彩钢瓦类建筑检测与定位方法研究

金楠; 伍永靖邦; 施钟淇; 岳清瑞; 郑则行

doi:10.3724/j.gyjzG23120810

彩钢瓦类建筑检测与定位方法研究

doi: 10.3724/j.gyjzG23120810

金楠^1,2,3,
伍永靖邦^1,2,3,
施钟淇^1,2,3, ,,
岳清瑞^1,2,3,
郑则行^1,2,3

1. 深圳市城市公共安全技术研究院有限公司, 广东深圳 518046;
2. 城市安全风险监测预警应急管理部重点实验室, 广东深圳 518038;
3. 城市安全发展科技研究院(深圳), 广东深圳 518046

基金项目:

深圳市科技计划资助项目(ZDSYS20210929115800001)。

详细信息

作者简介:
金楠,博士后,主要从事智能监测研究。

通讯作者:
施钟淇,正高级工程师,主要从事城市安全监测预警研究,shizq@szsti.org。

计量
- 文章访问数: 108
- HTML全文浏览量: 14
- PDF下载量: 4
- 被引次数: 1
出版历程
- 收稿日期: 2023-12-08
- 网络出版日期: 2024-04-23

Research on Methods for Detection and Localization of Color Steel Tile Buildings

JIN Nan^1,2,3,
WU Yongjingbang^1,2,3,
SHI Zhongqi^{1,2,3
, ,},
YUE Qingrui^1,2,3,
ZHENG Zexing^1,2,3

1. Shenzhen Urban Public Safety and Technology Institute, Shenzhen 518046, China;
2. Key Laboratory for Urban Safety Risk Monitoring and Early Warning of Emergency Management Department, Shenzhen 518038, China;
3. Safe Urban Development Institute of Science and Technology(Shenzhen), Shenzhen 518046, China

摘要

摘要: 违章建筑的检测和定位问题一直是城市管理的挑战,彩钢瓦类建筑是界定是否属于违建建筑的重点关注对象之一。现有研究都集中于如何检测彩钢瓦类建筑,而缺少此类建筑的具体街道定位。为了解决上述问题,提出一种将违建检测与定位相结合的框架,并以无人机拍摄的可见光图像作为数据集,在深圳市某街道区域进行了实证研究。首先使用无人机对检测区域进行图像数据采集,然后通过DINO进行彩钢瓦类建筑的检测,获取彩钢瓦类建筑边界框的中心点。经坐标系转换,获取彩钢瓦类建筑的经纬度,并使用Aruco码对定位精度进行验证,最后通过地图引擎的接口将经纬度与街道信息相关联。检测结果表明,DINO在彩钢瓦类建筑检测方面表现良好,检测精度达90%。定位试验结果表明,在无人机距待测物高度30 m内时,定位精度可以控制在1 m以内,但无人机距待测物高度的增加会导致定位精度下降。提出并验证了一种有效的彩钢瓦类建筑检测与定位框架,该方法不仅具有较高的检测精度,还可确定其具体街道定位,有助于更精准和高效的城市管理。
- 违建检测 /
- 无人机定位 /
- 深度学习 /
- 遥感 /
- 城市更新
Abstract: The detection and location of illegal buildings has always been a challenge for urban management, Color steel tile buildings are one of the key objects of attention in determining whether they are illegal buildings. Existing research focuses on how to detect colored steel and tile buildings, but lacks the specific street positioning of such buildings. This study aims to address this issue by proposing a framework that integrates illegal building detection with location. An empirical study in a particular street area in Shenzhen was conducted, aerial visible light images taken by drones were used as the dataset. Firstly, drones were used to collect image data from the detection area, and then DINO was applied to detect color steel tile buildings and obtain the center points of their bounding box. After coordinate system transformation, the coordinates of the color steel tile buildings were obtained, and Aruco codes were used to verify the accuracy of the location. Finally, the coordinates were correlated with street information through the interface of a map engine. The detection results indicated that DINO performed well in detecting color steel tile buildings, achieving a detection accuracy of 90%. The positioning test results indicated that when the drone was within 30 m of the measured object, the positioning accuracy could be controlled within 1 m, but the increase in the height of the drone from the measured object wouldcause the positioning accuracy to decrease. An effective framework for the detection and location of illegal buildings was proposed and validated. This method not only has high accuracy in detecting color steel tile buildings, but can also determine the specific street location of color steel tile buildings, which contributes to more accurate and efficient urban management.
- illegal building detection /
- drone positioning /
- deep learning /
- remote sensing /
- urban renewal

HTML全文

参考文献(20)

[1]	ZHAO J, XIAO Y, SUN S, et al. Does China's increasing coupling of'urban population'and'urban area'growth indicators reflect a growing social and economic sustainability?[J/OL]. Journal of Environmental Management, 2022, 301.[2023-12-08].https://doi.org/10.1016/j.jenvman.2021.113932.
[2]	LI B S, TAN Z C. Dynamic monitoring method of illegal buildings using spatiotemporal big data based on urban high lying zones[J]. The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, 2020, 42:17-23.
[3]	WENJING Y. Analysis of illegally built buildings monitoring based on UAV remote sensing image[J]. Urban Geotechnical Investigation&Surveying, 2016, 6:84-87.
[4]	HE Y, MA W, MA Z, et al. Using unmanned aerial vehicle remote sensing and a monitoring information system to enhance the management of unauthorized structures[J/OL]. Applied Sciences, 2019, 9(22).[2023-12-08].https://doi.org/10.3390/app9224954.
[5]	王爽,林璐,徐健,等.无人机辅助违章建筑监测[J].测绘标准化,2017,33(1):39-41.
[6]	梁哲恒,邓鹏,姜福泉,等.基于卷积神经网络的无人机影像违章建筑检测应用[J].测绘通报,2021(4):111-115.
[7]	LIU Y, SUN Y, TAO S, et al. Discovering potential illegal construction within building roofs from UAV images using semantic segmentation and object-based change detection[J]. Photogrammetric Engineering&Remote Sensing, 2021, 87(4):263-271.
[8]	张勇.遥感与GIS在厦门市违章建筑监测与管理中的应用[J].测绘与空间地理信息,2016,39(2):75-77.
[9]	郝明,田毅,张华,等.基于DeepLab V3+深度学习的无人机影像建筑物变化检测研究[J].现代测绘,2021,44(2):1-4.
[10]	王爽,林璐,徐健,等.无人机辅助违章建筑监测[J].测绘标准化,2017,33(1):39-41.
[11]	朱建伟,袁国辉.基于倾斜摄影测量技术的无人机城市建筑监测系统在违建查找中的应用[J].工程勘察,2017,45(7):59-62.
[12]	LIU Y, YU Y, GAO R, et al. IBR-Yolov5:illegal building recognition with UAV image based on improved YOLOv5[C]//20227th International Conference on Image, Vision and Computing (ICIVC). IEEE, 2022:586-592.
[13]	ZHENG C, PENG B, CHEN B, et al. Multiscale fusion network for rural newly constructed building detection in unmanned aerial vehicle imagery[J]. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 2022, 15:9160-9173.
[14]	SHI X, HUANG H, YANG Y, et al. CSA-UNet:channel-spatial attention-based encoder-decoder network for rural blue-roofed building extraction from UAV imagery[J]. IEEE Geoscience and Remote Sensing Letters, 2022,19:1-5.
[15]	TAN Y, LI G, CAI R, et al. Mapping and modelling defect data from UAV captured images to BIM for building external wall inspection[J/OL]. Automation in Construction, 2022, 139.[2023-12-08]. https://doi.org/10.1016/j.autcon.2022.104284.
[16]	ZHANG H, LI F, LIU S, et al. Dino:Detr with improved denoising anchor boxes for end-to-end object detection[J/OL]. arXiv, 2022[2023-12-08]. https://doi.org/10.48550/arXiv.2203.03605.
[17]	DJI Enterprise. DJ-Innovations[EB/OL].(2023-12-08)[2023-12-08]. https://www.dji.com/cn/mavic-2-enterprise.
[18]	GARRIDO S,MUÑOZ S R,MADRID F J,et al. Automatic generation and detection of highly reliable fiducial markers under occlusion[J]. Pattern Recognition,2014,47(6):2280-2292.
[19]	曹宇辉,解明扬,李嘉铭,等.空地协同作业场景下无人机快速路径规划与自主降落技术[J].电光与控制,2023,30(10):1-6.
[20]	贺勇,李子豪,高正涛.基于视觉导航的无人机自主精准降落[J].电光与控制,2023,30(4):88-93.