Abstract Point cloud semantic segmentation, a crucial research area in the 3D computer vision, lies at the core of many vision and robotics applications. Due to the irregular and disordered of the point cloud, however, the application of convolution on point clouds is challenging. In this article, we propose the “coordinate convolution,” which can effectively extract local structural information of the point cloud, to solve the inapplicability of conventional convolution neural network (CNN) structures on the 3D point cloud. The “coordinate convolution” is a projection operation of three planes based on the local coordinate system of each point. Specifically, we project the point cloud on three planes in the local coordinate system with a joint 2D convolution operation to extract its features. Additionally, we leverage a self‐encoding network based on image semantic segmentation U‐Net structure as the overall architecture of the point cloud semantic segmentation algorithm. The results demonstrate that the proposed method exhibited excellent performances for point cloud data sets corresponding to various scenes.

Zhaoxuan Zhang

Kun Li

Xuefeng Yin

Xinglin Piao

Yuxin Wang

Xin Yang

Baocai Yin

Computer Animation and Virtual Worlds

随着功能强大的个人电脑和高端显卡的出现，虚拟世界、实时计算机动画和模拟、游戏出现了令人难以置信的发展。但与此同时，出现了新的、更便宜的虚拟现实设备，允许与这些实时虚拟世界甚至通过增强现实与现实世界进行交互。 3D 角色，尤其是虚拟人现在具有卓越的品质，可以在电影行业中使用它们。但这只是一个开始，随着人工智能和智能体技术的发展，这些角色将变得越来越自主甚至智能。他们将与动物和植物一起在虚拟生活中居住在虚拟世界中。

With the rapid advancement of personal computers and high-end graphics cards, significant progress has been made in the fields of virtual worlds, real-time computer animation, simulation, and gaming. At the same time, the emergence of new and more affordable virtual reality devices allows for interaction with these real-time virtual worlds, even integrating augmented reality with the real world. 3D characters, especially virtual humans, now possess exceptional quality and can be utilized in the film industry. However, this is just the beginning. As artificial intelligence and agent technologies continue to evolve, these characters will become increasingly autonomous and even intelligent. They will coexist with animals and plants in virtual life within these virtual worlds.

随着个人电脑和高端显卡的快速发展，虚拟世界、实时计算机动画、模拟和游戏领域取得了显著进步。同时，新型、价格更低的虚拟现实设备的出现，使得人们能够与这些实时虚拟世界进行交互，甚至通过增强现实与现实世界相结合。3D角色，特别是虚拟人，已经具备了卓越的品质，能够在电影行业中得到应用。然而，这仅仅是个开始。随着人工智能和智能体技术的不断进步，这些角色将变得越来越自主，甚至具备智能。它们将与动物和植物共同生活在虚拟世界中。

Point cloud semantic scene segmentation based on coordinate convolution

The “coordinate convolution” is proposed,” which can effectively extract local structural information of the point cloud, to solve the inapplicability of conventional convolution neural network structures on the 3D point cloud.

Category	Quartile
计算机科学	4
计算机科学, 计算机软件工程	4

Point cloud semantic scene segmentation based on coordinate convolution

ivySCI AI Smartly Parses PDF, Answers Researchers' Questions, and Helps You Understand Papers in Seconds

Journal Info

Category	Quartile
COMPUTER SCIENCE, SOFTWARE ENGINEERING	4