一种高光谱图像中的桥梁检测方法  

Study on a Bridge Detection Method in Hyper-spectral Images

作　　者：孙澜澜 Sun Lanlan(School of Statistics,Yan'an University,Yan'an 716000)

机构地区：[1]延安大学数计学院,陕西延安716000

出　　处：《中阿科技论坛（中英文）》2024年第6期75-79,共5页China-Arab States Science and Technology Forum

基　　金：陕西省教育厅2023年科学研究项目一般专项项目(23JK0729)。

摘　　要：随着遥感技术的发展,高光谱图像的研究与应用逐渐进入加速期。作为一种常见的建筑物,桥梁在民用和军事方面都有具有重要意义。文章提出了一种基于高光谱图像特征的桥梁检测算法,该算法利用桥体轮廓的几何特性,将桥梁识别转换为检测图像中的两条平行直线,通过利用桥体的几何特性,成功实现了对高光谱图像中的桥梁检测。试验结果表明,桥梁有效检出率达98.31%,说明该算法能够有效检测高光谱图像中的桥梁。With the advancement of remote sensing technology,the research and application of hyper-spectral images have gradually entered an accelerated period.As a common building,bridges have significant importance in both civil and military applications.Accordingly,the article proposes a bridge detection algorithm based on hyperspectral image features,which converts the bridge into two parallel lines according to the geometric characteristics of the bridge contour Experimental results have shown that the detection rate reaches 98.31%,indicating that the algorithm can effectively detect bridges.

关 键 词：高光谱图像 桥梁检测 OSTU大津法 CANNY算法 霍夫变换 

分 类 号：TP75[自动化与计算机技术—检测技术与自动化装置]