基于无人机遥感的苏州市未审批建设监测与识别研究

Research on Monitoring and Identification of Unapproved Construction in Suzhou Based on UAV Remote Sensing

摘要: 随着我国城镇化进程进入高质量发展阶段，城市建设用地资源日趋紧张，各类未审批建设行为频发，严重干扰城市规划的有序实施，并对公共安全构成潜在威胁。针对当前未审批建设治理中普遍存在的“入户难、取证难、发现滞后”等突出问题，本文引入无人机航测技术，构建了基于高分辨率遥感影像的未审批建设动态监测体系，旨在提升城市治理水平，为未审批建设监管的治理提供新技术路径。研究首先系统梳理了国内外未审批建设治理的法律法规体系与技术研究进展；其次，深入阐述了无人机航摄技术的系统构成、影像获取的技术参数与质量控制要求，并重点分析了基于Pix4DMapper的空中三角测量原理、正射影像(DOM)与数字表面模型(DSM)的生成算法。以苏州市高新区为实证研究区，利用多旋翼无人机平台每季度获取一次地面影像数据，通过Pix4DMapper软件生成高精度DOM与DSM，结合ArcGIS空间分析工具对多期影像进行叠加比对与变化检测，识别并提取新增未审批建设图斑，完成数据的统计汇总与动态更新。研究结果表明：(1) 基于无人机遥感的高分辨率影像数据能够快速、精准地提取违法用地变化信息，DSM差值分析方法可有效识别高程变化区域，结合DOM目视解译可精确定位新增未审批建筑位置与范围；(2) 研究区2019~2020年间共识别疑似未审批建筑图斑1352个，经外业核查验证，识别准确率达90.4%，其中企业厂区加建占比最高(42.3%)，住宅楼顶加建次之(28.5%)；(3) 该技术体系显著缩短了未审批建筑发现周期，较传统人工巡查方式效率提升约70%，有效填补了高层建筑顶部、厂区内部等模糊地带。本研究拓展了遥感技术在城市治理领域的应用场景，丰富了未审批建设治理的方法体系，为推进城市治理提供了技术支撑与实践范式。

Abstract: With China’s urbanization process entering a high-quality development stage, urban construction land resources are becoming increasingly scarce, and various unapproved construction behaviors occur frequently, which seriously interferes with the orderly implementation of urban planning and poses a potential threat to public safety. In view of the outstanding problems in the governance of unapproved construction, such as “difficulty in entering the house, difficulty in obtaining evidence, and lag in discovery”, this paper introduces the aerial survey technology of unmanned aerial vehicles, and constructs a dynamic monitoring system of unapproved construction based on high-resolution remote sensing images, aiming at improving the level of urban governance and providing a new technical path for the governance of unapproved construction supervision. Firstly, the research systematically combs the legal and regulatory system and technical research progress of unapproved construction governance at home and abroad; Secondly, the system structure of UAV aerial photography technology, technical parameters and quality control requirements of image acquisition are expounded in depth, and the principle of aerial triangulation based on Pix4DMapper, the generation algorithm of orthophoto (DOM) and digital surface model (DSM) are emphatically analyzed. Taking Suzhou high-tech zone as an empirical research area, the multi-rotor UAV platform is used to obtain ground image data once every quarter, and high-precision DOM and DSM are generated by Pix4DMapper software. Combined with ArcGIS spatial analysis tools, multi-phase images are superimposed and compared, and changes are detected, so as to identify and extract new unapproved construction spots, and complete statistical summary and dynamic update of data. The research results show that: (1) The high-resolution image data based on UAV remote sensing can quickly and accurately extract the change information of illegal land use, DSM difference analysis method can effectively identify the area of elevation change, and combined with DOM visual interpretation, it can accurately locate the location and scope of new unapproved buildings; (2) During 2019~2020, a total of 1352 suspected unapproved building spots were identified in the study area, and the identification accuracy rate reached 90.4% after field verification, among which the addition of enterprise factory buildings accounted for the highest proportion (42.3%), followed by the addition of residential buildings (28.5%); (3) The technical system significantly shortens the discovery period of unapproved buildings, and improves the efficiency by about 70% compared with the traditional manual inspection method, effectively filling the vague areas such as the top of high-rise buildings and the inside of the factory. This study expands the application scenarios of remote sensing technology in the field of urban governance, enriches the method system of unapproved construction governance, and provides technical support and practical paradigm for promoting urban governance.