3D Modeling and WebVR Implementation using Azure Kinect, Open3D, and Three.js ◦Won Joon Yun and ◦;zJoongheon Kim ◦School of Electrical Engineering, Korea University, Seoul, Republic of Korea zArtificial Intelligence Engineering Research Center, College of Engineering, Korea University, Seoul, Republic of Korea E-mails:
[email protected],
[email protected] Abstract—This paper proposes a method of extracting an Algorithm 1: ICP-Registration RGB-D image using Azure Kinect, a depth camera, creating a 1: Input: The set of RGB-D nodes fragment, i.e., 6D images (RGBXYZ), using Open3D, creating it as a point cloud object, and implementing webVR using A = fa1; : : : ; am; : : : ; aM g, N : the number indicating three.js. Furthermore, it presents limitations and potentials for how many node pairs to divide. development. 2: Output: The set of point clouds P = fp1; : : : ; pk; : : : ; pK g I. INTRODUCTION s.t. M = NK_ where M; K and N 2 N 1 Recently,the mobile device with LiDAR sensor, e.g. Ipad 3: Initialize: P fg, T = fg Pro4, has been released and Apple has announced that it will 4: Definition: s: source node index, t: target node index s.t. release Apple Glass. After the VR industry emerged, it has 1 ≤ s < t < s + N ≤ M been receiving a lot of attention from the academia and the public about virtual environments such as VR, AR, and MR 5: for s in range M: around the world [1], [2]. 6: as : source node Bringing the real world into virtual reality is another area 7: for t in range [s; min(s + N; M)): of VR.