3D Reconstruction Based On Kinect

Three-dimensional reconstruction is an important target in computer vision, whichcan help people to digitalize the object in the real world and is widely used in differentfields. This paper introduces a cheap, fast and easy way for3D reconstruction. With thehelp of Microsoft’s Kinect as a depth sensor to get depth images and RGB color images,users can finally obtain3D point clouds model through the process of denoising,smoothing introduced in this paper.This paper have completed the following tasks: a new inter-frame filter methodbased on mean filtering and median filtering is proposed, which can filter the raw depthimages from the Kinect; by using the iterative cloud point (ICP) algorithm, thetransformation matrix of two sets of point clouds data can be calculated; the point cloudlibrary (PCL) is used for3D point clouds registration. The experiment shows that themethod can achieve the desired results of3D reconstruction.According to the work in this paper, users can reconstruct the real objects withoutexpensive equipment, what they need are just Kinect and PC. As the results shown,although the3D model is not accurate enough because of lower accuracy, the generaloutline is clear enough to meet the need in daily application.