GB/T 47022-2026Information technology - Technical specification for multi-view image-based 3D reconstruction systems for large-scale scenes (English PDF)

Foreword

This document complies with the provisions of GB/T 1.1-2020 "Standardization Work Guidelines Part

1.Structure and Drafting Rules of Standardization Documents". Drafting. Please note that some content in this document may involve patents. The issuing organization of this document assumes no responsibility for identifying patents. This document was proposed and is under the jurisdiction of the National Information Technology Standardization Technical Committee (SAC/TC28). This document was drafted by: Zhejiang SenseTime Technology Development Co., Ltd., Zhejiang University, China Electronics Technology Standardization Institute, and Beijing Microvision. Information Technology Co., Ltd., CESI (Shenzhen) Electronic Information Product Standardization Engineering Center Co., Ltd., Tianjin University, Shanghai Yingsu Intelligent Technology Co., Ltd. Limited Liability Company, Zhejiang Taihe Land Surveying and Planning Co., Ltd., Beijing Jinfa Technology Co., Ltd., Nanjing Jiangxing Lianjia Intelligent Technology Co., Ltd. The company, Chengdu Ruibo Technology Co., Ltd., Hunan Zhongke Xingtu Information Technology Co., Ltd., and Chongqing Dawa Hezhi Imaging Technology Co., Ltd. The company, Zhongyi Instech Technology Co., Ltd., Shanxi Provincial Smart Transportation Laboratory Co., Ltd., Beijing SenseTime Technology Development Co., Ltd. Taodianlian (Guangzhou) Information Technology Co., Ltd., China Communications Construction Road & Bridge Technology Co., Ltd., Yunnan Provincial Transportation Planning and Design Institute Co., Ltd. Jiangxi Nuclear Industry Surveying and Mapping Institute Group Co., Ltd., Tianjin Jiechuang Tiancheng Technology Co., Ltd., Yangtze River Three Gorges Surveying and Mapping Institute Co., Ltd. (Wuhan) Yangtze River Water Resources Commission Yangtze River Scientific Research Institute, Beijing Shenmou Technology Co., Ltd., Beijing Construction Engineering Quality Third Testing Co., Ltd. Beijing Longyifeng Technology Co., Ltd., Henan Zhongping Automation Co., Ltd., Beijing Suce Information Technology Co., Ltd., and Zhongpuda Technology Co., Ltd. Limited Liability Company, Shenzhen Pegasus Robotics Co., Ltd., Beijing Guanwei Technology Co., Ltd., Guangzhou Quancheng Multidimensional Information Technology Co., Ltd. Shandong Digital Human Technology Co., Ltd., Saifite Engineering Technology Group Co., Ltd., Wuhan Huafeng Electronic Engineering Co., Ltd., Beijing Yunxing Yutong Transportation Technology Co., Ltd., Sichuan Jianshan Technology Co., Ltd., China Chemical Civil Engineering Co., Ltd., Beijing Daoda Tianji Technology Joint-stock company, China National Chemical Engineering Heavy Machinery Co., Ltd., Lanzhou Surveying and Mapping Institute Co., Ltd., Guangdong Meidian Bellco Technology Group Corporation, Beijing Haoyu Tiandi Surveying and Mapping Technology Development Co., Ltd., Nanjing Model Sky Aviation Technology Co., Ltd., Donghua University Cheng Technology Co., Ltd., Hangzhou Kuangxing Technology Co., Ltd., Beijing Kantian Technology Co., Ltd., and Shenzhen Jimu Yida Technology Co., Ltd. Company, Chengdu Vision Imitation Technology Co., Ltd., Beijing Quanwang Digital Commerce Technology Co., Ltd., Beijing Zhongnuo Chainjie Digital Technology Co., Ltd. Beijing Xingyi Kaichen Culture Media Co., Ltd., Beijing Kuaiying Interactive Entertainment Media Co., Ltd., Feinuomenzhen (Beijing) Technology Co., Ltd., China

2 China Railway No.

10 Metallurgical Construction Group Co., Ltd., China Railway Sixth Group Co., Ltd., Xinjiang Transportation Science Research Institute Co., Ltd., and Beijing Ruike Tongchuang Technology Co., Ltd. Co., Ltd., Nanchang Lingxing Information Technology Co., Ltd. The main drafters of this document are. Ye Zhichao, Fan Kefeng, Zhang Guofeng, Pan Rong, Gai Meng, Lei Jianjun, Jiang Hanqing, Zhao Xinping, Sun Qifeng, and Dong Guiguan. Xiang Xiaojun, Wu Geng, Geng Yidan, Cao Qitong, Qian Chen, Peng Bo, Yu Hailin, Wang Hejun, Pan Zhaoqing, Li Yi, Pang Haitian, Zhao Qichao, Yang Ran, Yang Run Li Jianhui, Lu Qi, He Qin, Fan Libo, Zhang Jun, Jia Linxuan, Zhao Xiaojin, Jiang Lianghua, Li Junjie, Yan Wen, Zhu Zhihong, Fan Junlin, Huang Shuling Cao Qingan, Zhou Xu, He Meide, Dong Peng, Tan Lei, Liu Suhua, Li Shuai, Liu Qiaoguang, Yang Wanli, Li Mengwei, Sun Zhaohui, Wei Yu, Li Di, Yan Laifang Xin Guangyu, Wu Junhui, Gao Min, Duan Qiang, Yan Hongping, Yang Xiaodong, Zhang Wanhu, Liu Jianming, Zeng Chengqiang, Li Yonghui, Zhang Fan, Cai Liang, Wu Hao, Shu Jiangpeng Jiang Yanchuan, Mao Kai, Zheng Xing, Xu Huan, Liu Xiaopeng, Wang Jia, You Di, Shen Yushi, Xu Ning, Zou Qi, Wei Zhenxun, Song Jun, Su Hengjun. Information technology large-scale scene multi-view 3D Reconstruction System Technical Specifications

1 Scope

GB/T 47022-2026 is the Chinese national standard covering building a three-dimensional model of a city block or a construction site from photographs - the image capture and overlap required, the camera pose estimation, the dense reconstruction and meshing, and the accuracy and completeness the result must reach. First edition. It was issued on 28 January 2026 and has been in force since 1 August 2026, as a first edition. The document is under the responsibility of the Standardization Administration of China. This page is published from the official record of the 2026 edition; the clause text of a standard this recent is not yet in circulation, and the figures, limits and tables it contains are those of the document itself, delivered in full with the English translation.

This document defines the system framework for a large-scale scene multi-view 3D reconstruction system, specifying functional modules, functional requirements, and performance requirements. The question describes the corresponding testing methods. This document applies to the design, development, application, and maintenance of multi-view 3D reconstruction systems for large-scale scenes captured by digital cameras.

2 Normative references

The contents of the following documents, through normative references within the text, constitute essential provisions of this document. Dated citations are not included. For references to documents, only the version corresponding to that date applies to this document; for undated references, the latest version (including all amendments) applies. This document.

GB/T 38247-2019 Terminology for Augmented Reality in Information Technology

3 Terms and Definitions

The terms and definitions defined in GB/T 38247-2019, as well as the following terms and definitions, apply to this document. 3.1 3D Reconstruction The process of reconstructing three-dimensional information using the projection of an object onto several two-dimensional images or other depth information of the object. [Source: GB/T 38247-2019, 2.2.48]

3.2 large-scale scene A three-dimensional spatial environment covering an area ranging from hundreds of square meters to several square kilometers.

Note. This type of scenario includes enclosed indoor spaces such as multiple rooms and corridors inside buildings, as well as open outdoor areas such as urban blocks and road networks.

3.3 By using a collection of two-dimensional images acquired from two or more different perspectives, three-dimensional spatial information can be recovered and a geometric model reconstructed through computation. The process of texturing.

3.4 camera parameters A set of parameters describing camera imaging properties and spatial location, covering camera intrinsic and extrinsic parameters. Note

1.Camera intrinsic parameters determine the camera's internal imaging properties, such as optical center, focal length, and image sensor; camera extrinsic parameters determine the camera's position in the world coordinate system. and orientation. Note

2.The position and orientation described by camera extrinsic parameters are often collectively referred to as "pose". This concept is used to express the external orientation information of the camera.

3.5 Simultaneously infer camera parameters and sparse 3D structure of the target scene for each image from multi-view image sequences or video data.