3D Reconstruction from Integral Images Based on Interpolation Algorithm

Integral Imaging (II) is a technique that is capable of creating and encoding a true volume spatial optical model of the object scene in the form of a planar intensity distribution through micro-lens array. With recent progress in the theory and micro-lens manufacturing, integral imaging is becoming a practical and prospective 3D display technology and attracting much attention. In this application, the generation of depth map and 3D scene reconstruction from the planar recorded integral image is essential if real and/or computer generated objects are to be integrated within integral 3D images. It is also essential to enable content-based image coding and content-based interactive manipulation to be carried out within integral images. In this technology, 3D scene information is existed in a great deal of "Element Image" under each micro-lens. A true volume spatial optical model of the 3D scene can be perceived through the corresponding micro-lens array replaying. Since micro-lens sheet is used in recording, only one recording is necessary in obtaining three-dimensional information. The precision of depth information will affect the integral images evidently. This paper presented a method of extracting more precise and reliable depth information with the interpolation algorithm based on the present depth extraction algorithm. Contrast experiment and data analysis prove that the 3D reconstruction result is more confirmable with our current depth information, the precision of 3D reconstruction is improved. This is very important to acquisition of the exact 3D scene information, analysis of the 3D spatial resolution and data processing of the next generation II-based 3DTV.