Effective Depth Detection of Three-Dimensional Objects in Space with Picked-up and Computationally Reconstructed Integral Images

Abstract

A new approach to effectively extract depth data of three-dimensional (3D) objects in space using elemental images picked-up from 3D objects and their computationally reconstructed plane object images (POIs) is proposed. For this approach, two image concepts are introduced: one is the mapped elemental image (MEI) defined as an inversely flipped and magnified version of the picked-up elemental image and the other is the reconstructed area image (RAI) defined as a part of the computationally reconstructed POI segmented with the area exactly overlapped with the MEI. Then, depth data of the 3D objects can be extracted through correlations between the MEIs and RAIs along the output plane. That is, a depth map can be constructed from depth data extracted from each of the picked-up elemental images, and by analyzing this depth map, the specific location of the 3D objects in space can be determined. To show the feasibility of the proposed method, some experiments with virtual and real 3D objects are performed and the results are discussed.