2687-0517 2 217 3 2015

RAR RUS 9-18 El-Khatib Samer samer_elkhatib@mail.ru Skobtsov Yury skobtsov@kita.dgtu.donetsk.ua System of Medical Image Segmentation Using Ant Colony Optimization The article presents an image-processing system using a combined ant colony optimization algorithm andK-means. We have developed the possibility of working directly with «raw» medical images and integrationinto the segmentation algorithm of K-Means and ant colonies. We have implemented and analyzed amixed segmentation algorithm of K-means and ant colony optimization. The software system has beenimplemented to visualize and test the developed algorithm. We have tested the algorithm using data fromthe Ossiriss system. We have also compared the obtained data of the developed method with automaticalgorithms (C-Means) and interactive methods (the algorithm Magic Wand). The article shows the outputimages and values of heuristic coefficients of the algorithm. 10.5862/JCSTCS.217-222.1 004.932 segmentation ant colony optimization K-means algorithm image analysis medical system https://infocom.spbstu.ru/article/2015.44.1/ 01.pdf

RAR RUS 19-33 Glazunov Vadim neweagle@gmail.com Peter the Great St.Petersburg Polytechnic University Kurochkin Mikhail kurochkin_ma@spstu.ru

Polytechnicheskaya, 29, St.Petersburg, 195251, Russia

Peter the Great St.Petersburg Polytechnic University Popov Sergey popovserge@gmail.com

Polytechnicheskaya, 29, St.Petersburg, 195251, Russia

Wu Chunming wuchunming@zju.edu.cn Liu Leibo liulb@tsinghua.edu.cn Management Technology for the Cloud Service of the Telematics Map in the Intelligent Transportation System Continuous access to services from a mobile device improves safety and provides ecological functioning of the transport infrastructure of metropolitan agglomerations. The quality of these services is determined by two-way communication between the vehicle and the stationary high-performance computing (HPC) system. As the signal strength in global and local networks varies, there is a need for new approaches to provide a continuous data transfer between the vehicle and the stationary HPC system in highly dynamic networks. One of the promising ways to increase the duration of stable communication between the vehicle and the HPC system is through getting a list of available networks from an external source, e.g., a telematics map, which is a set of low-level methods that ensure the formation and management of the list of networks. Such a set allows retrieving data on the telematics environment from vehicles, summarizing it on a geographical and temporal basis, and implementing the vehicle requests for the data of the generalized map to plan connections to available telematics resources. The article describes a new technology that helps to establish the data networks of vehicles by using the telematics map and the multiprotocol node. The technology is based on collecting network data from multiprotocol nodes, transmitting the collected data to the cloud service of the telematics map, summarizing the data and executing queries about the available networks in a given region. Thus, the paper proposes a fully automatic technology to manage the data on the telematics environment of a region. In order to test the proposed approach, a prototype of the information system was implemented. Its tests demonstrated the feasibility of the technical solutions in regard to managing the lists of wireless networks as applied to vehicles. 10.5862/JCSTCS.217-222.2 004.77 telematics map cloud service intelligent transportation system multiprotocol node spatio-temporal database MESH-networks wireless networks WI-FI LTE https://infocom.spbstu.ru/article/2015.44.2/ 02.pdf

RAR RUS 34-40 Laboshin Leonid laboshin@ya.ru Lukashin Aleksey lukash@neva.ru Peter the Great St.Petersburg Polytechnic University Vladimir S. vlad@neva.ru

Polytechnicheskaya, 29, St.Petersburg, 195251, Russia

Applying Mapreduce and Network Traffic Analysis to Control Access to Information Resources Nowadays information security is an important issue. Network traffic analysis is widely used by Internet Service Providers to evaluate network performance, to collect statistics and to detect vulnerabilities. To analyze traffic traces collected from a large network it is required a computer system where both storage and computing resources can be easily scaled out to handle and process multi-Terabyte files. Cloud computing platforms and cluster file systems could provide resizable compute and storage capacity. The MapReduce programming model developed by Google in 2004 allows processing huge amounts of data in distributed manner by defining the map and reduce functions. The given paper proposes a cloud-computing framework based on a MapReduce approach for fast internet traffic analytics. 10.5862/JCSTCS.217-222.3 004.415 deep packet inspection Big Data MapReduce Hadoop https://infocom.spbstu.ru/article/2015.44.3/ 03.pdf

RAR RUS 41-48 Il’yin Anatoliy toly@rtc.ru Interpolation Algorithm for an Increasing Function by Exponential Splines The paper presents formulas for calculating the parameters of exponential and logarithmic basic functions for the interpolation of the increasing function set by the table. Cases with even and odd numbers of argument steps are considered. The main goal was to prevent interpolation convexity, so the joining of derivatives was relegated to minor positions. It has been suggested that jumps in a derivative should be regarded as quality factors for basic data when visual examination is performed. The test case involved comparing two variants of pasting basic functions and discussing the issue of form-keeping. An example of processing experimental calibration data is demonstrated. The procedure is presented in the Pascal language. 10.5862/JCSTCS.217-222.4 519.652 interpolation exponential splines calibration function formkeeping https://infocom.spbstu.ru/article/2015.44.4/ 04.pdf

RAR RUS 49-70 Solomenko Institute of Transport Problems of the Russian Academy of Sciences, University National Technology Initiative 2035 Seliverstov Yaroslav maxwell_8-8@mail.ru Methods and Models of the Construction of Transport Correspondence Matrix We have analyzed the methods and models how to construct correspondence matrix for urban transport processes. We have also identified the advantages and disadvantages of the models. Having analyzed these models, we chose the relational method of the construction of correspondence matrix. The EVA model of the construction of correspondence matrix was recognized as the most suitable. The article demonstrates how the distribution of traffic flows in St. Petersburg is changed due to the fact that new metro stations such as “Obvodny Kanal”, “Admiralteyskaya”, “Bukharestskaya” and “Mezhdunarodnaya” have been put into operation. We have also outlined technical prospects of the models of the construction of correspondence matrix. 10.5862/JCSTCS.217-222.5 656, 004.8, 007.5 , 51-74, 510.67 correspondence matrix transport modelling single factor growth method average growth rate Detroit method Fratara method gravity model entropy model competing centers model EVA model https://infocom.spbstu.ru/article/2015.44.5/ 05.pdf

RAR RUS 71-78 Akishin Alexandr Elmechtrans@mail.ru Savoskin A.N. elmechtrans@mail.ru Generating a Multidimensional Perturbed Stochastic Process in Railways Rolling Stock Dynamics When a train is moving along the tracks, it is exposed to perturbations in the form of vertical and horizontal irregularities of the right and left rails causing oscillations in the mechanical parts of the rolling stock. This raises the problem of generating a multidimensional stochastic process of disturbances in accordance with a given matrix of auto-correlation and cross-correlation functions or spectral densities. For this purpose, the forming mechanism in the temporary realm was designed. This mechanism allows generating multidimensional stochastic processes of irregularity at different speeds for any length of implementation or discretization step in the temporary realm. 10.5862/JCSTCS.217-222.6 629.4.015:519.246 generating multidimensional stochastic process perturbing factor causing oscillations of railway vehicles forming mechanism in the temporary realm https://infocom.spbstu.ru/article/2015.44.6/ 06.pdf

RAR RUS 79-92 Tchurkin Vitaly Estimation and Optimization of Spare Parts Kit Using the Method of Statistical Modeling The article considers a topical problem of the evaluation and optimization of stocks in the set of spare parts taking into account the direct connection of the set to the model of spare parts kit using the method of statistical modeling. The Monte Carlo method has not been used to design optimal sets of spare parts due to the lack of sufficient literature. It is not still well defined how to apply the Monte Carlo method in this area. The given article provides issues how to apply the Monte Carlo method to solve the given problem. 10.5862/JCSTCS.217-222.7 004.942 set of spare parts a method of statistical modeling Spare Parts Kit optimization Spare Parts Kit evaluation https://infocom.spbstu.ru/article/2015.44.7/ 07.pdf

RAR RUS 93-104 Kozionov Alexey alexey.kozionov@gmail.com Pyayt Alexander alexander.pyayt@siemens.com Mokhov Ilya Ivanov Yuri upi@mail.ru Research on Gap-filling Algorithms for Dike Health Monitoring Systems Earthen dam (dike/levee) health monitoring is a challenging task. Monitoring algorithms have to detect anomalies in dike behavior in on-line mode basing on measurement collected from the sensors installed in the dike. One of the important monitoring-related challenges for dike health monitoring is the quality of sensors measurements. There are often gaps that occur due to failures, outages of transmission or data collection systems, incorrect configurations of systems and other internal and external factors, therefore it is necessary to improve the quality measurement using specific algorithms. The proposed approach is based on adaptive algorithms filling in gaps in sensor measurements in the conditions of a priori uncertainty of signals models. The algorithms (based on the autoregressive model, Caterpillar-SSA, Fourier transform) presented in this paper use historical data for signal reconciliation. The description and analysis of the algorithms are also included in the paper. The algorithms have been tested at the Boston dike, Great Britain. The research findings and algorithms are implemented by Siemens in the UrbanFlood Early Warning System. 10.5862/JCSTCS.217-222.8 681.51 gap-filling measurement reconciliation autoregressive model spectrum singular analysis Fourier transform dike health monitoring intelligent signal processing https://infocom.spbstu.ru/article/2015.44.8/ 08.pdf

RAR RUS 105-114 Khurshudov Artem art1783@gmail.com Using an Ensemble of Transforming Autoencoders to Represent 3D Objects One of the key goals of computer vision-related machine learning is to obtain high-quality representations of visual data resistant to changes in viewpoint, area, lighting, object pose or texture. Current state-of-the-art convolutional networks, such as GoogLeNet or AlexNet, can successfully produce invariant representations sufficient to perform complex multiclass classification. Some researchers, however, (Hinton, Khizhevsky, et al.) suggest that this approach, while being quite suitable for classification tasks, is misguided in terms of what an efficient visual system should be capable of doing: namely, being able to reflect spatial transformations of learned objects in a predictable way. The key concept of their research is equivariance rather than invariance, or the model's ability to change representation parameters in response to different poses and transformations of a model-specific visual entity. This paper employs Hinton's architecture of transforming autoencoder neural networks to identify lowlevel spatial feature descriptors. Applying a supervised SVM classifier to these detectors, one can then represent a sufficiently complex object, such as a geometric shape or a human face, as a composition of spatially related features. Using the equivariance property, one can also draw distinctions between different object poses, e.g., a frontal face image or a profile image, and then, be able to learn about another, higher-leveled transforming autoencoder via the same architecture. To obtain initial data for first-level feature learning, we use sequences of frames, or movies, and apply computer vision algorithms to detect regions of maximum interest and track their image patches across the movie. We argue that this way of learning features represents a more realistic approach to vision than general naive feature learning from a supervised dataset. The initial idea came from the concept of one-shot learning (by Fei-Fei et al.), that suggests a possibility of obtaining meaningful features from just one image (or, as in this study, a rather limited set of images supervised by time and order). 10.5862/JCSTCS.217-222.9 004.923 transforming autoencoder one-shot learning equivariant representation capsules https://infocom.spbstu.ru/article/2015.44.9/ 09.pdf

RAR RUS 115-124 Khurshudov Artem art1783@gmail.com Constructing 3D Feature Maps from Video Sequences by Optic Flow Estimation The study presents a general case of structure-from-motion problem where the given data consists of a bunch of video sequences filmed in the same scene. Unlike the popular methods of photogrammetry and bundle adjustment, the proposed solution does not required specific knowledge of intrinsic camera parameters, could be applied to any type of consistent motion pictures and can handle large amounts of noise. During the process of reconstruction an object is viewed as a 3D map of robust sparse features, which at first hand are discovered in certain key frames (using existent computer vision techniques like Shi-Tomasi corner detector) and afterwards tracked across the following frames using sparse optic flow method. When camera motion (egomotion) data is available, it is became possible to estimate each feature's depth by using simple geometric properties of two-image disparity, and having each feature estimated from multiple video frames allows to effectively filter out the noise. Apart from sparse Lucas-Kanade optic flow the study also makes use of some properties of dense optic flow (Gunnar Farneback's algorithm), which is used for scene segmentation during the camera motion. The resulting 3D feature maps are designed to be used as a macro object detector that could be applied to any previously unknown single digital images, representing structures that are believed to store 3D visual memory of an object, and therefore being able to detect objects in spite of general invariant scene transformations. 10.5862/JCSTCS.217-222.10 004.923 object detection optical flow Lucas–Kanade algorithm sparse features watershed segmentation https://infocom.spbstu.ru/article/2015.44.10/ 10.pdf

RAR RUS 125-138 Martsenyuk Mikhail Seletkov Il`ya iseletkov@gmail.com Fuzzy Cellular Automata for Temperature Field Control The article considers the models of memoryless and memory-powered cellular automata based on fuzzy logic. It shows that the use of fuzzy logic methods in cellular automata algorithms allows increasing greater flexibility and simplifying the formalization of expert knowledge. We have discussed the problem how to calculate the dynamic changes of the temperature field with the given boundary conditions. A matrix approach of fuzzy logic has been used for all calculations. This approach simplifies the implementation of the algorithm and reduces the requirements for computational resources. We have also shown how to take into account temperature memory and how the results could be applied to solve a more complex problem of a temperature field control. 10.5862/JCSTCS.217-222.11 62.50 + 517.11 + 519.92 fuzzy cellular memoryless automata fuzzy cellular automata with memory temperature filed control temperature memory matrix approach of fuzzy logic https://infocom.spbstu.ru/article/2015.44.11/ 11.pdf

RAR RUS 139-161 Institute of oriental manuscripts of Russian Academy of Sciences Seliverstov Sviatoslav amuanator@rambler.ru Solomenko Institute of Transport Problems of the Russian Academy of Sciences, University National Technology Initiative 2035 Seliverstov Yaroslav maxwell_8-8@mail.ru Construction of Intelligent System for the Organization and Development of Transport System Metropolis The article proves the necessity to develop a radically new system of an intellectual spatial organization and the development of metropolis. We have analyzed the approaches of the modern metropolis taking into account the development of its operational environment and transport system. We have investigated the issues regarding the cooperation of land-use, transport and urban development and analyzed the existing architectures of intelligent transport systems. To develop the intellectual system of the organization and development of the metropolis transport system, a logical-algorithmic language has been used. We have firstly developed a system of REFISANIK to ensure structural and functional homogeneity. Using a logicalgorithmic language we have described the process of synthesis and functioning of main units of the system of REFISANIK including a block system, a selective filter, a combinatorial extender of indicator links and relational regulatory approval. Practical examples demonstrate the operation of the system and its components. 10.5862/JCSTCS.217-222.12 656, 004.8 intelligent system for the organization and development of transport system metropolis intelligent urbanization intelligent transport system https://infocom.spbstu.ru/article/2015.44.12/ 12.pdf