ABSTRACT. Complexity of both, applications and processors, is permanently increasing. While, designers must face with a lot of challenges (reduced time to market, performance and energy efficiency, among others), many times with a weak knowledge of the target technologies they intend to use. In this context, dataflow based methodologies and tools, like the Open RVC CAL Compiler Infrastructure (Orcc), have revealed powerful in supporting platform independent design. Within the Orcc design flow, applications are described using the RVC CAL language. A set of Orcc backends can generate different source codes (e.g., C, C++, VHDL, Java) that must be finally compiled for a target platform with a native compiler. This approach has several advantages, but the quality of the final product is limited by the quality of the currently implemented backends. In this work, an HEVC decoder implementation for DSP technology has been developed by using a combination of RVC-CAL descriptions and native C-code. The results of this approach are compared with those obtained by using both, a pure RVC CAL description and a pure native C code implementation.

ABSTRACT. Video Quality Assessment needs to correspond to human perception. Pixel-based metrics (PSNR or MSE) fail in many circumstances for not taking into account the spatio-temporal property of human’s visual perception. In this paper we propose a new pixel-weighted method to improve video quality metrics for artifacts evaluation. The method applies a psychovisual model based on motion, level of detail, pixel location and the appearance of human faces, which approximate the quality to the human eye’s response. Subjective tests were developed to adjust the psychovisual model for demonstrating the noticeable improvement of an algorithm when weighting the pixels according to the factors analyzed instead of treating them equally. The analysis developed demonstrates the necessity of models adapted to the specific visualization of contents and the model presents an advance in quality to be applied over sequences when a determined artifact is analyzed.

ABSTRACT. It is highly desirable to known in advance the transaction of data in the design of any electronic embedded system. It is of especial interest for data-intensive applications, such as complex video system, when the options available in the video decoder continuously change and/or the features of the input video sequences are different. This paper exposes the development of a profiling tool intended to assist the designer on the decision making based on the transaction of data during the decoding of video sequences using the H.264 Scalable Video Coding (SVC) standard. The tool incorporates some Python scripts that allow the designer to analyze the data transaction in the decoder, with several automating profiling utilities for the management of bitstreams and decoding tasks. Using the profiling scripts, the results show that the data transaction load changes based on the intrinsic characteristics of the video sequence and on the scalable options selected. Due to the huge number of functions that form part of the SVC video decoder, a set of modules that better describe the internal structure of the decoder has been defined. The assignation of functions to modules is open to the designer and can be changed at any time to accommodate changes in the system. This environment allows the system designer to make better decisions about the load distribution of data transactions, based on the modules defined inside the tool to create the profile.

ABSTRACT. The use of X-ray inspection for analyzing faults in printed circuit board components is becoming more important with the ongoing development of 3D integrated circuits. Achieving optimal X-ray inspection for a slab-like flat object depends on selecting an appropriate computed tomography (CT) geometry, such as an off-centered cone-beam geometry. However, because the use of flat components in tomographic reconstruction restricts the range of viewing angles, CT leads to the missing wedge problem, which causes missing angle artifacts in reconstructed 3D data. In this paper, we propose the use of post-processing to reduce the missing wedge effect via total variation minimization, when sinogram images have equal quality at all viewing angles and the accessible tilt range is restricted only by the physical limits of the oblique-view CT system. We demonstrate that axial slice reconstruction using the proposed processing method substantially reduces the missing wedge effect and that the proposed method provides overall image quality superior to that achieved with existing denoising methods.

ABSTRACT. This paper presents an interleaved sine wave generator using an efficient memory access technique for large touch screen controllers. It concurrently applies sine waves of different frequencies. The proposed circuit improves the speed of touch detection and reduces the chip area compared with conventional analog oscillators. The range of sine wave frequencies and the gap between each of frequencies can be configured by adjusting the address calculation algorithm to fetch samples values stored in memory. Therefore it provides higher flexibility with configurable sine wave frequencies than conventional schemes. It also minimizes the memory size required to regenerate all the sine waves needed. The proposed architecture has been implemented the proposed sine wave generator in a touch screen controller with FPGA and analog front end board.

ABSTRACT. In this work, fabricated 1550-nm short-wavelength infrared (SWIR) Ge-on-Si photodetectors coupled with a Si waveguide on silicon-on-insulator (SOI) for on-chip optical interconnect as an energy-efficient green technology for next-generation very-large-scale integration (VLSI) systems are characterized. Here, a particular emphasis is put on the back-end-of-the-line (BEOL) technology in device design. Comparison study on the effects of interconnection geometry on the electrical and optical DC characteristics of the device is made. Compared with a reference device with bulk contacts, device with holey contacts demonstrated an increased optical responsivity. Further, device with holey contacts on two metal layers showed the highest photocurrent. Also, dependence of forward and reverse currents in the heterojunction pn diode on effective contact area is empirically studied. As a result, it is found that external quantum efficiency (EQE) can be significantly improved by engineering the geometry of metal interconnect in the Ge-on-Si photodetector without being affected by reduction in the effective contact area.

ABSTRACT. In this paper we introduce a novel approach for multi room comfort control based on humidex index. The aim of the system is to reach and maintain comfort conditions through a fuzzy logic controller. The control system has been experimentally tested in an apartment in the central east coast of Italy. Temperature regulation performances have been compared with those of a classical PID based thermostat.

ABSTRACT. In this paper, a new approach capable of detecting water, ice and snow on the road surface is shown, focusing on the hardware needed for the practical implementation. Proposed platform uses a simple microphone and a general purpose microcontroller. The system is tightly integrated with vehicle ECUs (Electronic Control Units). A CAN bus allows it to read some of the variable dynamics provided by the sensors already installed in virtually every motor vehicle. Detector results are also published through the CAN bus to be displayed on the vehicle console and to be used by other ECUs. Proposed system is simple, cheap and easy to install. Initial tests for the wet/dry detector classes have shown a very high success rate.

ABSTRACT. In this paper the authors present a new middleware for mapping gestures, identified by the Kinect device, to key events which are consumed by a standard off-the-shelf application. The aim is twofold: accessibility for users which are not able to use the keyboard because of physical impairments and the use of standard games for doing physical exercises. Hereby, special attention is laid on the adaptability to user requirements and easiness of configuration for the user himself and non-expert assistants. The actual state of our system is compared to similar proposals in terms of usability and performance, finally future working directions are outlined.

ABSTRACT. In this paper, we develop a platform of medical instrumentation for real-time electroencephalogram(EEG) measurement and analysis system. This platform obtains EEG signal from electrode sensors, analyzes the signal and transmits the result to mobile device via Wi-Fi. Because the changes of EEG signal on the scalp received by electrode are in very small range, the signal is amplified by designed circuits. Low pass filter is applied in order to eliminate the component of breathing and high-pass filter is used to remove power supply noise. The digital logic in FPGA transforms time domain signal into frequency domain signal using 256 point radix-2 decimation-in-frequency as an FFT algorithm. The platform has developed on the FPGA development board.

ABSTRACT. This paper describes the development of a software complement for Blender (a freely available animation software) which allows to insert motion data obtained from the Kinect camera. The main target is to provide Blender with a tool to develop exergames, i.e. serious games for performing physical exercises, which are fully adaptive to the user’s needs and capabilities, especially addressing chronical patients or handicapped. The plugin incorporates both: motion control and recording, allowing to store user movements. Transmission is realized via the OSC (Open Sound Control) protocol. The tool is still under development, here a first version is presented together with a demo-game. In the future, sensor data from other devices is planned to be integrated.

ABSTRACT. Choice of an electroencephalogram (EEG) reference is a critical issue during measurement of brain activity. An appropriate reference may improve efficiency during diagnosis of psychiatric conditions, e.g., major depressive disorder (MDD). In literature, various EEG references have been proposed, however, none of them is considered as gold-standard [1]. Therefore, this study aims to evaluate 3 EEG references including infinity reference (IR), average reference (AR) and link-ear (LE) reference based on EEG data acquired from 2 groups: the MDD patients and healthy subjects as controls. The experimental EEG data acquisition involved 2 physiological conditions: eyes closed (EC) and eyes open (EO). Originally, the data were recorded with LE reference and re-referenced to AR and IR. EEG features such as the inter-hemispheric coherences, inter-hemispheric asymmetries, and power in different frequency bands were computed. These EEG features were used as input data to train and test the logistic regression (LR) classifier. Finally, the results were presented as classification accuracies, sensitivities, and specificities while discriminating the MDD patients from a potential population of healthy controls. According to the results, AR has provided the maximum classification efficiencies for coherence and power based features. The case of asymmetry, IR and LE performed better than AR. The study concluded that the reference selection should consider factors such as underlying EEG data, computed features and type of assessment performed.

ABSTRACT. To overcome the high cost and limited availability of indium tin oxide (ITO) used in manufacturing touch panels, in this work a novel polymer-based touch panel is proposed. The proposed touch panel interface is mainly made from a composition of acrylic resin and Polyethylene glycol which are abundant and low-cost materials. The touch-sensing technology adopted for this touch panel is based on detecting the induced voltage on the human body resulting from the surrounding 50/60Hz electromagnetic waves of power lines. In addition to its low cost and good performance, the implemented touch panel is flexible, transparent and above all, its interface is made from high-resistivity materials offering promising alternatives to ITO in touch panels manufacturing.

ABSTRACT. A novel and robust vision-based human-machine interface system to naturally interact with computers/smart devices is proposed. The key contribution is the introduction of a Compressive Sensing technique to largely reduce the dimensionality of highly discriminative feature descriptors (computed from depth imagery), which originally have an excessive and inoperative high dimension to be applied to a Support Vector Machine based classifier. The experimental results prove the appropriateness of this approach for the proposed system.

ABSTRACT. In this study, we carried out experiments to investigate the influence of display resolution on psychophysiological state while viewing various types of video content at 4K and 2K on a 65-inch 4K TV as a comparison between age groups. The participants viewed four kinds of TV video content (two types of scenic material and two kinds of material with movement and action). The results showed the psychological evaluation scores for participants in their 20s to be higher than those in their 50s when viewing 4K scenic content. However in both age groups, NIRS, an index of nervous system activity, during viewing tests of 4K scenic content, was significantly higher for 4K scenic content viewing than for 2K scenic content. On the other hand, there was a significant difference observed between NIRS for 4K and for 2K in the 50s participants, although there was no significant difference observed for NIRS between resolutions in the 20s participants. Our results suggest that content viewing at 4K can cause psychological elation and a surge in brain activity, although the effects varied somewhat according to content.

ABSTRACT. The current technology trend in automotive navigation systems is “Display Audio” devices. These are display devices that connect directly to and are controlled by the user’s smartphone. However, such devices suffer problems of driver distractibility as well as problems related to differences in the life cycle of the Display Audio device and the smartphone. In this study, we discuss the state of current on-vehicle information devices, and propose a new architecture for an on-vehicle information system using a smartphone and an accelerometer. Finally, we show positive results based on a usability evaluation.

ABSTRACT. In this paper, we propose a real-sense vestibular rehabilitation systems using sensory experience based on multiple sensory media. This system provides realistic vestibular rehabilitation exercises through IPTV channel without visiting a rehabilitation center. It provides sensory effects like wind, light, scent, vibration and so on in order to provide pleasure and durability.

ABSTRACT. This paper presents an approach to incorporate strong security in deploying Internet of Things (IoT) for smart home system, with due consideration given to user convenience in operating the system. The IoT smart home system runs on conventional wifi network implemented based on the AllJoyn framework, using an asymmetric Elliptic Curve Cryptography to perform the authentications during system operation. A wifi gateway is used as the center node of the system to perform the system initial configuration. It is then responsible for authenticating the communication between the IoT devices as well as providing a mean for the user to access and control the system through an Android based mobile device running appropriate application program.

ABSTRACT. In this paper, we investigate the symbol timing recovery when FTN signaling scheme is applied to DVB-S2 standard. In general, the conventional DVB-S2 receiver has not been taken into account for ISI problem except the impairment by satellite transponder. However, it’s inevitable to modify it for FTN signaling because ISI issue is more severe, significantly. We propose the robust timing recovery as for FTN signaling in DVB-S2 and to analyze the effect through performance evaluation.

ABSTRACT. We present an exhibition information model (EIM) by analyzing quantitative data for curators to configure the exhibition space. The proposed method consists of four steps: i) metadata classification for EIM, ii) probabilistic interest prediction model (IPM), iii) mutual interest-based time prediction, and iv) spatial analysis for path prediction. The proposed method can be simulated in advance using the information of visitors and characteristics of the exhibition. The proposed simulation method can be applied to analyze the visitor’s behavior by detecting and tracking moving objects.

ABSTRACT. QR decomposition is a key operation in many current communication systems. This paper shows how to reduce the area of a fixed-point QR decomposition implementation based on Givens rotations by using a new number representation system. This new representation allows performing round-to-nearest at the same cost of truncation. Consequently, the rounding errors of the results are halved, which allows it to reduce the word-length by one bit. This reduction positively impacts on the area, delay and power consumption of the design.

ABSTRACT. A large number of studies on small flying robots have been carried out. Flying robots need to land on specific heliports, thus necessitating the search for specific heliports. This paper reports a method for searching a heliport, in which data from an onboard camera and an ultrasonic sensor are processed using an onboard Raspberry Pi unit.

ABSTRACT. In this paper, we present a novel method for building detection using Lidar point clouds. The proposed method uses SFTA features to identify buildings contours. The evaluations carried out on the ISPRS LIDAR reference data set, show high detection performances (89 percent).

ABSTRACT. In this paper, adaptive PU mode estimation algorithm is proposed to achieve a fast High Efficiency Video Coding (HEVC) encoding. According to HEVC standard, a picture is composed of a number of Coding Tree Units (CTU's) These CTUs are further divided in quadtree structure to form Coding Units (CUs), and CUs have one or more Prediction Units (PUs). The PU mode decision processing to find a best PU mode has the most computational complexity. To reduce the computational complexity, the proposed algorithm can skip checking of both symmetric and asymmetric PU partition sizes using best PU information (i.e., partition size and mode) of temporally co-located CUs. Experimental results show that the proposed algorithm using HM 16.2 achieves an average 30% reduction in encoding time for random access configuration, with a Y BD-rate loss of 0.58%.

ABSTRACT. This paper presents a Ring Projection Transforms (RPT) for recognizing an object in automation system. Based on the Coarse-fine strategy and Compute Unified Device Architecture (CUDA), we developed a recognition model on GPGPU (General-Purpose computing on Graphics Processing Units) scheme. As experimental validation indicates, our scheme achieved a better performance in terms of processing time compared with the existing method.

ABSTRACT. Abstract—This paper presents an object recognition algorithm for machine vision in the factory robot automation application. Especially when the object is rotational states, the automated machine needs the exact angle and recognition algorithm of high reliability with fast processing speed. Therefore, in this paper, we propose a method focusing on the robust-rotation recognition algorithm by using a CCD Centroid Contour Distance) method. Furthermore, the proposed algorithm targets reducing the processing speed by using the CUDA (Compute Unified Device Architecture) for real-time processing.

ABSTRACT. “Selfie” capturing is one of the most popular functions used by smart phone users. Those selfies are usually uploaded and exchanged on social networking services such as Facebook, Instagram, or Twitter. We introduce an automatic cubic facial make-up method by using simple image processing techniques including resizing, Gaussian smoothing and face detection. The proposed algorithm is designed to run in a real-time application at Android smart phone and no human interactions are required. The experimental results outperform conventional image editing tools and seamless object boundaries are observed.

ABSTRACT. In this paper, we proposed real-time FPGA-based rectification algorithm implementation combined with stereo camera. Rectification is a necessary pre-required step for stereo matching to align left and right images, including correction of radial distortion. In our experiment, we implemented rectification module in the FPGA combined with stereo camera to verify in video circumstance. As a result, rectification for HD image is processed at 45 fps, with Zynq7020 FPGA.

ABSTRACT. In this paper, a limited-view computer tomography(CT) image reconstruction method was proposed to reduce the scan times and X-ray dose in the semiconductor industry. To reduce streak artifacts, which are caused by an insufficient number of views, we studied comparative the sinogram interpolation method based on the average pixel value. The change patterns were tracked by iteratively averaging the values. A Shepp-Logan phantom was used to evaluate the effectiveness of the comparative interpolation method. The results showed that the streak artifacts were reduced in the adapted interpolation method. In addition, we compared our method with the other interpolation method by using the Peak-Signal-to-Noise Ratio(PSNR) and performance time.

ABSTRACT. Face detection and recognition are key components in multiple camera-based devices and applications. Smart glasses are a type of optical head mounted displays that integrate first-person cameras and hands free displays with immediate access to processing power able to analyze first person images in real time with hands free operation. In this context, we have constructed an application prototype that detects and recognizes faces in real-time, and runs independently on the device. We provide a description of the embedded implementation at a system-level where we highlight the application development challenges and trade-offs that need to be dealt with battery powered wearable devices. The implementation includes a parallel pipeline that reduces the latencies of the application.

ABSTRACT. In this paper, we propose a novel iris center (IC) localization method. Given an eye region image, we first find an initial iris center and perform iterative circle fitting until the center of the circle converges on the final IC. In each iteration step, a certain number of feature points are extracted from radial sectors defined around the IC, and then grouped into several clusters for circle fitting. Through the iteration, the circle center converges on the final IC. Experimental results on the public dataset GI4E show that the proposed method achieves the superior performance as compared with conventional ones.

ABSTRACT. This paper presents an accelerating face detection algorithm using Coarse Grained Reconfigurable Architectures (CGRA). Face detection algorithms usually use several stages of cascaded face detectors and require large amount of feature data, while general processors have small internal memory. Since the latency from external memory is much longer than internal memory, efficient use of memory is important to make face detection fast. In this paper, we do the first-stage of cascaded face detection process for every line using the feature data in the internal memory first, and then do next stages for the candidates who pass the first-stage. In addition, by efficient use of software pipelining and vectorization, face detection process can be accelerated. The proposed method was implemented into CGRA@400MHz, and can run at 15.1@800x600 frame per second.

ABSTRACT. Detection of moving objects remaining static is a fundamental step in many computer vision applications, since it allows to identify potentially dangerous situations (abandoned objects) and people temporally static. Here, we propose a strategy to efficiently detect such static moving objects, which is based on three nonparametric background models (long term, medium term and short term) to detect moving objects and a novel Finite State Machine to identify when a moving object becomes static.

ABSTRACT. In recent years, the necessity to protect multimedia content from illegal copying has been made more critical by the advent of digital technology. Singular Value Decomposition (SVD) received much attention from the watermarking community due to its appealing properties, such as the invariance of singular values (SVs) to common image processing operations and geometric transforms, like rotation, translation and scaling. However, most of watermarking schemes require a lot of prior information to detect the watermark and in addition they are not secure. In this paper we investigate the efficiency of optimal decoding schemes for the watermark inserted into the SVD domain of the host images using an additive spread spectrum (SS) embedding framework. In order to use the singular values (SVs) with the SS embedding we adopt several restrictions that ensure that the values of the SVs remain real, positive and sorted. For both the optimal maximum likelihood decoder and sub-optimal decoders we assume that the PDF of SVs can be modeled by the Weibull distribution. Furthermore, we investigate the error probability behavior, i.e., the probability of detection and the probability of false detection, for the applied optimal decoders. By taking into account the efficiency and the necessary auxiliary information for decoding the watermark, we discuss the suitable decoder for various operating situations. Experimental results are carried out to show the imperceptibility and decoding efficiency of the proposed scheme against various attack scenarios.

ABSTRACT. Signal Integrity is now an important part of the design cycles for consumer electronics devices. One of the major issues in signal integrity is passivity check and restoration of passive devices s-parameters. This paper will first discuss the background theory of passivity and introduce a new efficient technique of passivity restoration based on finding the local minima in the singular value vs. frequency stemming from the system matrix. The regions between pairs of adjacent local minima will be scaled down when the adjoining maximum is greater than 1, thus restoring passivity. The proposed new technique for passivity enforcement has the potential to restore the passivity in a real or simulated system represented by S-parameters without making significant changes in the system’s original S-parameters or the systems characteristics. This new algorithm is also successfully implemented on more than one example using frequency data from real systems

ABSTRACT. This paper presents a novel VLSI module that implements a lightweight symmetric authentication protocol based on Keyed-Hash Message Authentication Code (HMAC). The cryptographic key and the random numbers needed by the protocol are generated efficiently by a SRAM acting as a Physical Unclonable Function (PUF).

ABSTRACT. In this paper, we propose a near-optimal multiple-input multiple-output (MIMO) detection algorithm with low and fixed complexity, which is based on the fixed-complexity sphere decoder (FSD). By employing the lattice reduction (LR) to the FSD, the computational complexity is reduced considerably while maintaining the near-optimal bit-error-rate (BER) performance. Although the application of LR to the FSD is not straightforward, an efficient solution is proposed in this paper. The numerical results demonstrate that the computational complexity of the proposed algorithm is reduced by about 43.02% and 76.49% for 64-QAM and 256-QAM compared to the optimal FSD while the performance degradation is negligible.