测绘通报 ›› 2023, Vol. 0 ›› Issue (1): 163-167.doi: 10.13474/j.cnki.11-2246.2023.0028

• 技术交流 • 上一篇    下一篇

深度学习支持下的宅基地复垦项目真实性智能审查技术研究与应用

向煜1,2,3, 黄志1,2,3, 华媛媛1,2,3, 邓剑峰4,5   

  1. 1. 重庆数字城市科技有限公司, 重庆 401121;
    2. 城市全息空间数据应用重庆市工程研究中心, 重庆 401121;
    3. 重庆市移动测量工程技术研究中心, 重庆 401121;
    4. 重庆市勘测院, 重庆 401121;
    5. 自然资源部国土空间规划监测评估预警重点实验室, 重庆 401121
  • 收稿日期:2022-05-30 修回日期:2022-11-22 发布日期:2023-02-08
  • 作者简介:向煜(1974-),男,高级工程师,主要研究方向为智慧城市建设与3S技术、人工智能。E-mail:xiangy@dcqtech.com
  • 基金资助:
    重庆市技术创新与应用发展专项(cstc2019jscx-dxwtBX0012)

Research and application of intelligent verification technology for authenticity of homestead reclamation project based on deep learning

XIANG Yu1,2,3, HUANG Zhi1,2,3, HUA Yuanyuan1,2,3, DENG Jianfeng4,5   

  1. 1. Chongqing Cybercity Sci-tech Co., Ltd., Chongqing 401121, China;
    2. Chongqing City Holographic Spatial Data Application Engineering Research Center, Chongqing 401121, China;
    3. Chongqing Mobile Mapping Engineering Technology Research Center, Chongqing 401121, China;
    4. Chongqing Survey Institute, Chongqing 401121, China;
    5. Key Laboratory of Monitoring, Evaluation and Early Warning of Territorial Spatial Planning Implementation, Ministry of Natural Resources, Chongqing 401121, China
  • Received:2022-05-30 Revised:2022-11-22 Published:2023-02-08

摘要: 全域综合整治背景下的宅基地复垦项目审查是保证复垦项目合法性合规性的有效手段,针对当前宅基地复垦项目审查效率低的问题,本文提出了基于遥感影像的农村宅基地复垦项目立项真实性审查方法。首先以重庆市农村宅基地复垦项目为例,建立宅基地复垦项目立项真实性审查模型,利用改进的U-Net模型准确提取建筑物,然后与复垦红线叠加实现复垦项目真实性的智能审查。试验结果表明,该方法能够快速准确地审查复垦项目真实性,对优化土地结构、促进土地流转、实现乡村振兴具有重要意义,可为全域综合整治背景下宅基地复垦项目真实性审查提供技术参考。

关键词: 全域综合整治, 宅基地复垦, 卷积神经网络, 语义分割, 建筑物提取

Abstract: The review of homestead reclamation project under the background of region-wide comprehensive renovation is an effective means to ensure the legality and compliance of the reclamation project. Aiming at the low review efficiency of the current homestead reclamation project, a authenticity review method of rural homestead reclamation project was proposed based on remote sensing images. Taking the rural housing land reclamation project in Chongqing as an example, the paper establishes the authenticity review model of housing land reclamation project, and uses the improved U-Net to accurately extract buildings, and then superimpose with the reclamation red line to realize the intelligent authenticity review of the housing land reclamation project. The experimental results show that this method can quickly and accurately examine the authenticity of residential land reclamation projects, which is of great significance to optimize land structure, promote land circulation and promote rural revitalization. It can provide technical reference for the authenticity examination of residential land reclamation projects under the background of regional comprehensive renovation.

Key words: comprehensive regulation in all areas, homestead reclamation, convolutional neural network, semantic segmentation, building extraction

中图分类号: