Sürdürülebilir Kalkınma Amaçları
Yayınlar
- MakaleAçık Erişim2023Investigation of geometric object and indoor mapping capacity of Apple iPhone 12 Pro LiDAR
Günen, Mehmet Akif; Erkan, İlker; Aliyazıcıoğlu, Şener; Kumaş, Cavit
LiDAR (light detection and ranging) sensors use laser beams to calculate distances in the surroundings. These sensors can be applied to a wide range of tasks, and they are frequently helpful in tasks like building 3D maps, navigating airplanes, robots, conducting mining operations, and automated driving. High-resolution distance measurements are taken by LiDAR sensors, but they also gather environmental data. This information aids in locating, identifying, and quantifying things and their surroundings. The iPhone 12 Pro, which Apple released in 2020, was evaluated for accuracy with various geometric shapes and its capacity to recognize indoor environments. Free of charge 3D Scanner and the Clirio Scan application were employed in this situation. However, it was found that the root mean square error and mean error in indoor mapping were ±1.41 cm and-0.56 cm in 3D Scanner and ±3.94 cm and-0.60 cm in the Clirio Scan application, respectively, despite the findings obtained showing low accuracy in scanning small geometric objects due to the scanning difficulty. Clirio does not reject the null hypothesis in the t-test that was conducted. The accuracy of the LiDAR sensor in indoor mapping has been shown to be more promising than that of small items. In order to evaluate the reliability and reusability of the indoor mapping application according to reference measurements, intraclass correlation test was performed and the results were determined to be reliable.
- MakaleAçık Erişim2022Sentinel-2 görüntüleri kullanılarak su yüzey alanlarının belirlenmesinde sınıflandırıcıların performanslarının incelenmesi
Günen, Mehmet Akif
Nehirler, akarsular, göller ve göletler gibi su yapılarının haritalanması, su kaynaklarının gözlenmesi, zaman içerisinde değişiminin belirlenmesi ve yönetilmesi açısından oldukça önemlidir. Su yüzey alanlarının tespit edilmesi birçok faktörden etkilenen karmaşık bir süreçtir. Su ve su olmayan alanların belirlenmesi problemi için çeşitli çok kanallı bantlı uydu görüntüleri ve sınıflandırma tabanlı yaklaşımlar kullanılmaktadır. Bu çalışmada, sınıflandırma algoritmalarının etkinliğinin araştırılmasında indeks tabanlı yöntemlerden daha başarı sonuç üreten parametrik olmayan (Destek Vektör Makinalar, k-En Yakın Komşuluk ve Karar Ağaçları), olasılıksal (Saklı Markov Model) ve derin öğrenme (Oto- Kodlayıcı) tabanlı danışmalı sınıflandırma yöntemleri kullanılmıştır. Çok bantlı ve yüksek mekansal çözünürlüklü uydu görüntüleri yüksek maliyetli olduğundan sınıflandırma için Sentinel-2 uydu görüntüsüne ait sadece kırmızı, yeşil ve mavi bantlar kullanılarak Arıklar ve Denizli Göletlerine ait su yüzey alanları belirlenmiştir. Deneysel sonuçlar doğruluk, özgüllük, kesinlik, duyarlılık, f-skor gibi karışıklık matrisinden elde edilen metrikler ve görüntü kalite belirlemede kullanılan ortalama karesel hata, yapısal benzerlik indeksi, pik sinyal-gürültü oranı ve Kohen’in Kappa metriği gibi istatistiksel araçları kullanılarak karşılaştırılmıştır. Elde edilen nicel ve nitel deneysel sonuçlarda derin öğrenme tabanlı oto-kodlayıcı istatistiksel olarak en başarılı yöntem olurken, zamansal karşılaştırma açısından karar ağaçları yönteminin daha hızlı çalıştığı belirlenmiştir.
- MakaleYalnız künye2022Performance comparison of deep learning and machine learning methods in determining wetland water areas using EuroSAT dataset
Günen, Mehmet Akif
Wetlands are critical to the ecology because they maintain biodiversity and provide home for a variety of species. Researching, mapping, and conservation of wetlands is a challenging and time-consuming process. Because they produce temporal and geographical information, remote sensing and photogrammetric approaches are useful tools for analyzing and managing wetlands. In this study, the water areas of five different wetlands obtained with Sentinel-2 images in Turkey were classified. Although obtaining large amounts of high-dimensional dataset labeled for various land types is costly, it is a significant advantage to use it after model training in a wide range of applications. In this paper, the EuroSAT dataset was used in the validation process. Proposed deep learning–based 1D convolutional neural networks (CNN) and traditional machine learning methods (i.e., support vector machine, linear discriminant analysis, K-nearest neighborhood, canonical correlation forests, and AdaBoost.M1) were compared quantitatively (i.e., accuracy, recall, precision, specificity, F-score, and image quality assessment metrics) and qualitatively. Finally, pairwise comparison was made with chi-square-based McNemar’s test. There is a statistical difference between 1D CNN and machine learning method (except the support vector machine vs linear discriminant analysis in Test 1 area). CNN models outperform machine learning algorithms in terms of non-linear function approximation and the ability to extract and articulate data features. Since 1D CNNs can process data in a highly complex and unique feature space, they are very successful in segmenting strongly related and highly correlated discrete signals. It also has advantages over machine learning methods for water body extraction in that it can be integrated with sophisticated image pre-processing and standardization tools, is less susceptible to low-level random noise, and provides shift in variations and contrast-invariant image local transforms.