P. Sykora, Patrik Kameneay, R. Hudec, M. Benco, M. Šinko
{"title":"Comparison of Feature Extraction Methods and Deep Learning Framework for Depth Map Recognition","authors":"P. Sykora, Patrik Kameneay, R. Hudec, M. Benco, M. Šinko","doi":"10.23919/NTSP.2018.8524109","DOIUrl":"https://doi.org/10.23919/NTSP.2018.8524109","url":null,"abstract":"In this paper a comparison between three feature extraction methods (Fourier Transform, Radon Transform, Canny Edge Filter) and Convolutional Neural Network is presented. These methods are tested on set of depth maps. The Microsoft Kinect camera is used for capturing the images. For the image classification the Support Vector Machine with Radial Basis Function kernel was used. The experimental results from each tested method are stored in confusion matrix. Each row in this matrix represents actual class of tested data and each column represents predicted class. The quality of the Convolutional Neural Networks features has been compared with traditional methods of feature extraction. From the experimental results, we have shown that the Convolutional Neural Network based deep learning framework achieve better classification performance than test feature extraction methods (Fourier Transform, Radon Transform, Canny Edge Filter).","PeriodicalId":177579,"journal":{"name":"2018 New Trends in Signal Processing (NTSP)","volume":"43 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124926994","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Synchronization Verification Improvement of Two Tree Parity Machines Using Polynomial Function","authors":"Martin Javurek, M. Turčaník","doi":"10.23919/NTSP.2018.8524106","DOIUrl":"https://doi.org/10.23919/NTSP.2018.8524106","url":null,"abstract":"One of the problems of cryptography based on the tree parity machine is how to recognize that two tree parity machines are synchronized. In this paper, one of the possible solutions based on polynomial function is described. The proposed method is focused on obtaining of one unique number by polynomial function. Next is improvement of security during verification of the synchronization process using memory of weights and polynomial function.","PeriodicalId":177579,"journal":{"name":"2018 New Trends in Signal Processing (NTSP)","volume":"20 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116847255","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Small-Sized Seismic-Acoustic Sensor System Based on MEMS Accelerometer and MEMS Microphone","authors":"Juraj Guzma, R. Berešík, Jozef Puttera","doi":"10.23919/NTSP.2018.8524100","DOIUrl":"https://doi.org/10.23919/NTSP.2018.8524100","url":null,"abstract":"There are several sensor systems concepts designed based on high performance digital signal processors and high-resolution analog to digital converters, which are capable to act as decentralized sensor system in military and security applications. However, there is also need to design the small-seized, low cost sensor systems which are able to detect events by multiple sensors in particular spot of monitored area and at the same time to be part of the higher structure of multisensor systems. The key part of multisensor systems design is to use the tasks oriented sensors and keep the complexity of sensor system from hardware and software site as simple as it is possible. As a variety and number of sensors used in multi-sensor system increases, there is challenge to achieve the real-time signal processing. The paper deals with the design of acoustic-seismic sensor system based on digital Micro-Electro-Mechanical accelerometer, analog Micro-Electro-Mechanical microphone and 8-bit microcontroller. The main part of the paper is focused on sensor system parameter optimization and design of algorithm for acoustic and seismic events detection and subsequent data transferring for a further signal processing in real-time. In conclusion, outcomes of experiments with designed sensor system are evaluated in terms of the ability to detect seismic and acoustic events in real-time.","PeriodicalId":177579,"journal":{"name":"2018 New Trends in Signal Processing (NTSP)","volume":"81 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114305162","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Tomas Hnilicka, V. Schejbal, Ondrej Fisher, T. Zálabský
{"title":"Design of Square Waveguide Corner","authors":"Tomas Hnilicka, V. Schejbal, Ondrej Fisher, T. Zálabský","doi":"10.23919/NTSP.2018.8524076","DOIUrl":"https://doi.org/10.23919/NTSP.2018.8524076","url":null,"abstract":"The paper presents a full-wave design for matching a square mitered waveguide corner. We compare the full-wave numerical simulations with experimental results. The corner can operate with both polarizations, i.e. TE10for magnetic plane bend and TE01for electric plane bend. That is very useful for intended meteorological radar. That could use dual polarization transmit and receive both horizontal and vertical polarizations to assess additional properties of the weather targets.","PeriodicalId":177579,"journal":{"name":"2018 New Trends in Signal Processing (NTSP)","volume":"31 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130323653","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Time-Frequency Peak Filtering Versus Kalman Filter in the Kinematic Trajectory Estimation","authors":"D. Mikluc, M. Andric","doi":"10.23919/NTSP.2018.8524127","DOIUrl":"https://doi.org/10.23919/NTSP.2018.8524127","url":null,"abstract":"The kinematic trajectory of point targets with time-frequency peak filtering and Kalman filtering has been used in this paper. The comparison of the named filters is done on position estimation of several characteristic generated trajectories. Firstly, the error estimation of a point target trajectory was analysed in one dimension and then in the plane. The criterion of a root mean square error is used to compare the estimation results of the used filters. It is shown that time-frequency peak filtering can be used in the trajectory estimation in various velocity values as well as in different variances of Gaussian noise.","PeriodicalId":177579,"journal":{"name":"2018 New Trends in Signal Processing (NTSP)","volume":"31 4","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114119226","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Accelerometer Calibration Based on Spectral Analysis","authors":"K. Draganová, P. Lipovský, M. Šmelko","doi":"10.23919/NTSP.2018.8524056","DOIUrl":"https://doi.org/10.23919/NTSP.2018.8524056","url":null,"abstract":"The calibration method using the spectral analysis is based on the continuous sensor rotation around the sensitivity axes placed perpendicularly to the plane of the vector of gravity. The rotation can be continuous, because for the calculations based on the spectral analysis the assignment of samples to the particular degrees of the sensor rotation is not required. For this calibration methodology only the final amount of measured samples is substantial and that enables creation of an automatized workstation for the accelerometer calibration with the focus on the determination of bias and scale factor constants of the sensors' channels.","PeriodicalId":177579,"journal":{"name":"2018 New Trends in Signal Processing (NTSP)","volume":"106 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"117233388","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"New Trends for Establish an Innovative Theoretical Approach Between a Train of Digital Pulses and a Set of Real Polynomials Using a Nonlinear Operator","authors":"M. Silveira, G. Figueiredo, R. G. Caputo","doi":"10.23919/NTSP.2018.8524114","DOIUrl":"https://doi.org/10.23919/NTSP.2018.8524114","url":null,"abstract":"The main purpose of this paper is to introduce an original theoretical approach for establish a one to one linear equivalence between a class of an arbitrary train of digital pulses with a class of real polynomials. This innovative technological modeling the authors will define as the polynomial signal processing, which can serve as an alternative approach for those systems that will demand the theoretical modeling using the digital signal processing, which is a very recognized tool that have been used for a wide class of communications systems in the current literature. We have a direct application of these studies for those analog or digital systems that demand a very high transmission rate, and some experimental results will be presented in this manuscript confirms that the approach presented for the first time in the current literature have a very consistent and compatible performance.","PeriodicalId":177579,"journal":{"name":"2018 New Trends in Signal Processing (NTSP)","volume":"29 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131704732","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Comparison of Speed and Accuracy of Chosen Magnetometer Calibration Algorithms in the Presence of External Interference","authors":"T. Kliment, P. Lipovský, K. Draganová, V. Moucha","doi":"10.23919/NTSP.2018.8524105","DOIUrl":"https://doi.org/10.23919/NTSP.2018.8524105","url":null,"abstract":"The paper deals with the robustness of the calibration algorithms based on the neural networks against external interference. The chosen neural networks distinguish only in the learning rule. For a comparison the backpropagation, backpropagation with momentum and resilient propagation learning algorithms were chosen. The algorithms were examined from the calibration speed and accuracy point of view defined by the specified criterions. Tests were performed in the form of simulations with the applied measured periodical interference and the periodical interference with the random occurring amplitude modulation. The goal of the article is to prove the fundamental robustness of the algorithm against the interference and the response of the chosen algorithms to the various external interference types.","PeriodicalId":177579,"journal":{"name":"2018 New Trends in Signal Processing (NTSP)","volume":"21 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128610786","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}