruction plane can be selected automatically without any prior knowledge. Thus, the integral imaging system can record the light field of 3D scene without focusing on them and image focusing will be completed in the computational post-processing process. To our knowledge, this is the first time to apply auto-focus method to 3D integral imaging reconstruction. Furthermore, the depth-of-field can be controlled by adjusting the transmission angle of light ray from the elemental images. Experimental results are presented to demonstrate the feasibility and effectiveness of the proposed method.