Hardware implementation of microlens array calibration for light field video processing
Sven-Hendrik Voss, Erwin Prizkau, Marc Geh · 2018
Modern microlens-array based light field cameras are capable of capturing both angular and spatial information of light, thus delivering a 2D image together with the depth information of the whole scene (4D light field). This is accomplished by placing a microlens array in front of the camera sensor, i.e. a standard CMOS sensor. A sophisticated optical design and complex post-processing algorithms are required to reconstruct the scene and to generate multiple new images with different focus depth and perspective. The calibration of the microlenses is one decisive step in the post-processing of the captured image. In this paper we present a new calibration approach suitable for efficient FPGA implementation. State-of-the-art algorithms and implementations focus on offline non-realtime processing - mainly in software - for still images. The approach presented herein substantially underlines and improves the real-time performance of a hardware implementation of 4D video content.