Naslov
Detecting and filtering unwanted features from GPR recordings using stationary wavelet transform

Autori
Štifanić, Daniel ; Musulin, Jelena ; Baressi Šegota, Sandi ; Glučina, Matko ; Car, Zlatan

Vrsta, podvrsta i kategorija rada
Sažeci sa skupova, prošireni sažetak, znanstveni

Izvornik
Abstract Book - 6th International Workshop on Data Science / Organiser - Centre of Research Excellence for Data Science and Cooperative Systems Research Unit for Data Science / Lončarić, Sven ; Šmuc, Tomislav - , 2021, 32-34

Skup
6th International Workshop on Data Science (IWDS 2021)

Mjesto i datum
Zagreb, Hrvatska, 24.11.2021

Vrsta sudjelovanja
Predavanje

Vrsta recenzije
Recenziran

Ključne riječi
GPR recordings ; stationary wavelet transform ; filtering

Sažetak
Detecting and identifying underground objects without excavation can be demanding, time- consuming, and at the same time, very challenging. Nowadays, with different approach i.e., by utilizing ground-penetrating radar (GPR), underground utilities such as pipes, metals, cables, concrete etc. can be observed and explored from the ground surface. Since such a method is nondestructive, sub-surface surveying has been achieved using electromagnetic radiation. First, a high-frequency electromagnetic wave is emitted into the ground where the signal is reflected, scattered, or refracted from different subsurface structures. Afterwards, the returning signal is received and recorded by the GPR. The received signal can be visually interpreted as an image where the x-axis represents the position of the radar, while the y-axis represents the depth of wave penetration. After the ground has been observed, and recordings acquired, the next step is analyzing and post-processing the data. In real-world applications, GPR recordings can be very noisy due to the presence of different unwanted objects or specific soil layers that can produce interference. Such unwanted features are contained in the same recordings as the valuable features of the observed objects. However, using a mathematical tool called Wavelet Transform (WT), the unwanted features can be partially filtered out.

Izvorni jezik
Engleski

Znanstvena područja
Računarstvo, Interdisciplinarne tehničke znanosti

POVEZANOST RADA