Paper Title, Authors, Abstract (Issue 1, Volume 4, 2010)

Pages

The Sky-Scanner System for Air Traffic Management: A Simulation Software
M. Salerno, G. Costantini, M. Carota, D. Casali

Abstract: Laser detection and tracking of aircrafts based systems (LIDARs, LIgth Detection And Ranging systems) are emerging as a critical design trend in development of new generation ATM (Air Traffic Management) paradigms, of which they are the main innovations. A novel laser tracking technology (SKY-Scanner System) capable to detect and track of aircrafts up to at least 6 nautical miles from the Aerodrome Traffic Zone (ATZ) has been proposed. The proposed methodology is considered at the frontier of technological research but it represents the only realistic way to put solid basis for the fabrication of effective radar and lidar integrated systems for incorporation in new generation ATM paradigms. The present paper is mainly focused on the simulation software of the above mentioned system. The simulation software is necessary in order to predict the behavior that the system, which is currently under development, will have. The software consists of two modules: Sim-module and Scen-module, both integrated in a single software package. The Sim-module simulates the mechanical system, the interaction of the laser with atmosphere, and the reflection on the surface of the airplane. The Monte-Carlo method will be used in order to take into account of random variables involved in the system, for example noise, turbulence, and error in mechanical positioning system of the lidar. The Scen-module simulates a scenery in which one or more aircraft will move along trajectories the user will specify.
 

Fault Tolerance of Artificial Neural Networks: An Open Discussion for a Global Model
Fernando Morgado Dias, Ana Antunes

Abstract: It is commonly assumed that neural networks have a built in fault tolerance property mainly due to their parallel structures. The international community of Neural Networks discussed these properties only until 1994 and afterwards the subject has been mostly ignored. Recently the subject was again brought to discussion due to the possibility of using neural networks in nano-electronic systems where fault tolerance and graceful degradation properties would be very important. In spite of these two periods of work there is still need for a large discussion around the fault model for artificial neural networks that should be used. One of the most used models is based on the stuck at model but applied to the weights. This model does not cover all possible faults and a more general model should be found. The present paper proposes a model for the faults in hardware implementations of feedforward neural networks that is independent of the implementation chosen and covers more faults than all the models proposed before in the literature.
 

Shape Classification Using Contour Simplification and Tangent Function
Chi-Man Pun, Cong Lin

Abstract: In this paper we propose a new approach for image classification by simplifying contour of shape and making use of the tangent function as image feature. We firstly extract shapes from a sample image and connecting pixels of its contour. The extracted contour is simplified by our algorithm and converted into tangent function which is regarded as a feature. The tangent function represented a shape is input into classified system and compared with tangent function from existed classes by computing their distance. The input sample image will finally be classified into a class that has minimum distance with it. The experimental results show the proposed method can achieve high accuracy.
 

    Paper Title, Authors, Abstract (Issue 2, Volume 4, 2010)

Pages

Image Edge Detection Using Ant Colony Optimization
Anna Veronica Baterina, Carlos Oppus

Abstract: Ant colony optimization (ACO) is a population-based metaheuristic that mimics the foraging behavior of ants to find approximate solutions to difficult optimization problems. It can be used to find good solutions to combinatorial optimization problems that can be transformed into the problem of finding good paths through a weighted construction graph. In this paper, an edge detection technique that is based on ACO is presented. The proposed method establishes a pheromone matrix that represents the edge information at each pixel based on the routes formed by the ants dispatched on the image. The movement of the ants is guided by the local variation in the image’s intensity values. The proposed ACO-based edge detection method takes advantage of the improvements introduced in ant colony system, one of the main extensions to the original ant system. Experimental results show the success of the technique in extracting edges from a digital image.
 

Frontier-based Exploration with Two Cooperating Mobile Robots
M. Al-Khawaldah, S. Livatino, D. Lee

Abstract: In this paper, a frontier-based algorithm is used with two cooperating mobile robots to explore unknown environment. The aim is to decrease the exploration time. The proposed frontier-based exploration is based on a new bidding function in which we introduced a special parameter to decrease the overlap between the robots in addition to the utility and cost parameters. Tens of thousands experiments have been conducted, each experiment is executed with different weight values set, to see the relative importance of the weight parameters used in this technique. As a result of these experiments, the weight values can be chosen, according to the environment characteristics, to guarantee short exploration time. The proposed algorithm has been tested with a set of environments with different shapes and different numbers of obstacles. Finally, the results of our algorithm were compared with the results of one of known exploration algorithms available in the literature. The new technique led to promising results.
 

Low Power Semiconductor Devices at 65nm Technology Node
Kiran Bailey, K. S. Gurumurthy

Abstract: This paper attempts to analyze the performance of 65 nm CMOS device structures for low power applications. It indicates that the historical trend of scaling of MOS devices can be sustained by innovative CMOS Structures such as Ultra-thin body SOI devices and multiple gate MOSFETS (such as FinFETS), that can withstand the adverse effects of Scaling. A particular issue of great concern in logic design is the power dissipation. For high- performance logic with increased leakage currents, chip static power dissipation is expected to become a bottleneck to meet aggressive targets for performance scaling. Innovations in circuit design and architecture for performance management as well as utilization of multiple transistors on chip are required for chip design. Multiple transistors having different threshold voltages (Vt) are used selectively with the low Vt, high leakage devices being used mainly in the critical paths and higher Vt, lower leakage devices being used in the rest of the chip area to control static power dissipation. This paper presents the low static power dissipation CMOS devices at 65 nm technology node and compares the performance of SOI CMOS with the conventional planar Bulk CMOS and establishes that SOI CMOS is better suited for low stand-by power applications. A low leakage current of 0.2 pA/μm for NMOS and 0.1 pA/μm for PMOS was observed for SOI devices at a supply voltage of 1.5V as compared to 10nA/μm for bulk CMOS devices at a supply voltage of 1.2V.
 

Modern Algorithms by Filtration and Automated Classification of Products
Marius-Constantin Popescu, Marius Buzera, Nikos E. Mastorakis, Jean-Octavian Popescu

Abstract: Researches concerning classification automatic installations, on the basis of colour, shape, size of both industrial and vegetal products, have lately increased in number. Almost all studies hint to replace the old mechanic installations with newer modern machine vision – based methods. This technique has the advantage of ensuring the evaluation of some features such as colour, faults detection, impossible to be assessed through any mechanic procedures, still, both in the case of mechanic products and especially regarding vegetal ones, the variation fields of colour and shape may vary at a very large scale. That is why, before starting to classify a new set of industrial products, or a new variety of vegetal products an establishing stage of the variation limits of colour, and features of the shape, is going to take place. Techniques, of image preprocessing focusing on the linear ones are going to be tested throughout the article, so as the best techniques of improving the images acquired and the fastest ones, to be chosen. In this aim we designed an application named Filters Quality Analyzer.
 

An Enhanced Rate Control Based on Mode Decision and Early Motion Estimation for H.264/AVC
Siavash Es'Haghi, Hassan Farsi

Abstract: The H.264/AVC video coding standard delivers a significantly better performance compared to previous standards, supporting higher quality video over lower bit rate channels. Rate control plays an important role in real-time video communication applications using H.264/AVC. An important step in many existing rate control algorithms is to determine the target bits for each P frame. This paper aims in improving video distortion by allocating more bits to frames with higher complexity and fewer bits to low complexity frames. In this work, the distribution of Macro Block (MB) modes in a frame is considered as a measure of its complexity. Also, an early motion estimation approach is introduced and used for complexity estimation. The bit budget is then allocated to frames according to their complexity and buffer status. Simulation results show that the proposed method effectively improves the PSNR average and meets the target bit rate more closely. In addition the proposed technique is less complex than other existing frame layer bit allocation schemes that are based on frame complexity.
 

Watermarking Technique Based on DWT Associated with Embedding Rule
Jih Pin Yeh, Che-Wei Lu, Hwei-Jen Lin, Hung-Hsuan Wu

Abstract: Information hiding has been an important research topic for the past several years. Techniques to solve the problem of unauthorized copying, tampering, and multimedia data delivery through the internet are urgently needed. Today’s information hiding techniques consist mainly of steganography and digital watermarking. In this paper, we shall focus on the digital watermarking and propose an improved version of the integer discrete wavelet transform (integer-DWT)-based watermarking technique proposed by Chang et al. [17]. Our method is able to achieve ownership protection. First, the original image is performed with the Discrete Wavelet Transformation (DWT) and embedded with the watermark in the HL and LH blocks associated with an embedding rule. The experimental results show that the proposed approach indeed produces better results than the compared method in terms of the quality of the stego image, the extracted watermark with or without attack, and time efficiency.
 

    Paper Title, Authors, Abstract (Issue 3, Volume 4, 2010)

Pages

Studies on Implementation of Harr and daubechies Wavelet for Denoising of Speech Signal
Mahesh S. Chavan, Nikos Mastorakis

Abstract: In hands free speech communication environments situation occurs that speech is superposed by background noise. Over the past few decades there is tremendous increase in the level of ambient environmental noise. This has been due to growth of technology. Noise is added by various factors like noisy engines, heavy machines, pumps, vehicles, over noisy telephone channel or using radio communication device in an aircraft cockpit. As speech is transmitted and received using various media it introduces distortions and have bandwidth constraints. These degradations lower intelligibility of speech message causing severe problems in downstream processing and user perception of speech signal. There has been a lot of research in speech denoising so far but there always remains room for improvement. The motivation to use wavelet as a possible alternative is to explore new ways to reduce computational complexity and to achieve better noise reduction performance. The wavelet denoising technique is called thresholding. It is divided in three steps. The first one consists in computing the coefficients of the wavelet transform (WT) which is a linear operation. The second one consists in thresholding these coefficients. The last step is the inversion of the thresholded coefficients by applying the inverse wavelet transform, which leads to the denoised signal. This technique is simple and efficient. In this paper wavlet is used as denoising algorithm. Performance of the Haar and Daubechies wavelets are experimentally evaluated.
 

FPGA Prototype of Robust Watermarking JPEG2000 Encoder
Pankaj U. Lande, Sanjay N. Talbar, G. N. Shinde

Abstract: In this paper we have presented a novel hardware for watermarking which can be used with the loss less JPEG2000 compression standard. The aim of hardware assisted watermarking is to achieve low power usage, real-time performance, reliability, and ease of integration with existing consumer electronic devices. We have implemented CDF5/3 wavelet filters with lifting scheme which requires less hardware and they are also the basis of lossless JPEG2000. The experimental result shows that the proposed scheme of watermarking is robust against most of the geometric attacks scaling and rotation.
 

Transport Supervision of Perishable Goods by Embedded Context Aware Objects
A. Wessels, R. Jedermann, W. Lang

Abstract: Intelligent freight objects can be introduced in order to simplify the solution of complex logistical planning tasks. This enables splitting logistical problems to local executable subtasks. The concept of Ubiquitous Computing (UbiComp) presents a model to realize such intelligent objects. A local task can be handled autonomously by a group of objects. The realization of UbiComp needs a platform that first provides the computational resources for the implementation of decision algorithms and then, secondly, the position information in order to enable context-aware features. This article presents a concept for the autonomous supervision of perishable goods and introduces the required soft- and hardware. Freight objects are represented by individual software components, which are realized with the JAVA framework OSGi. The whole software is designed to run on wireless sensor nodes to create embedded objects. Because temperature values can differ inside a reefer container, the signal strength of an RFID reader is used to provide position information by a cell based localization. A new approach is presented, where four RFID antennas are used to locate goods inside a container.
 

A Low-Power Synthesis of Submicron Interconnects with Time and Area Constraints
A. Mahdoum, R. Benmadache, A. Chenouf, M. L. Berrandjia

Abstract: Technology scaling has resulted in interconnect delay increasing significantly. Buffer-insertion is a well-known technique to reduce wire delays of critical signal nets in a circuit. However, the power consumption of buffers has become a critical concern with the increase of the number of buffers. In this paper, it is shown that this problem is not polynomial in time. Thus, we developed a geneticbased algorithm that provides optimal or near optimal solutions for reducing the power dissipation while meeting the time and area constraints.
 

Segmentation of Compound Signals Using Higher-Order Activity Indexes
Damjan Zazula, Rok Istenic

Abstract: The so called activity index is defined on the secondorder statistics. In multiple-input multiple-output (MIMO) models, it estimates the level of activity of individual input sources. In this paper, we study an extended definition of activity index, which deploys higher-order statistics instead of the second-order. Experiments with synthetic models show that noise resistance at higher even orders increases. In spite of the fact that the level of superimpositions of source activity also increases, which is disturbing, a difference of the 4th- and 2nd-order activity indexes proves to be a reliable relative measure of the number of simultaneously active input sources. The measure is not influenced either by the properties of sources, or by the level of additive random Gaussian noise, or the over- or underdeterminancy of the model output observations. This can be of considerable help when analysing and, in particular, decomposing compound MIMO output signals.
 

Cooperative Carrying Task Control based on Receding Horizon Control for Mobile Robots
Kou Nakamura, Tohru Kawabe

Abstract: A new control method for a cooperative carrying task control problem by two mobile robots is proposed. In the problem, a leading robot is assumed to be controlled by a human directory or remotely. On the other hand, a following robot must be run autonomously anytime without dropping a carrying object. The proposed method based on the receding horizon control (RHC) generates optimal left and right wheel motor torques of the following robot at each sampling time to hold constraint condition of relative position. Numerical examples are shown to demonstrate the effectiveness of the method.
 

    Paper Title, Authors, Abstract (Issue 4, Volume 4, 2010)

Pages

Control of Discrete-Time Systems with State Equality Constraints
A. Filasova, D. Krokavec

Abstract: Design conditions for existence of memory-less feedback control for stabilization of discrete-time systems with equality constraints given on the state variables are presented in the paper. The design problem is addressed for linear discretetime systems with equality constraints tying together some state variables. Using the classical memory-less feedback control principle LMI-based procedures are provided for computation of the gain matrix of state control laws, and influence of equality constraints is explained if a tracking problem be considered. The approach is successfully illustrated on simulation examples, where the validity of the proposed method is demonstrated with an equality constraint tying together all state variables of the system.
 

A Real Time 3D Foot Pose Estimation using Silhouette based 2.5D Database
Ho-Geun Song, Ha-Sung Koo

Abstract: This paper proposes a real time 3D foot pose estimation method using silhouette based 2.5D database. Shape descriptor for 3D pose estimation should be chosen to hold robustness for geometrical transformations, and to require short processing time. To reduce processing time, total 13,500 silhouette-based foot image database is built and meta information which involves 3D pose and feature vectors of the foot image is appended to the database. And we proposed a modified Centroid Contour Distance whose size of the feature space is small and performance of pose estimation is better than the others. In order to analyze performance of the proposed descriptor, we evaluate time and spatial complexity with retrieval accuracy, and then compare with the previous methods. According to the result of our experiment, proposed descriptor has only 63 feature values. It is about 20% of the average spatial complexity. Furthermore, our method takes about 0.00296 seconds to extract feature values. It is about 86% of the average time complexity. The results show that the proposed descriptor is more effective than the previous methods on feature extraction time. Finally, if we estimate retrieval accuracy as a reciprocal of the average error, then the proposed method improves about 36% of the average retrieval accuracy in comparison with the other methods. The experimental results show that the proposed descriptor is more effective than the previous methods on feature extraction time and pose estimation accuracy.
 

Applying Multiple KD-Trees in High Dimensional Nearest Neighbor Searching
Shwu-Huey Yen, Chao-Yu Shih, Tai-Kuang Li, Hsiao-Wei Chang

Abstract: Feature matching plays a key role in many image processing applications. To be robust and distinctive, feature vectors usually have high dimensions such as in SIFT (Scale Invariant Feature Transform) with dimension 64 or 128. Thus, accurately finding the nearest neighbor of a high-dimension query feature point in the target image becomes essential. The kd- tree is commonly adopted in organizing and indexing high dimensional data. However, in searching nearest neighbor, it needs many backtrackings and tends to make errors when dimension gets higher. In this paper, we propose a multiple kd-trees method to efficiently locate the nearest neighbor for high dimensional feature points. By constructing multiple kd-trees, the nearest neighbor is searched through different hyper-planes and this effectively compensates the deficiency of conventional kd-tree. Comparing to the well known algorithm of best bin first on kd-tree, the experiments showed that our method improves the precision of the nearest neighbor searching problem. When the dimension of data is 64 or 128 (on 2000 simulated data), the average improvement on precision can reach 28% (compared under the same dimension) and 53% (compared under the same number of backtrackings). Finally, we revise the stop criterion in backtracking. According to the preliminary experiments, this revision improves the precision of the proposed method in the searching result.
 

A Single-Scaled Hybrid Filtering Method for IPTV Program Recommendation
Kyusik Park, Jongmoo Choi, Donghee Lee

Abstract: In this paper, a single-scaled hybrid filtering algorithm is proposed to recommend user preferred IPTV-VOD program. A proposed system is implemented with hybrid filtering method that can cooperatively complements the shortcomings of the content-based filtering and collaborative filtering. For a user program preference, a single-scaled measure is designed so that the recommendation performance between content-based filtering and collaborative filtering is easily compared and reflected to final hybrid filtering procedure. In order to provide a high quality of program recommendation, we use not only the user watching history, but also the user program preference and mid-subgenre program preference which are updated weekly as a user preference profile. System performance is evaluated with modified IPTV data from real 24-weeks cable TV watching data provided by Nilson Research Corp. in Korea and it shows quite comparative quality of recommendation.
 

Facial Feature Extraction for Face Modeling Program
Ha-Sung Koo, Ho-Guen Song

Abstract: In this paper suggests to define 20 facial features being used in 3D face modeling program and the method of extraction of 20 facial features from 2D image. The experimental image is to be restricted to input one person's image, background of the image can be taken in anywhere, the lighting condition is not uniform, and there is no racial restriction. The suggested method is to seek facial candidate region by Harr Classifier and to decide eye candidate region and extract eye features by dilate operation then decide lip candidate region using the features. The relative color difference of a* in the L*a*b* color space was used to extract lip feature and to seek nose candidate region and detected 20 features from 2D image by analyzing end of nose.