Search

Low-cost 3D sensors are nowadays widely diffused and many different solutions are available on the market. Some of these devices were developed for entertaining purposes, but are used also for acquisition and processing of different 3D data with the aim of documentation, research and study. Given the fact that these sensors were not developed for this purpose, it is necessary to evaluate their use in the capturing process. This paper shows a preliminary research comparing the Kinect 1 and 2 by Microsoft, the Structure Sensor by Occipital and the O&P Scan by Rodin4D in a medical scenario (i.e. human body scans). In particular, these sensors were compared to Minolta Vivid 9i, chosen as reference because of its higher accuracy. Different test objects were analysed: a calibrated flat plane, for the evaluation of the systematic distance error for each device, and three different parts of a mannequin, used as samples of human body parts. The results showed that the use of a certified flat plane is a good starting point in characterizing the sensors, but a complete analysis with objects similar to the ones of the real context of application is required. For example, the Kinect 2 presented the best results among the low-cost sensors on the flat plane, while the Structure Sensor was more reliable on the mannequin parts.

Since the advent of the first Kinect as motion controller device for the Microsoft XBOX platform (November 2010), several similar active and low-cost range sensing devices have been introduced on the mass-market for several purposes, including gesture based interfaces, 3D multimedia interaction, robot navigation, finger tracking, 3D body scanning for garment design and proximity sensors for automotive. However, given their capability to generate a real time stream of range images, these has been used in some projects also as general purpose range devices, with performances that for some applications might be satisfying. This paper shows the working principle of the various devices, analyzing them in terms of systematic errors and random errors for exploring the applicability of them in standard 3D capturing problems. Five actual devices have been tested featuring three different technologies: i) Kinect V1 by Microsoft, Structure Sensor by Occipital, and Xtion PRO by ASUS, all based on different implementations of the Primesense sensor; ii) F200 by Intel/Creative, implementing the Realsense pattern projection technology; Kinect V2 by Microsoft, equipped with the Canesta TOF Camera. A critical analysis of the results tries first of all to compare them, and secondarily to focus the range of applications for which such devices could actually work as a viable solution.

Cultural Heritage documentation and preservation has become a fundamental concern in this historical period. 3D modelling offers a perfect aid to record ancient buildings and artefacts and can be used as a valid starting point for restoration, conservation and structural analysis, which can be performed by using Finite Element Methods (FEA). The models derived from reality-based techniques, made up of the exterior surfaces of the objects captured at high resolution, are - for this reason - made of millions of polygons. Such meshes are not directly usable in structural analysis packages and need to be properly pre-processed in order to be transformed in volumetric meshes suitable for FEA. In addition, dealing with ancient objects, a proper segmentation of 3D volumetric models is needed to analyse the behaviour of the structure with the most suitable level of detail for the different sections of the structure under analysis. Segmentation of 3D models is still an open issue, especially when dealing with ancient, complicated and geometrically complex objects that imply the presence of anomalies and gaps, due to environmental agents such as earthquakes, pollution, wind and rain, or human factors. The aims of this paper is to critically analyse some of the different methodologies and algorithms available to segment a 3D point cloud or a mesh, identifying difficulties and problems by showing examples on different structures.