A new Wavelet Prediction method for GPR clutter elimination Based on LSTM network
- Resource Type
- Conference
- Authors
- Juan, He; Longxiang, Wang; Hongxia, Ye
- Source
- 2021 International Applied Computational Electromagnetics Society (ACES-China) Symposium Applied Computational Electromagnetics Society (ACES-China) Symposium, 2021 International. :1-2 Jul, 2021
- Subject
- Computing and Processing
Engineering Profession
Fields, Waves and Electromagnetics
Ground penetrating radar
Computational modeling
Neural networks
Time series analysis
Predictive models
Data models
Unmanned aerial vehicles
Long short-term memory network
wavelet extraction
- Language
Ground penetrating radar (GPR) wavelet greatly affects the inversion results of underground structures. This paper presents a new wavelet extraction method based on the LSTM neural network. The simulation data with FDTD method and the measured data collected by the self-built unmanned aerial vehicle ground penetrating radar (UAV-GPR) are used for test, and the predicted wavelets are almost all offset by the wavelet components of the original data. The results show that the LSTM neural network can effectively predict the wavelets and their tailing oscillations for different detection scenes.