The use of photogrammetry and NeRF techniques to document built heritage

Neural Radiance Field (NeRF) and photogrammetry are two methods that are used to replicate real-life objects and environments to produce high-quality models. This paper examines the use of NeRF and Photogrammetry techniques for documenting cultural built heritage. In order to compare and examine the capabilities of both techniques, the documentation of the 14th century church of Panagia Karmiotissa located in the Limassol region in Polemidia was utilized. The church was originally a Carmelite monastery from the 14th century A.D. Today, only the church exists, which was restored by the Department of Antiquities in 2001, with the name of Panagia Karmiotissa. On the north side of the church are the ruins of the monastery complex that are reduced to heaps of stones. The church of Panagia Karmiotissa is the only preserved Gothic church in the Limassol area during the time of the Frankish House of Lusignan in Cyprus. The church and Monastery area have been declared as an ancient monument by the Antiquities Department. <br/> <br/> Photogrammetry is a technique that calculates the 3D structure from a sequence of images taken from different viewpoints. It identifies common features (such as points or keypoints) in the images and then determines their spatial relationships. The Neural Radiance Field (NeRF) is a machine-learning technique that uses AI to create 3D models. The AI automatically interprets the different images where the algorithm can fill in the missing data where images do not exist. NeRF is faster and more efficient than photogrammetry in terms of processing. The neural network that powers the AI processing provides a partial compensation for noise, distortions and missing information. The use of these two techniques for cultural applications will be examined using the case study of Panagia Karmoitissa.