Depth Estimation and Object Recognition using Integral Imaging (Invited Paper)

Abstract

In this invited talk, depth estimation and object recognition using integral imaging are presented. The computational method reconstructs three-dimensional information at arbitrary depth-levels, eliminating the occluding effect in order to visualize the object of interest. The depth of the object is estimated where the uncertainty of the corresponding intensities is minimized. Various applications including edge detection and statistical pattern recognition can be performed using the 3D information acquired by integral imaging.