A Framework of Linear Sensor Networks with Unmanned Aerial Vehicle for Rainfall-Induced Landslides Detection

Abstract

This paper proposes a real-time monitoring framework for a landslide susceptibility area based on wireless sensor network using multiple Unmanned Aerial Vehicles (UAVs). Many researchers have considered building a landslide susceptibility map to distinguish different levels of landslide susceptible zones. However, to prevent damage from landslides, it is more important for the disaster control center to identify the time and location of the landslide occurrence in those highly susceptible areas. Hence, a rain-triggered landslide monitoring system is proposed herein for local mountain areas. First, a wireless sensor network framework is constructed to inform the control center as immediately as possible when landslides occur. Second, multiple UAV sensors will be responsible for collecting the stereo images of the slope in highly sensitive zones on schedule. Based on the stereo images and the binocular model, in-depth information can be obtained. With the depth information and Speeded Up Robust Features (SURF) detection, the key point characteristic information is constructed as the input data for Support Vector Machine (SVM). An SVM algorithm is designed with Python program language and executed in real time. Using this algorithm, the real-time images collected by UAVs and the landslide warning information will be sent to the control center for further analysis. Finally, a field experiment is conducted to demonstrate the effectiveness of the proposed method.