International and National Journals
A. Balsamo, A. Chimienti, P. Grattoni, R. Nerino, G. Pettiti, M. L. Rastello, M. Spertino
Active vision applications to cultural heritage acquisition and monitoring
Journal of Cultural Heritage, Vol. 7, no. 2, pp. 98-109, 2006.
This paper presents features and performances of a portable stereo active vision system (AVS), designed and built for data acquisition and monitoring of medium-scale monuments and environments of interest for cultural heritage. The research activity was partially carried out in the framework of PARNASO-SIINDA project, a national project aimed at the development of new techniques for cultural heritage preservation, and the case-of-study of the activity was the Roman Theatre of Aosta (IT). Reliable data acquisition and monitoring of the surface alterations related to the monument degrade was a project demand that motivated the stereo vision approach because of its high spatial sampling rate, high accuracy and close correlation of shape and color measurements. The AVS system architecture is designed around three CCD cameras mounted on three independent pan-tilt stages under computer control. The cameras are equipped with different focal lens objectives, allowing the system to acquire both geometric and colorimetric information of an architectural surface (monuments, frescoes…) at different scales over areas of tens of square meters. Two cameras form a stereo pair and are equipped with long-focal objectives to sample a limited portion of the scene at high spatial resolution, while the third one is equipped with a wide-angle objective and it is intended for surveying the whole scene at low-resolution. Geometric measurements are carried out both by forward triangulation of the optical axes of the stereo pair during fixation of an object and by three-dimensional (3D) object reconstruction from the images of the stereo pair. The fixation process is performed by a PC-controlled imagebased servo loop. The fixation approach supplies sparse but accurate 3D measurement in the AVS reference system, while the stereo acquisition of image pairs allows a dense 3D local reconstruction of the surface around the fixation point. Colorimetric measurements are performed at pixel level by processing the images of one of the stereo cameras, which is equipped with tristimulus filters. Results are supplied in terms of CIE-Lab components. A user-friendly interface for the system management allows performing several measurements automatically. On-site data processing allows the user to decide for further investigations immediately, if necessary, without expensive delays. The features of the AVS system, namely the working range of depths (2-10 m), the high spatial sampling rate, the accuracy and the close correlation of shape and color measurements, makes the system competitive with respect to other available instruments and techniques for data acquisition and monitoring of small and medium-scale works of art. In the following the AVS architecture is presented, and laboratory tests made to asses its performances are described. Results of the tests and of the acquisition campaign, made on the Roman Theatre of Aosta (IT), are reported and discussed in comparison with other techniques commonly employed in cultural heritage surveys.
A. Cumani, S. Denasi, A. Guiducci, P. Lanza, G. Quaglia
Visual pose registering in space robotics applications
WSEAS Transactions on Circuits and Systems, Vol. 5, no. 10, pp. 1550-1555, 2006.
This work sketches the vision system designed for the Eurobot Flight Model, a robotic system planned by ESA to assist astronauts on the International Space Station during extra-vehicular activities. In particular, the algorithm for registering the robot's pose with respect to the station is presented, as well as some preliminary results from laboratory simulations.
A. Cumani, A. Guiducci
Stereo-based visual odometry for robust rover navigation
WSEAS Transactions on Circuits and Systems, Vol. 5, no. 10, pp. 1556-1562, 2006.
This work proposes a method for estimating the trajectory of an autonomous rover relying on passive stereo vision only. This visual odometry estimate is obtained by tracking pointwise image features and then evaluating the robot motion by robust bundle adjustment of the stereo matched features, before and after the motion. Some results of the application of the algorithm, both to simulated data and to actual image sequences acquired by a mobile platform, are presented and discussed.
R. Nerino
Automatic registration of point-based surfaces
WSEAS Transactions on Computers, Vol. 5, no. 12, pp. 2984-2991, 2006.
Shape modelling of 3D (three-dimensional) structures is an important issue in several fields, among other, manufacturing and quality control, reverse engineering, medicine and cultural heritage. In the first step of the shape modelling pipeline, the data acquisition, the increasing use of range measurement devices as laser scanners, structured light and stereo systems makes available large amounts of data sets as clouds of 3D points, or scans. An important issue in the modelling pipeline is the alignment, or registration, of partially overlapping scans in a common coordinate system. A new approach to automatic coarse pair-wise registration of partially overlapping scans is presented in this paper. The approach is based on the characterization of the pair of scans by two sets of sparse feature points invariant to Euclidean transformations, which are robustly extracted from the data by means of a multiscale procedure. The feature points are grouped in sets of triplets and the triplets are then characterized by signatures, which are vectors of parameters with the same invariance properties of the feature points. Pairs of triplets are matched in the signature space according to their signature distance. The alignment transformation able to register the two scan pairs is then estimated from the matched pairs of triplets. A verification process asses the quality of the match, which is defined as the number of feature points put in correct correspondence by the estimated transformation associated to a matched pair. This number of features must be greater than a prescribed threshold, defined by the expected percentage of overlapping of the two data sets. The proposed approach to registration has been evaluated on original scan sets and on standard ones available on the web. Preliminary results, presented and discussed in the paper, confirm the validity of the approach. Other work has to be done to analyze the robustness of the approach to noisy data and to optimize the computational time.
Conference Proceedings
A. Balsamo, A. Chimienti, P. Grattoni, R. Nerino, G. Pettiti, M. L. Rastello, M. Spertino
An active vision system for 3D surface colour measurements
Proceedings of the ISCC/CIE Expert Symposium, Ottawa (Canada), May 2006.
Common surfaces have reflection characteristics that differ considerably from those of a reference standard for colorimetry. They are neither totally diffusing nor regularly reflecting, and their reflectance strongly depends on the viewing angle and the illumination geometry. As a consequence, reliable measurements can be achieved only if the measuring geometry is fixed or known. To solve this problem, the Commission Internationale de l'Eclairage (CIE) recommended four standard geometries, defining both irradiation and observation conditions. Unfortunately, when dealing with 3D objects, with large dimensions in space, like for instance monuments or automobiles, geometry can hardly controlled and new-concept instrumentation is needed to obtain results reproducible in different times and/or locations. Traditional instrumentation on the market rarely offers colorimetric and geometric measurements combined in a single device and when it happens one function is just a support to the other without any accuracy indication. In fact, accurate geometric and colorimetric data permit detecting changes of surfaces at a given resolution (e.g. erosion, mould growth, chemical alterations) when these data are strictly correlated for effective surveying analyses. Moreover, the assessment of the measurement accuracy would allow establishing a possible correlation between the geometric and colorimetric data of surfaces and the chemical-physical changes of the surrounding environment with a number of possible implications of interest. Concerning geometric measurements, devices and techniques for the acquisition of three-dimensional structures can be grouped in three categories: topographic, photogrammetric and laser-based techniques. Due to their physical working principle, each of them has a specific range of operation within which it supplies its best performances. In particular, all techniques merged together leave almost uncovered the range from two - three meters to twelve-fifteen meters. Concerning colorimetric measurements, techniques and devices on the market can perform accurate measurements on single spots of at least some mm in size, or more. Colorimeters can be roughly grouped in two categories based on their measuring characteristics: in-contact and not in-contact. In-contact devices are equipped with an internal calibrated source of light while the other ones need external sources of light, possibly satisfying CIE recommendation on their spectral content. Even if all devices require less than some seconds to carry out a single measurement, the dense sampling of a surface can be very time-consuming because of the time needed for repositioning. In addition, the color-to geometry correlation is not immediate and can be difficult establishing it. In this context, there is a definite demand for a flexible, multifunctional (geometric and colorimetric) instrumentation which should integrate the traditional peculiar ones to easier the on-site data collection and promote its diffusion. The Active Vision System (AVS) described in this paper has been designed and developed to answer these needs. AVS works over a range from two to ten meters in depth and carries out integrated colorimetric and geometric measurements with assessed accuracies. Thanks to its computerized control for the automatic management of the operations, it allows the in-field processing of the acquired data and their comparison with databases for onitoring purposes. The global functions of the SVA are essentially: the measurement of the 3D position of a point in the scene of the imaged surface and the measurement of the tristimulus values of this point. Then, the whole large scene is reconstructed with high resolution by scene tessellation and image mosaicing, and the 3D surface is obtained from sparse points or by dense reconstruction from the stereo TLs images. All these functions are automatically performed under computer control and a man-machine graphical interface has been developed for managing the whole system easily. The acquisition and registration of accurate geometric and colorimetric parameters concerning a given survey, such as the relative position between AVS and scene, the spatial co-ordinates of the test points, and the spatial position of artificial light sources, allow the system to carry out automatic and accurate repetition of that survey in successive measuring campaigns. AVS is composed of three B/W TV cameras aligned along a common axis γ . Two of these cameras (TL1 and TL2) are equipped with long focal-length lenses to frame only small portions of a scene at high resolution. They can be rotated by known angles both around the parallel pan axes α1 and α2 and the tilt axis γ, to perform the fixation of some points of a scene. One TL is equipped with spectral filters, allowing the acquisition of high accuracy colour images of the examined surface. The third camera is equipped with a wide-angle lens (WA) to frame the whole region of interest at a lower resolution. When the field size framed by TL cameras is too small for analysis, a wider field can be acquired as a sequence of partially overlapping tiles by automatically scanning the Region Of Interest (ROI) with TLs. The ROI can be interactively selected by an operator looking at the WA image on the computer display. In addition to texture and colour information, the spatial position, orientation and gaze direction of each tile are acquired at each step so that the entire framing geometry is completely controlled by the system and can be saved for reliable repetitions of the measurements at different times, i.e. for monitoring tasks. The reconstruction of the whole ROI information is obtained by image mosaicing. The mosaicing procedure differs from the ones described in literature mainly because the transformation of the acquired images is aimed at compensating systematic acquisition errors and parallax effects independently of the image contents, and not at minimizing the matching errors between adjacent images. Therefore, this method is intrinsically non-iterative and offers the advantages of being simple and fast, but accurate enough to satisfy the application requirements.
P. Lanza, A. Cumani, S. Denasi, A. Guiducci, G. Quaglia
Image Processing applied on the WET robot prototype
Proceedings of the International Space System Engineering Conference DASIA 2006, Berlin (Germany), May 2006.
R. Nerino
Invariant features for automatic coarse registration of point-based surfaces
6th WSEAS International Conference on Signal Processing, Computational Geometry and Artificial Vision (ISCGAV'06), Elounda, Agios Nikolaos (Greece), August 2006 (on cd rom).
Automatic coarse alignment, or registration, of partially overlapping three dimensional (3D) shapes is a fundamental problem of the shape acquisition and modelling pipeline. This paper describes a new approach to automatic coarse pair-wise registration of partially overlapping 3D point sets which are commonly generated by laser scanners, structured light systems or stereo. The approach is based on the characterization of the two 3D point clouds by sparse feature points invariant to Euclidean transformations, which are robustly extracted from the data by a multiscale procedure. The feature points are grouped in sets of triplets and the triplets are then characterized by signatures, which are vectors of parameters with the same invariance properties of the feature points. A matching procedure selects probable corresponding triplets in the signature space according to their signature distance. The alignment transformation able to register the two scan pairs is estimated from the matched triplets. A verification process asses the quality of match, which is defined as the number of feature points put in correct correspondence by the estimated transformation associated to a matched pair. This number of features must be greater than a prescribed threshold, defined by the expected percentage of overlapping of the two data sets. The proposed approach to registration has been evaluated on standard scan sets available on the web, and some preliminary results, presented and discussed in the paper, confirm its validity. Other work has to be done to analyze the robustness of the approach to noisy data and to optimize the computational time.
A. Cumani, S. Denasi, A. Guiducci, P. Lanza, G. Quaglia
Vision on the Eurobot WET prototype
Proceedings of the TAROS 2006 Conference - Towards Autonomous Robotic Systems, pp. 40-45, Guildford (UK), September 2006.
This paper sketches the vision system designed to equip ESA's Eurobot Flight Model, a robotic system planned to assist astronauts on the International Space Station during extra-vehicular activities. The vision algorithms for registering the robot's pose with respect to the station are presented, together with some preliminary results from laboratory simulations.
A. Cumani, A. Guiducci
Visual odometry for robust rover navigation by binocular stereo
Proceedings of the 6th WSEAS International Conference on Signal, Speech and Image Processing (SSIP-06), pp. 74-79, Lisbona (Portugal), September 2006.
This work deals with the problem of estimating the trajectory of an autonomous rover by passive stereo vision only (visual odometry). The proposed method relies on the tracking of pointwise image features and on the estimation of the robot motion by robust bundle adjustment of the stereo matched features, before and after the motion. Preliminary results from the application of the algorithm both to simulated data and to actual image sequences acquired by a mobile platform are presented and discussed.
A. Cumani, S. Denasi, A. Guiducci, P. Lanza, G. Quaglia
Visual registering of arm pose: a space robotics application
Proceedings of SSIP-06, pp. 80-84, 2006.
This paper deals with the vision algorithms designed for ESA's Eurobot Flight Model, a robotic system planned to assist astronauts on the International Space Station during extra-vehicular activities. The algorithm for registering the robot's pose with respect to the station is presented, together with some preliminary results from laboratory simulations.
M. Cadoni, A. Chimienti, R. Nerino
Automatic coarse registration by invariant features
7th International Symposium on Virtual Reality, Archaeology and Cultural Heritage, VAST 2006, Cyprus, October 2006.
The increasing availability of relatively low-cost range sensors such as scanner lasers and structured light systems in cultural heritage applications has dramatically changed the traditional approaches to the documentation, monitoring and fruition of cultural heritage findings. Three-dimensional shape modelling is often the final goal of the processing pipeline which starts from the acquisition of overlapping scans of the entire work of art. An important step of the processing pipeline is the optimal alignment of the scan set in a common coordinate system, the so called registration step. This paper presents a new feature-based approach to the coarse registration between partially overlapping range images. Our approach extracts from the range images "feature points" and then characterises them by invariants to Euclidean transformations. The novelty of the approach is that the choice and the design of the invariants is supported by the theory of moving frames recently developed by J.Olver. This provides us with an algorithm to find the fundamental sets of invariants necessary to parameterise a signature manifold that characterises the original manifold up to Euclidean transformations. To maximise performance against noise we can design invariants that depend on distances and 1st order derivatives only. To reduce the overall computational complexity the invariant are not estimated on all the points of the scans, but only on a reduced subset of them. This subset, the feature points, are determined by the Gaussian curvature maxima of the surface underlying the data. Preliminary results on standard 3D data sets from web repositories and on original scans of works of art show the effectiveness of the proposed registration algorithm.