基于多尺度特征的工程测量遥感影像建模方法  

Modeling method for engineering surveying remote sensing images based on multi-scale features

作　　者：何洪涛[1] HE Hongtao(Sichuan Road Bridge Engineering Co.,Ltd.,Chengdu 610017,China)

机构地区：[1]四川路桥桥梁工程有限责任公司,四川成都610017

出　　处：《经纬天地》2024年第6期62-66,81,共6页Survey World

摘　　要：针对传统遥感技术在细节精度和建模效率上的局限性等问题,提出了一种基于多尺度特征的工程测量遥感影像构建模型,该模型采用三维建模技术和无人机倾斜摄影测量技术对工程测量中的建筑信息进行采集并建模。试验结果表明:在数据量为800时,多尺度特征融合技术的结构信息损失值为0.04。多尺度特征融合技术在模型重建精度、信噪比和建模效率方面均优于其他方法,能够提高地表信息的细节精度和建模效率,从而更好地满足精细化地理信息获取的需求。Addressing such limitation problems of traditional remote sensing technology as the detail accuracy and modeling efficiency.A multi-scale feature-based engineering surveying remote sensing image construction model is proposed,which uses 3D modeling technology and unmanned aerial vehicle oblique photogrammetry technology to collect building information and build models in engineering surveying.The experimental results show that when the data volume is 800,the structural information loss value of multi-scale feature fusion technology is 0.04.It is seen that multi-scale feature fusion technology outperforms other methods in terms of model reconstruction accuracy,Signal-to-Noise Ratio and modeling efficiency.It can improve the detail accuracy and modeling efficiency of surface information,thereby better meet the needs of fine-grained geographic information acquisition.

关 键 词：遥感影像 三维建模技术 倾斜摄影 多尺度特征 信噪比 

分 类 号：TP751.1[自动化与计算机技术—检测技术与自动化装置]