Depth Map Estimation of Single-View Image using Smartphone Camera for a 3-Dimension Image Generation in Augmented Reality

Abstract

Immersive experience exhibits such as augmented reality (AR) are a way to enjoy museums and art galleries because this way not only may update information easily but also does not require physical existence. However, 3-dimension (3D) images of AR are difficult to apply with one image as depth information either is unknown, or is cost and time consuming. This paper designs a depth map estimation method from single-view image by using smartphone camera to generate a 3D image in AR. Some functions are added into AR Depth Lab [7] in a personal computer (PC). Then, the improve AR Depth Lab [7] is installed into the smartphone. The depth map can be obtained by using patch-based depth estimation [8] and saved into the smartphone. The appropriate parameters of the proposal such as patch size, the cropped image size and the appropriate image models are set by experiments.