A block-based 3D map: A crude sketch of steric environment

Abstract

A 3D map provides useful information for various services. Traditional 3D maps, however, consist of a raw image data and are not suitable for real-time applications. In this paper, we propose a block-based 3D map, which forms three dimensional spaces in a collection of cubic blocks. The block-based 3D map has two major parameters: an object ratio and a block size. The object ratio is defined as the rate of object pixels to air-filled pixels within a block and determines the type of a block. The block size determines actual length of the side of a cubic block and is a matter of resolution and focal distance of images. Experimental results show benefits of the block-based 3D map in reducing sensitivity to noise and in saving amounts of data to be processed. Block-based 3D maps would support a variety of new services by providing additional information to image data in intelligent real-time applications.