Francesco Leporati

Real-Time Implementation of the Vertex Component Analysis Algorithm on GPUs

In this letter, we present a new parallel implementation of the vertex component analysis (VCA) algorithm for spectral unmixing of remotely sensed hyperspectral data on commodity graphics processing units. We first developed a C serial version of the VCA algorithm and three parallel versions: one using NVIDIAs Compute Unified Device Architecture (CUDA), another using CUDA basic linear algebra subroutines library CUBLAS, and the last using the CUDA linear algebra library CULA. Experimental results, based on the analysis of hyperspectral images acquired by a variety of hyperspectral imaging sensors, show the effectiveness of our implementation, which satisfies the real-time constraints given by the data acquisition rate.

A Hybrid CPU–GPU Real-Time Hyperspectral Unmixing Chain

Hyperspectral images are used in different applications in Earth and space science, and many of these applications exhibit real- or near real-time constraints. A problem when analyzing hyperspectral images is that their spatial resolution is generally not enough to separate different spectrally pure constituents (endmembers); as a result, several of them can be found in the same pixel. Spectral unmixing is an important technique for hyperspectral data exploitation, aimed at finding the spectral signatures of the endmembers and their associated abundance fractions. The development of techniques able to provide unmixing results in real-time is a long desired goal in the hyperspectral imaging community. In this paper, we describe a real-time hyperspectral unmixing chain based on three main steps: 1) estimation of the number of endmembers using the hyperspectral subspace identification with minimum error (HySime); 2) estimation of the spectral signatures of the endmembers using the vertex component analysis (VCA); and 3) unconstrained abundance estimation. We have developed new parallel implementations of the aforementioned algorithms and assembled them in a fully operative real-time unmixing chain using graphics processing units (GPUs), exploiting NVIDIAs compute unified device architecture (CUDA) and its basic linear algebra subroutines (CuBLAS) library, as well as OpenMP and BLAS for multicore parallelization. As a result, our real-time chain exploits both CPU (multicore) and GPU paradigms in the optimization. Our experiments reveal that this hybrid GPU-CPU parallel implementation fully meets real-time constraints in hyperspectral imaging applications.

An FPGA-Based Embedded System for Fingerprint Matching Using Phase-Only Correlation Algorithm

Biometric identification systems are defined as systems exploiting automated methods of personal recognition based on physiological or behavioural characteristics. Among these, fingerprints are very reliable biometric identifiers. Trying to fasten the image processing step makes the recognition process more efficient, especially concerning embedded systems for real-time authentication. In this paper we propose an FPGA-based architecture that efficiently implements the high computationally demanding core of a matching algorithm based on phase-only spatial correlation. Moreover, we show how it is possible to use COTS components to embed an entire AFIS on chip and so reducing cost, space and energy used.

A Localized Surface Plasmon Resonance-Based Portable Instrument for Quick On-Site Biomolecular Detection

In recent years, several approaches have been developed to carry out biosensors based on localized surface plasmon resonance (LSPR). However, the high costs of nanostructure fabrication and the absence of autonomous portable devices strongly limit the extensive use of LSPR biosensors outside the research laboratories. We designed, implemented, and tested a novel low cost multiparametric stand-alone LSPR imaging instrument for biosensing applications. This compact device (15 × 6 × 17 cm3 size and <;500-g weight) consists of a nanohole array biochip integrated with a microfluidic layer and a processing system. An optical apparatus focuses a light beam from an IR LED source and a digital image sensor captures the reflected light from the biochip surface. The signals are processed by the embedded ARM processor and shown on a touchscreen display by a user-friendly application, without the need for other external computational devices. Moreover, we propose an extremely simple analytical method to reduce the image noise without any sophisticated temperature control or external luminosity change compensation. The device sensitivity of

Data acquisition system for measurements in free moving subjects and its applications

6 \times 10^{\mathrm {-5}}

An Accelerator for Physics Simulations

refractive index unit was measured using glycerol solutions with different concentrations. We demonstrated the efficiency of our system in biomolecular detection by monitoring the Ab-PTX3 antibody in a test that showed the instruments potentialities in the detection of antibodies. These results confirmed the potential usefulness of the proposed system in several biomedical applications such as medical diagnostic procedures, immunoassays, or fast in loco preliminary analyses without the aid of specialized laboratory or trained personnel.

Flow Rate Profiler: an instrument to measure blood velocity profiles.

A low-cost, portable acquisition system for monitoring and processing human biomechanical parameters is presented. It is equipped with 16 input channels, each one linked to an external transducer by a suitable connector. Input signals from sensors are converted into a digital form by a 12-bit analog-to-digital converter and stored in a removable memory (memory card) respecting the PCMCIA standard interface, allowing the download of acquired data toward the host computer. The acquisition operating mode is programmable by a host PC, writing proper values into the memory card; then, the instrument acquires the defined number of channels at the selected sampling rate. The instrument is battery powered; then, it can be used in all those applications, like rehabilitation and sports medicine, where the freedom for subject movement is a constraint for the test. In fact, this instrument does not require an arranged environment for measurements, and it is not connected to a PC. Three sample applications are presented in which the instrument is used to evaluate human motor capability, physical parameters in amputees, and motor performance in athletes.

The Human Brain Project

An accelerator based on field-programmable gate array (FPGA) technology accelerates double-precision floating-point operations present in the energy calculation of Monte Carlo-Metropolis simulations. The accelerator uses COTS components and is scalable in terms of clock frequency, memory capability, and number of computing units. It could also be part of a cluster of accelerated workstations.

An embedded multi-core biometric identification system

In this paper we present Flow Rate Profiler (FRP), an instrument for measuring the blood velocity by means of ultrasound-based techniques. The velocity is directly related to the shear rate, which is in turn proportional to the shear stress, a parameter expressing the pressure exerted by the blood on the vessel walls. The knowledge of this value is important in medicine to establish the state of the vessels, directly related to vascular diseases. FRP provides a non-invasive measure of the blood velocity by exploiting the red corpuscles property of diffusing ultrasound waves: in practice blood velocity is determined by a cross-correlation technique, which analyses the time shift between correlated subsequent echo waves, instead of frequency shift characteristic of the Doppler technique. The acquired data are then processed on a personal computer by means of mathematical techniques based on the evaluation of the correlation function, giving a reconstructed velocity profile and showing a good adherence with experimental data, since the average error is nearly the 10%. The reconstructed profile is displayed to the operator, who can follow the vessel status in real time. A few comparisons between the reconstructed and the experimental profiles are also presented, together with a study on a small set of patients suffering from artery hypertension.

An accelerator for double precision floating point operations

Studying and understanding human brain is one of the main challenges of 21st century scientists.The Human Brain Project was conceived for addressing this challenge in an innovative way, enabling collaborations between 112 partners spread in 24 European countries.The project is funded by the European Commission and will last until 2023.This paper describes the ongoing activity at one of the Italian units focused on innovative brain simulation through high performance computing technologies. Simulations concern realistic models of neurons belonging to the cerebellar cortex. Due to the level of biological realism, the computational complexity of this model is high, requiring suitable technologies. In this work, simulations have been conducted on high-end Graphical Processing Units (GPUs) and Field Programmable Gate Arrays (FPGAs). The first technology is used during model tuning and validation phases, while the latter allows to achieve real time elaboration, aiming at a possible development of embedded implantable systems. Simulations performance evaluations are discussed in the result section.