Estimate the Unknown Environment with Biosonar Echoes-A Simulation Study.
Sensors (Basel)
; 21(12)2021 Jun 18.
Article
em En
| MEDLINE
| ID: mdl-34207193
ABSTRACT
Unmanned aerial vehicles (UAVs) have shown great potential in various applications such as surveillance, search and rescue. To perform safe and efficient navigation, it is vitally important for a UAV to evaluate the environment accurately and promptly. In this work, we present a simulation study for the estimation of foliage distribution as a UAV equipped with biosonar navigates through a forest. Based on a simulated forest environment, foliage echoes are generated by using a bat-inspired bisonar simulator. These biosonar echoes are then used to estimate the spatial distribution of both sparsely and densely distributed tree leaves. While a simple batch processing method is able to estimate sparsely distributed leaf locations well, a wavelet scattering technique coupled with a support vector machine (SVM) classifier is shown to be effective to estimate densely distributed leaves. Our approach is validated by using multiple setups of leaf distributions in the simulated forest environment. Ninety-seven percent accuracy is obtained while estimating thickly distributed foliage.
Palavras-chave
Texto completo:
1
Base de dados:
MEDLINE
Assunto principal:
Árvores
/
Florestas
Idioma:
En
Revista:
Sensors (Basel)
Ano de publicação:
2021
Tipo de documento:
Article
País de afiliação:
Estados Unidos