Cultural landscapes are those that are shaped through the combined forces of cultural and natural activity, typically over a long period that may reach back to historic and prehistoric times. Amongst these, the documenting, modelling and visualisation of cave landscapes have always been particularly important for scientific communities such as speleology. In the current state of the art, terrestrial laser scanning provides a level of documentation of objects whose accuracy cannot be achieved by conventional methods. In addition to the highly accurate metric and geometric information, visual information brings incredible richness to the presentation of data. Virtual reality (VR) allows individuals to immerse themselves within virtual environments to explore monuments and other cultural heritage sites up close. More and more, VR systems are available at lower prices and are not only limited to VR labs. In this paper, we present research on the generation of a virtual 3D model of the İnceğiz caves, located at the Çatalca district of Istanbul, Turkey, and its integration within the Unity 3D game engine. This project, carried out as a collaboration between BİMTAŞ, a company of the Greater Municipality of Istanbul, Turkey and the Photogrammetry & Laser Scanning Lab of the HafenCity University Hamburg, Germany, aimed at developing an immersive and interactive VR visualisation of the cave for the HTC Vive Pro VR system. The entire workflow, from data acquisition to VR visualisation, is described here in detail with particular emphasis given to the 3D modelling of the cave and its integration within a VR environment.

The spatial and spectral information brought by the Very High Resolution (VHR) and multispectral satellite images present an advantage for Satellite-Derived Bathymetry (SDB), especially in shallow-water environments with dense wave patterns. This work focuses on Tavşan Island, located in the Sea of Marmara (SoM), and aims to evaluate the accuracy and reliability of two machine learning (ML) regression methods, Multi-Layer Perceptron (MLP) and Random Forest (RF), for bathymetry mapping using Worldview-2 (WV-2) imagery. In situ bathymetry measurements were collected to enhance model training and validation. Pre-processing techniques, including water pixel extraction, sun-glint correction, and median filtering, were applied for image enhancement. The MLP and RF regression models were then trained using a comprehensive dataset that included spectral bands from the satellite image and corresponding ground truth depth values. The accuracy of the models was assessed using metrics such as Root-Mean-Square Error (RMSE), Mean Absolute Error (MAE), and R² value. The RF regression model outperformed the MLP model, with a maximum R² value of 0.85, lowest MAE values from 0.65 to 1.86 m, and RMSE values from 0.93 to 2.41 m at depth intervals between 6 and 9 m. These findings highlight the effectiveness of ML regression methods, specifically the RF model, for SDB based on remotely sensed images in wave-dense shallow-water environments.

Advancements in the geometric resolution of space images have improved the conditions for generations of large-scale topographic maps. Using WorldView-1, WorldView-2, and GeoEye-1, images can now be captured from space with a 0.5 m ground sampling distance (GSD). Geometric accuracy and information content are the most significant components of mapping from space images. Depending on the resolution, image quality, and shadows, the identification and classification of ground objects may prove challenging. In this research, the geometric accuracy and information content, of panchromatic WorldView-1 images, were analyzed by covering parts of Istanbul and Zonguldak in Turkey. Each of these locations has various topographic characteristics. For the orientation and investigation of the geometric accuracies of images, a number of ground control points (GCPs) were developed as independent checkpoints. Based on bias-corrected rational polynomial coefficients with one GCP, a standard deviation of independent checkpoints on the range of one GSD was obtained. The information content of images was analyzed by mapping all buildings, in both test areas, and comparing the results with reference $1/5000$ scaled topographic maps. The results verified that the WorldView-1 images can be utilized for generating and updating $1/5000$ scaled topographic maps of urban areas.