2024 |
Betsas, T.; Georgopoulos, A.; Doulamis, A. INCAD: 2D VECTOR DRAWINGS CREATION USING INSTANCE SEGMENTATION Journal Article In: The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, vol. XLVIII-2/W4-2024, pp. 65–72, 2024. Abstract | Links | BibTeX | Tags: 2D instance segmentation, 2D vector drawings, Cultural heritage, orthophotos, YOLOv8 @article{isprs-archives-XLVIII-2-W4-2024-65-2024,Orthoimages are a common product used as a base in CAD software for vectorization purposes. In fact, vectorization of orthoimages constitutes a tedious and labour-intensive process which should be supervised by experts e.g. architects, chemical engineers etc. On the one hand, deep learning algorithms are used extensively nowadays achieving high quality results. On the other hand, deep learning algorithms require a huge amount of manually annotated data to be trained on, which is a very difficult process especially at pixel level applications like semantic segmentation and instance segmentation. However, the transformation of 2D CAD drawings into a suitable deep learning dataset (CAD2DLD) is underexplored ignoring a large source of data, created by experts. In this effort, the InCAD algorithm is proposed, which aims to automatically create 2D vector drawings using the YOLOv8 instance segmentation algorithm which was trained on CAD2DLD data. Additionally, a methodology for transforming 2D CAD drawings into a suitable deep learning dataset for instance segmentation, is presented. Finally, the proposed workflow is evaluated on the creation of 2D vector drawings of stones of a fortification wall achieving promising results (78.34 mIoU). |
2016 |
Georgopoulos, A; Oikonomou, C; Adamopoulos, E; Stathopoulou, E K EVALUATING UNMANNED AERIAL PLATFORMS FOR CULTURAL HERITAGE LARGE SCALE MAPPING Journal Article In: ISPRS - Int. Arch. Photogramm. Remote Sens. Spat. Inf. Sci., vol. XLI-B5, pp. 355–362, 2016, ISSN: 2194-9034. Abstract | Links | BibTeX | Tags: dense matching, large scale mapping, orthophotos, structure from motion, UAS @article{Georgopoulos2016,When it comes to large scale mapping of limited areas especially for cultural heritage sites, things become critical. Optical and non-optical sensors are developed to such sizes and weights that can be lifted by such platforms, like e.g. LiDAR units. At the same time there is an increase in emphasis on solutions that enable users to get access to 3D information faster and cheaper. Considering the multitude of platforms, cameras and the advancement of algorithms in conjunction with the increase of available computing power this challenge should and indeed is further investigated. In this paper a short review of the UAS technologies today is attempted. A discussion follows as to their applicability and advantages, depending on their specifications, which vary immensely. The on-board cameras available are also compared and evaluated for large scale mapping. Furthermore a thorough analysis, review and experimentation with different software implementations of Structure from Motion and Multiple View Stereo algorithms, able to process such dense and mostly unordered sequence of digital images is also conducted and presented. As test data set, we use a rich optical and thermal data set from both fixed wing and multi-rotor platforms over an archaeological excavation with adverse height variations and using different cameras. Dense 3D point clouds, digital terrain models and orthophotos have been produced and evaluated for their radiometric as well as metric qualities. |
2024 |
INCAD: 2D VECTOR DRAWINGS CREATION USING INSTANCE SEGMENTATION Journal Article In: The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, vol. XLVIII-2/W4-2024, pp. 65–72, 2024. |
2016 |
EVALUATING UNMANNED AERIAL PLATFORMS FOR CULTURAL HERITAGE LARGE SCALE MAPPING Journal Article In: ISPRS - Int. Arch. Photogramm. Remote Sens. Spat. Inf. Sci., vol. XLI-B5, pp. 355–362, 2016, ISSN: 2194-9034. |