1,720,991 research outputs found
Phase difference stereo disparity computation on a SIMD parallel machine
No full textA parallel version of the phase-based algorithm for disparity estimation in stereo image pairs for the reconstruction of the third dimension is presented. The algorithm is implemented on the Quadrics massively parallel SIMD machine. An analysis of performance as a function of image size and processors number is given. The obtained processing times are compared with two other HW architectures both sequential and parallel. © 1997 Springer-Verlag Berlin Heidelberg
HIPERCLASS: High performance industrial inspection and defect classification in steel industry
No full textHIPERCLASS is a project of the European ESPRIT Special Action named CAPRI. This Special Action is aimed at the diffusion of the Parallel Computing among the italian industries. In such environment, the project HIPERCLASS is directed to the porting of the image treatment and defect classification of a surface inspection pilot system for steel rolled strip developed by CSM, from the hardwired implementation to the software environment of the Quadrics parallel supercomputer. The resulting SW implementation offers both performance and a large flexibility of the system as compared to the dedicated electronic circuitry. The prototype developed on this supercomputer is able to sustain a data rate around 15 Mpixel/s, performing the image processing, the defect detection and the defect neural classification. © 1997 Springer-Verlag Berlin Heidelberg
Traffic Request Generation through a Variational Auto Encoder Approach
Full text linkTraffic and transportation forecasting is a key issue in urban planning aimed to provide a greener and more sustainable environment to residents. Their privacy is a second key issue that requires synthetic travel data. A possible solution is offered by generative models. Here, a variational autoencoder architecture has been trained on a floating car dataset in order to grasp the statistical features of the traffic demand in the city of Rome. The architecture is based on multilayer dense neural networks for encoding and decoding parts. A brief analysis of parameter influence is conducted. The generated trajectories are compared with those in the dataset. The resulting reconstructed synthetic data are employed to compute the traffic fluxes and geographic distribution of parked cars. Further work directions are provided
A CNN-based passive optical range finder for real time robotic applications
No full textThe paper presents a new CNN for real-time stereo vision, useful as a passive optical range finder for autonomous robots and vehicles. The stereo matching as energy minimization is discussed and former neural approaches to the problem are analyzed. Experimental results with the new CNN both with synthetic and real images are reported, demonstrating the performance of the system
A Mobile Small Sized Device for Air Pollutants Monitoring Connected to the Smart Road: Preliminary Results
No full textThe work in progress on a small sized air pollution monitoring system mountable on board urban vehicles is described. The system exchanges data exploiting a “Smart Road” infrastructure with a central computing facility, the Smart City Platform, a GIS-based Decision Support System designed to perform real time monitoring and interpolation of data with the aim of possibly issuing alarms with respect to different town areas. Early experimental data gathering in the Rome urban area and subsequent spline interpolation processing are presented. Thus, air pollutants distribution maps have been produced. Finally, protocols for data exchange have been designed. Work is in progress on algorithms for data fusion among different monitoring systems and interpolation of data for a geographically denser map
A CNN-based passive optical range finder for real-time robotic applications
No full textThe paper presents a new cellular neural network cellular neural network (CNN) for real-time stereo vision, useful as a passive optical range finder for autonomous robots and vehicles. The stereo matching as energy minimization is discussed, and former neural approaches to the problem are analyzed. Experimental results with the new CNN both with synthetic and real images are reported, demonstrating the performance of the system
Swarm Underwater Acoustic 3D Localization: Kalman vs Monte Carlo
Full text linkTwo three-dimensional localization algorithms for a swarm of underwater vehicles are presented. The first is grounded on an extended Kalman filter (EKF) scheme used to fuse some proprioceptive data such as the vessel's speed and some exteroceptive measurements such as the time of flight (TOF) sonar distance of the companion vessels. The second is a Monte Carlo particle filter localization processing the same sensory data suite. The results of several simulations using the two approaches are presented, with comparison. The case of a supporting surface vessel is also considered. An analysis of the robustness of the two approaches against some system parameters is given
Visual and laser sensory data fusion for outdoor robot localisation and navigation
No full textAn architecture for robot localization and navigation performing fusion among odometry, laser range data and range from a neural stereoscopic vision system is presented. The estimate robot position is used to safely navigate through the environment. The stereoscopic sub system delivers dense information in the entire field of view. This feature allows a safer navigation of the robot, since, for example, arch-like obstacles may be avoided. Experimental results are presented concerning localization and obstacle detection. The precision attained by the system allows safe navigation. The use of visual data is of great importance in many operative scenario
Urban Air Pollutant Monitoring through a Low-Cost Mobile Device Connected to a Smart Road
Full text linkAir pollutant monitoring is a basic issue in contemporary urban life. This paper describes an approach based on the diffused use of low-cost sensors that can be mounted on board urban vehicles for more abundant and distributed measures. The system exchanges data, exploiting a “Smart Road” infrastructure, with a central computing facility, the CIPCast platform, a GIS-based Decision Support System designed to perform real-time monitoring and interpolation of data with the aim of possibly issuing alarms with respect to different town areas. Experimental data gathering in the Rome urban area and subsequent processing results are presented. Algorithms for data fusion among different simulated monitoring systems and interpolation of data for a geographically denser map were utilised. Thus, in the framework of the Smart Road, protocols for data exchange were designed. Finally, air pollutant distribution maps were produced and integrated into the CIPCast platform. The feasibility of a full system architecture from the sensors to the real-time pollutant maps is shown
New board for CNN stereo vision algorithm
No full textArtificial vision for environment recognition is a very useful tool in autonomous robotics. Specifically the use of stereo vision algorithms implemented via a hardware neural architecture allows real time scene reconstruction. In this paper the follow-on of previous work on a analogue hardware Cellular Neural Network implementation of the algorithm is presented. In this paper a new CNN based PCI electronic board will be presented
- …
