1,721,284 research outputs found
MPC-based cooperative transportation: towards real-world dynamic environments
Full text linkopenBasandosi sul lavoro del professor Cenedese Angelo e dei suoi colleghi, intitolato “Model Predictive Control for Cooperative Transportation with Feasibility-Aware Policy”, che propone un approccio basato su MPC per la navigazione cooperativa di un sistema multi-robot in ambienti con ostacoli statici e dinamici, questo studio adotta il loro obiettivo di progetto come base per la progettazione di un nuovo sistema. Nel nuovo framework, i robot omnidirezionali sono sostituiti da robot a guida differenziale e viene introdotta una politica più sofisticata per la gestione dei casi limite. Questo documento spiega le motivazioni alla base di questi cambiamenti e descrive il flusso di sviluppo, supportato da simulazioni condotte in MATLAB. Presenta inoltre esperimenti in cui viene iniettato rumore artificiale nel sistema e si conclude delineando le basi per una futura implementazione in tempo reale su piattaforme robotiche fisicheBuilding on the work of Professor Cenedese Angelo and colleagues, titled “Model Predictive Control for Cooperative Transportation with Feasibility-Aware Policy,” which proposes an MPC-based approach for cooperative navigation of a Multi-Robot System in environments with static and dynamic obstacles, this study adopts their project goal as a foundation to design a new system. In the new framework, omnidirectional robots are replaced by differential-drive robots, and a more sophisticated policy for managing edge cases is introduced. This document explains the motivations behind these changes and describes the development workflow, supported by simulations conducted in MATLAB. It also presents experiments where artificial noise is injected into the system, and concludes by outlining the groundwork for a future real-time implementation on physical robotic platforms
On triangulation algorithms in large scale camera network systems
No full textGeometric triangulation is at the basis of the estimation of the 3D position of a target from a set of camera measurements. The problem of optimal estimation (minimizing the L2 norm) of the target position from multi-view perspective projective measurements is typically a hard problem to solve. In literature there are different types of algorithms for this purpose, based for example on the exhaustive check of all the local minima of a proper eigenvalue problem [2], or branch- and-bound techniques [3]. However, such methods typically become unfeasible for real time applications when the number of cameras and targets become large, calling for the definition of approximate procedures to solve the reconstruction problem.
In the first part of this paper, linear (fast) algorithms, computing an approximate solution to such problems, are described and compared in simulation. Then, in the second part, a Gaussian approximation to the measurement error is used to express the reconstruction error’s standard deviation as a function of the position of the reconstructed point. An upper bound, valid over all the target domain, to this expression is obtained for a case of interest. Such upper bound allows to compute a number of cameras sufficient to obtain a user defined level of position estimation accuracy
A ROBUST ACTIVE CONTOUR APPROACH FOR STUDYING CELL DEFORMATION FROM NOISY IMAGES
No full textThis work presents a generalized formulation of the Snake model defining new terms for the internal and the external energy functionals. These modifications conjugate features of the object contour as well as the inside of the shape. The obtained model is significantly more accurate spatially on the image plane and temporally on the frame sequence. In particular, the application to single cell analysis is in focus: In this context, we show how to cast the specific problem into the extended framework we propose. Shape descriptors and suitable metrics are then derived from the curve representation. The boundary identification produced through the classic formulation shows a poor and imprecise segmentation and leads to misleading metrics. The new model instead represents the boundary and the derived shape parameters in a way more consistent with the visual perception of shape evolution and deformation
Teseo: A Multi-Agent Tracking Application in Wireless Sensor Networks
No full textIn this work the design and implementation of an application to track multiple agents in a indoor Wireless Sensor Actor Network (WSAN) is proposed. We developed a tracking algorithm that falls into the category of the radio frequency localization/tracking methods, that exploit the strength of the wireless communications among fixed and mobile agents to establish the position of the mobile ones. The algorithm resorts to an Extended Kalman Filter to process the agents measurements and reach a desired level of tracking performance. The tracking application, namely Teseo, is composed by a low-level NesC management software for the agents side and a Java graphical interface provided to users connected to mobile agents. A detailed description of the operations performed by Teseo is given, accompanied both by simulations to validate the tracking algorithm and experiments on a real testbed to test Teseo
Structure-based approach for optimizing distributed reconstruction in Motion Capture systems
No full textThe diffusion of visual sensor networks, and in particular of smart camera networks, is motivating an increasing interest on the research of distributed solutions for several vision problems. Specifically, in this paper we propose a distributed solution to the problem of reconstructing target positions in large Motion Capture (MoCap) systems. Real time reconstruction by means of centralized procedures is practically unfeasible for very large systems, while the use of distributed computation allows to significantly reduce the computational time required for reconstruction, thus allowing the development of real time solutions.
Then the proposed distributed reconstruction procedure is optimized by exploiting information about the structure of the system: the visibility matrix states which objects in the scene are somehow measurable by a sensor (sensor-object matrix). Often, the typical localization of data from real application scenarios induces an underlying structure on the visibility matrix, that can be exploited to improve the performance of the system in understanding the surrounding environment. Unfortunately, usually these data are not properly organized in the visibility matrix: for instance, listing the sensors in a pseudo-random order can hide the underlying structure of the matrix. This paper considers the problem of recovering such underlying structure directly from the visibility matrix and designs an algorithm to perform this task.
Our simulations show that the distributed reconstruction algorithm optimized by means of the estimation of the structure of the visibility matrix allows a particularly relevant computational time reduction with respect to the standard (centralized) reconstruction algorithm
Multi-agent tracking in wireless sensor networks: implementation
No full textIn this work the design and implementation of an application to track multiple agents in a indoor Wireless Sensor Actor Network (WSAN) is proposed. The adopted embedded hardware for the network nodes is theTmote Sky, an ultra low power IEEE 802.15.4 compliant wireless device, which has become a reference in the academia for the early development of algorithms and applications for Wireless Sensor Actor Networks (WSANs). These devices are based on the TinyOS operative system and are programmed in NesC a C-derived language specifically developed for embedded systems. NesC has become indispensable for low-level management ofindividual agents while Java was chosen to provide the user with a simple and intuitive graphical interface with whom showing and coordinating the tracking
Multi-agent tracking in wireless sensor networks: model and algorithm
No full textIn this work an algorithm to track multiple agents in an indoor Wireless Sensor Actor Network (WSAN) is proposed. The algorithm falls into the category of the radio frequency localization methods, since it exploits the strength of the wireless communications among nodes to establish the position of a set of mobile nodes within a network of fixed nodes placed in known locations. In this sense, a radio channel model is introduced that allows to estimate the distances among nodes to attain localization and tracking (range-based approach). Moreover, to compensate for the scant robustness of power measurements, the loss effects induced by wireless communication,the intrinsic uncertainty of unstructured environments, the algorithm resorts to an Extended Kalman Filter to process the node measurements and reach a desired level of localization performance. Finally, the design phase is validated through the implementation and the experiments on a real testbed
Channel Model Identification in Wireless Sensor Networks Using a Fully Distributed Quantized Consensus Algorithm
Full text linkIn this paper, we consider the problem of designing a distributed strategy to estimate the channel parameters for a generic Wireless Sensor-Actor Network (WSAN). To this aim, we present a distributed least-square algorithm that complies with the constraint of transmitting only integer data through the wireless communication, which often characterizes WSAN embedded architectures. In this respect, we propose a quantized consensus strategy that mitigates the effects of the rounding operations applied to the wireless exchanged floating data. Moreover, the approach is based on a symmetric random gossip strategy, making it suitable for the actual deployment in multiagent networks. Finally, the effectiveness of the proposed algorithm and of its implementation as an open-source application is assessed and the employment of the procedure is illustrated through the application to radio-frequency localization experiments in a real world testbed
Multistep hybrid learning: CNN driven by spatial–temporal features for faults detection on metallic surfaces
No full textSolutions for the quality control of metallic surfaces are proposed. Specifically, we study a deflectometric apparatus based on coaxial structured light and the related algorithmic procedure, which is able to detect the faulty surface of a sample captured by a video sequence. First, by considering the metallic surface a dynamic scene illuminated under different light conditions, we develop the descriptor residuals of linear evolution of light (RLEL) that extracts the defectiveness information starting from the movement of the object without explicitly considering the physical characteristics of the light structure. Then, leveraging on RLEL, we present a hybrid learning (HL) technique capable of overcoming the data-driven approach used in classic deep learning (DL). By exploiting a multisteps training process, we combine the model-based descriptor RLEL and a classical data-driven convolutional neural network (CNN) to obtain an unconventional gray-box CNN, which exceeds the performance of popular DL solutions such as 3-D-inception and 3-D-residual DL networks. Remarkably, HL also shows its validity in comparing the performance of the same network structures trained not in a hybrid way, namely without the injection of the model-based information given by RLEL
How to Represent the Shape of a Deformable Object and Ease the Control of the Deformation?
No full textThe research on the reconstruction and the representation of the shape of objects in motion is nowadays particularly active, both on the theoretical ground and on the field of applications. It is particularly active the discussion present in the scientific community on how to make the mathematical formalism of shape representation compatible with the intuitiveness and perception of physical phenomena. In this paper we pose the problem of how to describe the shape of an object through the definition of a framework that allows not only the analysis but also the control of the shape of objects under continuous deformation
- …
