A modification of the descriptor in a human detector using Histogram of Oriented Gradients (HOG) and Support Vector Machine (SVM) is presented. The proposed modification requires inserting the values of average cell brightness resulting in the increase of the descriptor length from 3780 to 3908 values, but it is easy to compute and instantly gives ≈ 25% improvement of the miss rate at 10‒4 False Positives Per Window (FPPW). The modification has been tested on two versions of HOG-based descriptors: the classic Dalal-Triggs and the modified one, where, instead of spatial Gaussian masks for blocks, an additional central cell has been used. The proposed modification is suitable for hardware implementations of HOG-based detectors, enabling an increase of the detection accuracy or resignation from the use of some hardware-unfriendly operations, such as a spatial Gaussian mask. The results of testing its influence on the brightness changes of test images are also presented. The descriptor may be used in sensor networks equipped with hardware acceleration of image processing to detect humans in the images.
The paper presents two algorithms as a solution to the problem of identifying fraud intentions of a customer. Their purpose is to generate variables that contribute to fraud models’ predictive power improvement. In this article, a novel approach to the feature engineering, based on anomaly detection, is presented. As the choice of statistical model used in the research improves predictive capabilities of a solution to some extent, most of the attention should be paid to the choice of proper predictors. The main finding of the research is that model enrichment with additional predictors leads to the further improvement of predictive power and better interpretability of anti-fraud model. The paper is a contribution to the fraud prediction problem but the method presented may generate variable input to every tool equipped with variableselection algorithm. The cost is the increased complexity of the models obtained. The approach is illustrated on a dataset from one of the European banks.
Stealth in military sonars applications may be ensured through the use of low power signals making them difficult to intercept by the enemy. In recent years, silent sonar design has been investigated by the Department of Marine Electronic Systems of the Gdansk University of Technology. This article provides an analysis of how an intercept sonar operated by the enemy can detect silent sonar signals. To that end a theoretical intercept sonar model was developed with formulas that can numerically determine the intercept ranges of silent sonar sounding signals. This was tested for a variety of applications and water salinities. Because they are also presented in charts, the results can be used to compare the intercept ranges of silent sonar and traditional pulse sonar.
The article presents the detection of gases using an infrared imaging Fourier-transform spectrometer (IFTS). The Telops company has developed the IFTS instrument HyperCam, which is offered as a short- or long-wave infrared device. The principle of HyperCam operation and methodology of gas detection has been shown in the paper, as well as theoretical evaluation of gas detection possibility. Calculations of the optical path between the IFTS device, cloud of gases and background have been also discussed. The variation of a signal reaching the IFTS caused by the presence of a gas has been calculated and compared with the reference signal obtained without the presence of a gas in IFTS's field of view. Verification of the theoretical result has been made by laboratory measurements. Some results of the detection of various types of gases has been also included in the paper.
Leak detection in transmission pipelines is important for safe operation of pipelines. The probability of leaks may be occurred at any time and location, therefore pipeline leak detection systems play a key role in minimization of the occurrence of leaks probability and their impacts. During the operation of the network there are various accidents or intentional actions that lead to leaks of gas pipelines. For each network failure, a quick reaction is needed before it causes more damage. Methods that are used to detect such network failures are three-staged-: early identification of leakage, an accurate indication of its location and determine the amount of lost fluid. Methods for leak detection can be divided into two main groups: external methods (hardware) and internal methods (software). External leak detection methods require additional, often expensive equipment mounted on the network, or use systems that could display only local damage on the pipeline. The alternative are the internal methods which use available network measurements and signalling gas leakage signal based on the mathematical models of the gas flow. In this paper, a new method of leak detection based on a mathematical model of gas flow in a transient state has been proposed.
Although the phenomenon of otoacoustic emission has been known for nearly 30 years, it has not been fully explained yet. One kind of otoacoustic emission is distortion product of the otoacoustic emission (DPOAE). New aspects of this phenomenon are constantly discovered and attempts are made to interpret correctly the obtained results. This paper discusses a new method of measuring DPOAE signals based on double phase-sensitive detection, which makes possible a real-time measurement of the DPOAE signal amplitude and phase. The method was applied for measurements of DPOAE signals in guinea pigs. Sample records are presented and the obtained results are discussed.
In this study, an artificial neural network application was performed to tell if 18 plates of the same material in different shapes and sizes were cracked or not. The cracks in the cracked plates were of different depth and sizes and were non-identical deformations. This ANN model was developed to detect whether the plates under test are cracked or not, when four plates have been selected randomly from among a total of 18 ones. The ANN model used in the study is a model uniquely tailored for this study, but it can be applied to all systems by changing the weight values and without changing the architecture of the model. The developed model was tested using experimental data conducted with 18 plates and the results obtained mainly correspond to this particular case. But the algorithm can be easily generalized for an arbitrary number of items.
The article reports three experiments conducted to determine whether musicians possess better ability of recognising the sources of natural sounds than non-musicians. The study was inspired by reports which indicate that musical training develops not only musical hearing, but also enhances various non-musical auditory capabilities. Recognition and detection thresholds were measured for recordings of environmental sounds presented in quiet (Experiment 1) and in the background of a noise masker (Experiment 2). The listener’s ability of sound source recognition was inferred from the recognition-detection threshold gap (RDTG) defined as the difference in signal level between the thresholds of sound recognition and sound detection. Contrary to what was expected from reports of enhanced auditory abilities of musicians, the RDTGs were not smaller for musicians than for non-musicians. In Experiment 3, detection thresholds were measured with an adaptive procedure comprising three interleaved stimulus tracks with different sounds. It was found that the threshold elevation caused by stimulus interleaving was similar for musicians and non-musicians. The lack of superiority of musicians over non-musicians in the auditory tasks explored in this study is explained in terms of a listening strategy known as casual listening mode, which is a basis for auditory orientation in the environment.
Tomato farms in Arusha, Morogoro, Dodoma, Iringa, Kilimanjaro and Coast regions of Tanzania were surveyed to assess the incidence of the yellow leaf curl disease, and to collect infected tomato leaf samples for sero-diagnosis. The triple antibody sandwich enzyme linked immunosorbent assay (TAS-ELISA) format was adopted for the detection of disease using commercial polyclonal antiserum and monoclonal antibodies SCRI 17, SCRI 20, SCRI 23 and SCRI 33. ELISA readings were rated on a scale of 0–4. The results of the tests indicated that all the Tomato yellow leaf curl virus (TY-LCV) isolates recorded high reaction values (4) with the polyclonal antibody. However, the Dodoma and Arusha isolates were rated highest in optical density (OD) reading with MAb SCRI 20 and 23. The remaining isolates produced lower OD values. All the isolates rated low (2) when tested with SCRI 33. The differences in reaction to the monoclonal antibodies of TYLCV indicated that variability exists between the coat protein epitopes of TYLCV and Tomato yellow leaf curl Tanzania virus (TYL-CTZV) on one hand, and among the TYLCTZV isolates on the other. Only the isolates from Arusha and Dodoma share a high sequence homology in coat protein with the European and related TYLCV isolates. Furthermore, the reaction with either SCRI 20 or SCRI 23 show that the isolates from Arusha and Dodoma share a high degree of homology, and could belong to one serotype. The other isolates from Morogoro, Coast and Kilimanjaro could form another serotype, while the isolate from Iringa is a different serotype. On the other hand, reaction with SCRI 17 groups the isolates in two serotypes, the Dodoma isolate alone, and another that groups the other five isolates together. It is recommended that other procedures such as DNA-DNA hybridization assays, polymerase chain reaction, restriction fragment length polymorphisms and sequencing can be combined with the use of monoclonal antisera for the detection and prediction or inference of Tomato yellow leaf curl disease (TYLCD) virus relationships at the quasi-species or strain levels in Tanzania.
Diagnostics of composite castings, due to their complex structure, requires that their characteristics are tested by an appropriate description
method. Any deviation from the specific characteristic will be regarded as a material defect. The detection of defects in composite castings
sometimes is not sufficient and the defects have to be identified. This study classifies defects found in the structures of saturated metallic
composite castings and indicates those stages of the process where such defects are likely to be formed. Not only does the author
determine the causes of structural defects, describe methods of their detection and identification, but also proposes a schematic procedure
to be followed during detection and identification of structural defects of castings made from saturated reinforcement metallic composites.
Alloys examination was conducted after technological process, while using destructive (macroscopic tests, light and scanning electron
microscopy) and non-destructive (ultrasonic and X-ray defectoscopy, tomography, gravimetric method) methods. Research presented in
this article are part of author’s work on castings quality.
The paper presents analyses of current research projects connected with explosive material sensors. Sensors are described assigned to X and γ radiation, optical radiation sensors, as well as detectors applied in gas chromatography, electrochemical and chemical sensors. Furthermore, neutron techniques and magnetic resonance devices were analyzed. Special attention was drawn to optoelectronic sensors of explosive devices.
The contribution presents a novel approach to the detection and tracking of lanes based on lidar data. Therefore, we use the distance and reflectivity data coming from a one-dimensional sensor. After having detected the lane through a temporal fusion algorithm, we register the lidar data in a world-fixed coordinate system. To this end, we also incorporate the data coming from an inertial measurement unit and a differential global positioning system. After that stage, an original image of the road can be inferred. Based on this data view, we are able to track the lane either with a Kalman filter or by using a polynomial approximation for the underlying lane model.
Biometric identification systems, i.e. the systems that are able to recognize humans by analyzing their physiological or behavioral characteristics, have gained a lot of interest in recent years. They can be used to raise the security level in certain institutions or can be treated as a convenient replacement for PINs and passwords for regular users. Automatic face recognition is one of the most popular biometric technologies, widely used even by many low-end consumer devices such as netbooks. However, even the most accurate face identification algorithm would be useless if it could be cheated by presenting a photograph of a person instead of the real face. Therefore, the proper liveness measurement is extremely important. In this paper we present a method that differentiates between video sequences showing real persons and their photographs. First we calculate the optical flow of the face region using the Farnebäck algorithm. Then we convert the motion information into images and perform the initial data selection. Finally, we apply the Support Vector Machine to distinguish between real faces and photographs. The experimental results confirm that the proposed approach could be successfully applied in practice.
In this paper we present the numerical simulation-based design of a new microfluidic device concept for electrophoretic mobility and (relative) concentration measurements of dilute mixtures. The device enables stationary focusing points for each species, where the locally applied pressure driven flow (PDF) counter balances the species’ electrokinetic velocity. The axial location of the focusing point, along with the PDF flowrate and applied electric field reveals the electrokinetic mobility of each species. Simultaneous measurement of the electroosmotic mobility of an electrically neutral specie can be utilized to calculate the electrophoretic mobility of charged species. The proposed device utilizes constant sample feeding, and results in time-steady measurements. Hence, the results are independent of the initial sample distribution and flow dynamics. In addition, the results are insensitive to the species diffusion for large Peclet number flows (Pe > 400), enabling relative concentration measurement of each specie in the dilute mixture.
This paper presents a methodology for contact detection between convex quadric surfaces using its implicit equations. With some small modifications in the equations, one can model superellipsoids, superhyperboloids of one or two sheets and supertoroids. This methodology is to be implemented on a multibody dynamics code, in order to simulate the interpenetration between mechanical systems, particularly, the simulation of collisions with motor vehicles and other road users, such as cars, motorcycles and pedestrians.
The contact detection of two bodies is formulated as a convex nonlinear constrained optimization problem that is solved using two methods, an Interior Point method (IP) and a Sequential Quadratic Programming method (SQP), coded in MATLAB and FORTRAN environment, respectively. The objective function to be minimized is the distance between both surfaces. The design constraints are the implicit superquadrics surfaces equations and operations between its normal vectors and the distance itself. The contact points or the points that minimize the distance between the surfaces are the design variables. Computational efficiency can be improved by using Bounding Volumes in contact detection pre-steps. First one approximate the geometry using spheres, and then Oriented Bounding Boxes (OBB).
Results show that the optimization technique suits for the accurate contact detection between objects modelled by implicit superquadric equations.
In this work the construction of experimental setup for MEMS/NEMS deflection measurements is presented. The system is based on intensity fibre optic detector for linear displacement sensing. Furthermore the electronic devices: current source for driving the light source and photodetector with wide-band preamplifier are presented.
The pathologists follow a systematic and partially manual process to obtain histological tissue sections from the biological tissue extracted from patients. This process is far from being perfect and can introduce some errors in the quality of the tissue sections (distortions, deformations, folds and tissue breaks). In this paper, we propose a deep learning (DL) method for the detection and segmentation of these damaged regions in whole slide images (WSIs). The proposed technique is based on convolutional neural networks (CNNs) and uses the U-net model to achieve the pixel-wise segmentation of these unwanted regions. The results obtained show that this technique yields satisfactory results and can be applied as a pre-processing step for automatic WSI analysis in order to prevent the use of the damaged areas in the evaluation processes.
An intelligent boundary switch is a three-phase outdoor power distribution device equipped with a controller. It is installed at the boundary point on the medium voltage overhead distribution lines. It can automatically remove the single-phase-to-ground fault and isolation phase-to-phase short-circuit fault. Firstly, the structure of an intelligent boundary switch is studied, and then the fault detection principle is also investigated. The single-phase-to-ground fault and phase-to-phase short-circuit fault are studied respectively. A method using overcurrent to judge the short-circuit fault is presented. The characteristics of the single-phase-to-ground fault on an ungrounded distribution system and compositional grounded distribution system are analyzed. Based on these characteristics, a method using zero sequence current to detect the single-phase-to-ground fault is proposed. The research results of this paper give a reference for the specification and use of intelligent boundary switches.