S. Tosovic

Bio: S. Tosovic is an academic researcher from Vienna University of Technology. The author has contributed to research in topics: 3D reconstruction & Silhouette. The author has an hindex of 5, co-authored 5 publications receiving 87 citations.

3D modeling of archaeological vessels using shape from silhouette

Abstract: An algorithm for the automatic construction of a 3D model of archaeological vessels is presented. In archeology the determination of the exact volume of arbitrary vessels is of importance since this provides information about the manufacturer and the usage of the vessel. To acquire the shape of objects with handles in 3d is complicated, since occlusions of the object's surface are introduced by the handle and can only be resolved by taking multiple views. Therefore, the 3d reconstruction is based on a sequence of images of the object taken from different viewpoints. The object's silhouette is the only feature which is extracted from an input image. Images are acquired by rotating the object on a turntable in front of a stationary camera. The algorithm uses an octree representation of the model, and builds this model incrementally, by performing limited processing of all input images for each level of the octree. Beginning from the root node at the level 0 a rough model of the object is obtained quickly and is refined as the processed level of the octree increases. Results of the algorithm developed are presented for both synthetic and real input images.

Computer Aided Classification of Ceramics: Achievements and Problems.

Abstract: The first aim of this project is to establish objective criteria for the definition of the form of a vessel and to create an open classification system. Secondly the main part of the classification, that is the segmentation of the profile, should be carried out on a computer aided basis. The material basis for this exemplary attempt is provided by the late-roman burnished ware from the legionary fortress of Carnuntum. The profile sections are achieved automatically by a 3D-measurement system based on structured (coded) light and two laser-technique. Classification and reconstruction of archaeological fragments is based on the profile, which is the cross-section of the fragment in the direction of the rotational axis of symmetry. The input data for the estimation of the profile is a set of points produced by the acquisition system. A function fitting this set should be constructed and later on processed to find the extrema and inflection points necessary to classify the original fragment. The one we propose is based on cubic B-splines. This paper shows a method for shape classification of archaeological fragments, which is based on the profile.

Octree-based fusion of shape from silhouette and shape from structured light

Abstract: An algorithm for the automatic construction of a 3d model of archaeological vessels using two different 3d algorithms is presented. In archeology the determination of the exact volume of arbitrary vessels is of importance since this provides information about the manufacturer and the usage of the vessel. To acquire the 3d shape of objects with handles is complicated, since occlusions of the object's surface are introduced by the handle and can only be resolved by taking multiple views. Therefore, the 3d reconstruction is based on a sequence of images of the object taken from different viewpoints with different algorithms; shape from silhouette and shape from structured light. The output of both algorithms are then used to construct a single 3d model. Results of the algorithm developed are presented for both synthetic and real input images.

Next view planning for a combination of passive and active acquisition techniques

Abstract: In order to create a complete three-dimensional model of an object based on its two-dimensional images, the images have to be acquired from different views. An increasing number of views generally improves the accuracy of the final 3D model but it also increases the time needed to build the model. The number of the possible views can theoretically be infinite. Therefore, it makes sense to try to reduce the number of views to a minimum while preserving a certain accuracy of the model, especially in applications for which the performance is an important issue. We show an approach to next view planning for a fusion of shape from Silhouette, as an example of a passive 3D reconstruction technique, and shape from structured light, as an example of an active 3D reconstruction technique in order to get 3D shape reconstruction with minimal different views. Results of the algorithm developed are presented for both synthetic and real input images.

Combining shape from silhouette and shape from structured light for volume estimation of archaeological vessels

Abstract: An algorithm for the automatic construction of a 3d model of archaeological vessels using two different 3d algorithms is presented. In archeology the determination of the exact volume of arbitrary vessels is of importance since this provides information about the manufacturer and the usage of the vessel. To acquire the 3d shape of objects with handles is complicated, since occlusions of the object's surface are introduced by the handle and can only be resolved by taking multiple views. Therefore, the 3d reconstruction is based on a sequence of images of the object taken from different viewpoints with two different algorithms; shape from silhouette and shape from structured light. The output of both algorithms are then used to construct a single 3d model. Results of the algorithm developed are presented for both synthetic and real input images.

Active vision in robotic systems: A survey of recent developments

Abstract: In this paper we provide a broad survey of developments in active vision in robotic applications over the last 15 years. With increasing demand for robotic automation, research in this area has received much attention. Among the many factors that can be attributed to a high-performance robotic system, the planned sensing or acquisition of perceptions on the operating environment is a crucial component. The aim of sensor planning is to determine the pose and settings of vision sensors for undertaking a vision-based task that usually requires obtaining multiple views of the object to be manipulated. Planning for robot vision is a complex problem for an active system due to its sensing uncertainty and environmental uncertainty. This paper describes such problems arising from many applications, e.g. object recognition and modeling, site reconstruction and inspection, surveillance, tracking and search, as well as robotic manipulation and assembly, localization and mapping, navigation and exploration. A bundle of solutions and methods have been proposed to solve these problems in the past. They are summarized in this review while enabling readers to easily refer solution methods for practical applications. Representative contributions, their evaluations, analyses, and future research trends are also addressed in an abstract level.

Next best view planning for active model improvement

Abstract: We propose a novel approach to determining the Next Best View (NBV) for the task of efficiently building highly accurate 3D models from images. Our proposed method deploys a hierarchical uncertainty driven model refinement process designed to select vantage viewpoints based on the model’s covariance structure and appearance, as well as the camera characteristics. The developed NBV planning system incrementally builds a sensing strategy by sequentially finding the single camera placement, which best reduces an existing model’s 3D uncertainty. The generic nature of our system’s design and internal data representation makes it well suited to be applied to a wide variety of 3D modeling algorithms. It can be used within active computer vision systems as well as for optimized view selection from the set of available views. Experimental results are presented to illustrate the effectiveness and versatility of our approach.