Abstract:With the development of computer vision technologies, 3D reconstruction has become a hotspot. At present, 3D reconstruction relies heavily on expensive equipment and has poor real-time performance. In this paper, we aim at solving the problem of 3D reconstruction of an indoor scene with large vertical span. In this paper, we propose a novel approach for 3D reconstruction of indoor scenes with only a Kinect. Firstly, this method uses a Kinect sensor to get color images and depth images of an indoor scene. Secondly, the combination of scale-invariant feature transform and random sample consensus algorithm is used to determine the transformation matrix of adjacent frames, which can be seen as the initial value of iterative closest point (ICP). Thirdly, we establish the relative coordinate relation between pair-wise frames which are the initial point cloud data by using ICP. Finally, we achieve the 3D visual reconstruction model of indoor scene by the top-down image registration of point cloud data. This approach not only mitigates the sensor perspective restriction and achieves the indoor scene reconstruction of large vertical span, but also develops the fast algorithm of indoor scene reconstruction with large amount of cloud data. The experimental results show that the proposed algorithm has better accuracy, better reconstruction effect, and less running time for point cloud registration. In addition, the proposed method has great potential applied to 3D simultaneous location and mapping.

Study of 3-D Reconstruction Method Based on Turntable Array Photos

A Simple 3D Reconstruction Method Based on Single Camera

A Flexible Method for 3D Reconstruction of Urban Building

Benchmarking Large-Scale Multi-View 3D Reconstruction Using Realistic Synthetic Images

Precise 3d reconstruction from a single image

Model-driven sketch reconstruction with structure-oriented retrieval

3-D Map Reconstruction in Cross-country Environment

Sketch-based 3-D modeling for piecewise planar objects in single images

Visual Odometry Based 3D-Reconstruction

Outdoor Scene 3D Reconstruction from Multiple Point Cloud

Texture Mapping on Real World Models from Multiple Photographic Images

An Interactive Modeling System from Multiple Images

3-D Reconstruction of Three Views Based on Manifold Study

Studies on Object Reconstruction from Space Using Three-Line Scanner Imagery

Research Progress of Texture Reconstruction in Color Three-Dimensional Scanning

3D Reconstruction of Outdoor Scenes from Synchronized Laser Range Data and Visual Image Data

3D Reconstruction Using a Linear Laser Scanner and a Camera

3D Reconstruction of Indoor Scenes Via Image Registration

3D Reconstruction From Traditional Methods to Deep Learning

Research on Object Panoramic 3D Point Cloud Reconstruction System Based on Structure From Motion

A Method of Subway 3D Reconstruction from Monocular Image