Soil is the substance most likely to meet nature and dirt people, vehicles, and clothing, especially in outdoor. Both source material and soil samples can be damaged during industrial and criminal investigations. Therefore, there is a need for detection, examination, and identification systems that can minimize contact with forensic evidence and provide accurate results with fewer samples. The study aims to determine the type of soil using a low-cost, easily accessible, and highly sensitive system that can be used easily without interference from the surface properties of the textile or destruction of the structure of the dirt. The working sites and areas of samples to be collected were determined according to the purpose of the study. In this context, samples of the most common soil types were taken from the lands in the Aegean Region of Turkey. Different types of substances were applied and dirtying on the collected samples. The newly formed samples were heated with a heating surface and allowed to cool. During this process, a thermal video was recorded, and feature extraction was performed. 165 samples were obtained from 55 tests. As a result, it is seen that the proposed method can detect samples with 97% accuracy.
Thermal Image Fabric Non-Destructive Testing Material Soil Machine Learning
Birincil Dil | İngilizce |
---|---|
Konular | Yazılım Mühendisliği (Diğer) |
Bölüm | Araştırma Makalesi |
Yazarlar | |
Erken Görünüm Tarihi | 25 Aralık 2023 |
Yayımlanma Tarihi | 31 Aralık 2023 |
Gönderilme Tarihi | 7 Ağustos 2023 |
Yayımlandığı Sayı | Yıl 2023 Cilt: 7 Sayı: 3 |
Uluslararası 3B Yazıcı Teknolojileri ve Dijital Endüstri Dergisi Creative Commons Atıf-GayriTicari 4.0 Uluslararası Lisansı ile lisanslanmıştır.