摘要
边坡变形监测是滑坡灾害防治和预警的重要手段,基于图像分析的近景摄影测量方法凭借全场、非接触的优势在边坡变形监测中的应用越来越广泛。而在长期野外监测应用中,环境光照变化对边坡图像的灰度一致性产生很大影响,导致该方法的测量精度受到严重制约,甚至出现测量失败的情况。从环境光照的物理过程出发,深入分析光照变化特征及其对图像灰度的影响规律,提出了一种基于灰度梯度方向特征的图像匹配方法,从而使其具备了抗环境光照影响的能力。室内边坡缩比模型实验和真实野外边坡观测实验的结果表明,该方法可以有效抑制环境光照影响,提升变形测量的精度和稳定性。研究为边坡变形摄像监测提供了一种新的手段,对该方法的实用性发展起到了支撑作用。
Slope deformation monitoring is an important means in the prevention and early warning of landslide disaster.The close-range photogrammetry based on image analysis has been widely used in the field of slope deformation monitoring with the advantages of full-field and non-contact.However,considering the fact that the change of environmental light reduces the gray consistency of captured slope images in the long-term field monitoring application,the close-range photogrammetry will be seriously limited in measurement accuracy and even lost.Based on the physical process of environmental lighting,the characteristics of lighting changes and their impact on image grayscale were deeply analyzed,and an image matching method based on gray gradient direction characteristics was proposed with the ability to resist the interference of environmental light.The results of indoor slope shrinkage model experiment and real field slope observation experiment show that this method can effectively suppress the influence of environmental light,improve the accuracy and stability of deformation measurement.The proposed research provides a new way for slope deformation monitoring by photogrammetry and plays a supporting role in the practicality of photogrammetry.
作者
王寿旭
马沁巍
郭海旭
邢同振
WANG Shou-xu;MA Qin-wei;GUO Hai-xu;XING Tong-zhen(School of Aerospace Engineering,Beijing Institute of Technology,Beijing 100081,China;School of Mechanics and Civil Engineering,China University of Mining and Technology(Beijing),Beijing 100083,China)
出处
《科学技术与工程》
北大核心
2021年第7期2773-2781,共9页
Science Technology and Engineering
基金
国家自然科学基金面上项目(11772053)。
关键词
边坡变形监测
近景摄影测量
环境光照
特征提取
slope deformation monitoring
close-range photogrammetry
environmental lighting
feature extraction
