基于光谱和雷达的输电线路树障遥测技术研究

输电线路树障是威胁输电网安全运行的重要因素,为实现输电线路树障的高效、广域监测,文中研究了基于高光谱图像的输电线路树障区域植被类型识别以及基于雷达卫星影像的输电线路植被高度检测。首先通过机载高光谱识别树障分布区域,完成了航拍图像的树障区域类型识别,选取输电线路验证了高光谱用于树障区域植被类型识别的可行性。然后,研究了一种改进三阶段植被高度反演算法,基于算法提出了一种输电线路植被高度检测方法,利用SAR影像数据,结合植被类型识别结果进行了工程应用分析。结果表明：高光谱识别输电线路植被准确率最高可达97.5%,SAR影像检测植被高度精度最高可达86.72%,文中方法能较准确地检测输电线路树障植被类型及高度。

Study on telemetry technology of transmission line tree barrier based on spectrum and radar

The transmission line tree barrier is an important factor that threatens the safe operation of the transmission network. In order to achieve efficient and wide-area monitoring of the transmission line vegetation growth, vegetation type identification of tree barrier areas. In this paper, the vegetation type identification based on the hyperspectral image and the vegetation height detection based on the radar satellite image are studied. Firstly, the tree barrier distribution area is identified by airborne hyperspectral, and the type recognition of aerial image is completed. A transmission line is selected to verify the feasibility of hyperspectral to identify the vegetation type of tree barrier areas. Then, an improved three-stage vegetation height inversion algorithm is studied. Based on the algorithm, a transmission line vegetation height detection method is proposed. Using SAR image data, combined with the vegetation type identification results, an engineering application analysis is carried out. The results show that hyperspectral recognition accuracy of transmission line vegetation is up to 97.5%. The highest accuracy of SAR image detection vegetation height is 86.72%. And the proposed method can accurately detect the vegetation type and height of the transmission line tree barrier area.