Sezgin, MehmetAlpdemir, Mahmut Nedim2023-11-102023-11-102023https://hdl.handle.net/11467/6986https://doi.org/10.1109/IC_ASET58101.2023.10150717In this study, we present the results of two-class identification of buried objects using convolutional neural networks on real GPR dataset with 1080 images. The dataset includes GPR images of clutter objects and surrogate mines. While clutter class consist of stones, cans, bottles, nails and similar objects, the surrogate mine class consists of metallic and non-metallic anti-personnel and anti-tank surrogate mines. We obtained nearly 100% classification results for two-class classification.eninfo:eu-repo/semantics/embargoedAccessGround penetrating radar, identification, convolutional neural networkConference ObjectN/A2-s2.0-8516425416610.1109/IC_ASET58101.2023.10150717