Proceedings of the 2015 Federated Conference on Software Development and Object Technologies

In conventional monitoring systems video previews from stationary cameras are overseen only by a human supervisor, who may easily overlook alarming events recorded by a camera. Because surveillance system must be reliable, its capabilities can be improved by applying computer vision algorithms to a video signal in order to detect objects in an automated fashion. Also its autonomy can be extended by the use of mobile robots capable of monitoring tight and occluded areas and by the use of smart cameras with integrated embedded systems. In this paper we introduce an architecture of the autonomous monitoring system based on object shape detection. Our approach is conformable with the concept of Internet of things. It consists of the set of smart objects with video sensors, controlled by the Shape Identification Cloud. Our work is aimed at building the real-time system efficient at reliable recognition of objects on the basis of their approximate shape and with the option to be used as a web service in a cloud. To monitor the environment the system uses mobots equipped with video sensors as well as surveillance cameras capable of remote position control. For object identification task we use the Query by Shape (QS) method which decomposes objects into simple graphical primitives like lines, circles, ellipses etc. and then it identifies them in a shape database.

The efficient usage and development of embedded multi- and many-core systems is an important field of research for years and decades. In the last decade, utilizing embedded and especially low-power architectures for high performance computing (HPC) applications became an important part of research. The reason for this are the constantly increasing energy costs along with an increasing awareness of energy consumption in general. As suitable low-power HPC architectures are not yet available at a larger scale, simulation of new and appropriate architectures becomes an important factor for the success of low-power systems and clusters. In order to speed up simulation, at the cost of accuracy, different levels of abstraction were introduced. Currently the class of instruction accurate simulations seems to yield the best trade-off between speed and precision, especially when simulating complex multi- and many-core systems. In this paper we present our investigations about the accuracy of the state-of-the-art instruction accurate embedded multi- and many-core simulation environment Open Virtual Platforms (OVP) in comparison to an actual embedded hardware system from Altera. Our investigations include the actual usage of the same operating system running on both, the target hardware and the simulation as well as serial and parallel software benchmarks. We analyze the current accuracy of the simulation environment with respect to a performance model, based on the execution time of the simulation and the actual embedded hardware system. Using the instruction accurate simulation technology from OVP is for the simulation of embedded/low-power HPC hardware architectures and applications. Furthermore, we point out first steps towards further possibilities for obtaining a better performance model by the use of our simple memory model.

The parallel logic controllers (PLC) developing process can be both, simplified and improved by means of formal methods.

The paper presents approach based on Petri nets specification and model checking techniques used for formal verification, synthesis and implementation. Interpreted petri net (IPN) is selected as a formal behavioural model for parallel logic controllers. It is proposed to use graphical modelling tools for formal behavioural PLC modelling, authors ICPN is such example. The use of common standard, like Petri Net Markup Language (PNML) enables integration with formal analysis tools. The model is simplified (optimised) by means of formal reasoning system (Gentzen). The transformation of the simplified model is made automatically into VHDL description and NuSMV model. The resulting VHLD model can be used for simulation and next for synthesis and implementation. The reliability of the PLC is improved by formal verification. The paper presents the application of the additional specification described in the temporal logic. Model received from reasoning system can be verified by such specification. The formal verification enables to locate deviations from the specification. Proposed approach is useful especially in PLC rapid prototyping approach. The changes in the specification can be verified immediately, towards general specification requirements. Discrepancies between specification and the prototype are localised and can be removed before next iteration. Proposed approach improves visual analysis and fast modifications, ensuring high reliability of the constructed logic controller. The formal methods increases the reliability and quality of the parallel logic controller.

This article brings new approach to optical network optimization. When routing data in fast optical networks, defined parameters of each path are taken into account and evaluated to find optimized path for data routing. Our fuzzy logic approach implements fuzzy weighed parameterization to achieve optimal path setup in decision process, over WDM network structure. Complex path evaluation and dynamic routing using fuzzy logic can provide better results and faster decisioning for data transport in optical networks.

Based on data fetched via the On-Board- Diagnostics interfaces of cars, a comprehensive solution was created to provide functionality like automatically maintained driver’s logbooks within a mobile phone application. Advanced on-board hardware modules and a dynamic client/server architecture allowed the design and implementation of an extensible system with optional integration of third-party software and infrastructure. This includes a mobile phone application which can be extended with new functionality according to customer configuration settings and payments. The project posed challenges like time and location independent availability, privacy and security issues, the dynamic extension of mobile applications with platform-independent code, and also economic restrictions. This paper contains a description of these challenges, the technologies and concepts explored and utilized, and the final system architecture and implementation decisions taken during product development. First experiences gained during deployment and commercialization already show the aptitude of the system and the impact of the design decisions taken.

Rapid development of miniaturization technologies helps create small and easily portable devices. This enables users to work with their information systems (IS) almost anywhere. The only obstacle, however, may be a poor network connection, which can make the system inaccessible. One of the solutions is creating a distributed information system with asynchronous data update. Then the user operates over the local data, and they do not necessarily need an instant network connection through a server. Later, with a connection available, the local data are merged with the rest of the system. This concept will effectively solve the situations in which two users are editing the same data. Below we present a new architectural pattern for designing such IS.

The paper deals with principal component analysis and data whitening. The research is done in the area of main economic indicators. This means the data preprocessing problem. The main aim of this paper is to present and discuss the possible ways of data preprocessing. The paper deals with four main approaches. There are compared the results from raw data, absolute differences, relative differences and logarithmic differences. The classic principal component analysis is also used with some improvement, there is described the basement of data whitening. The main aim is to get the good data visualisation. The next aim of such approach can be to identify the similarities between some states and their main trends. For this reason there is presented the comparison of states of Visegrad Group. At this moment there is no aim to deeply discuss the reasons of development in detail. This paper suggests new point of view to time series connected to economic development. The deep analysis of all relationships is the topic for further research.

Although currently the bulk of the most common algorithmic tasks is included in libraries of programming languages, it is necessary to realize, that upon completion of object oriented design of application, we still do not avoid of using more complex algorithmic constructions. For its visual projection and easier understanding several graphic languages are used. Architecture First methodology for its purpose prefers kopenograms, as one of the most suitable method of displaying structured algorithm. This paper deals with the tool, which was added to IDE BlueJ in order to improve support of Architecture First methodology by this IDE and which allows students to show kopenogram of selected method in a simple manner in interactive mode.

The aim of hydrological models is to represent physical processes of catchments. The models provide information about hydrological processes. Evapotranspiration is a main model output. The quality of the model output is highly dependent on a model calibration which is not a simple process. SAC-SMA is one of rainfall-runoff models. In SAC-SMA, evapotranspiration is defined by 12 discreet values. Optimization algorithms do not return valid calibrations for specific basins or time periods because model parameters are correlated or also, because the optimal solution found by the algorithm is not applicable in hydrology. Some model parameters can be interpreted by a continuous function. It provides higher precision of evapotranspiration and a faster optimization run than it is provided by the definition of evapotranspiration by 12 discreet values. The main goal of this article is to determine a continuous function which can replace the 12 values definition of evapotranspiration. It should make the optimization process faster because the continuous functions are described by less than 12 parameters. Evapotranspiration is very similar to the Gaussian function. Genetic algorithm has been modified to SAC-SMA model optimization which uses the continuous functions for a description of evapotranspiration. This modification has brought interesting results. The continuous definition of evapotranspiration provides better results for specific time periods. The optimization speed can be up to a one third faster. The next step will be to confirm the approach by another catchments and time periods.

This paper focuses on implementation and benefits of the cache module for the dictionary writing system (DWS) Alexis. At first, the DWS Alexis that is being developed at the Institute of the Czech Language is introduced. We shortly present architecture of the system and briefly describe its main modules including list of entries, editing, or output modules. The main emphasis of this contribution is put on the recently developed cache module. We explain design and implementation of the cache, then we present its benefits for the dictionary writing system. At first, we analyze performance improvement of the output module, then we explain its roles in the fulltext search in the web interface of the dictionary and source of offline data for mobile devices. We conclude with a current state of development and plans for the nearest future.

The control process can be understand as the specification of the manufacturing process in factory or specification of the control program for logic controller. The aim of the control process management can be understood quite widely, especially as the resources usage optimisation or time reduction. The paper presents the application of the evolutionary algorithm towards the loops reduction. The unnecessary repeating of the tasks can increase the time and costs of the control process. The UML formal models are popular method of the control process specification. The Proposed approach enables optimisation of the control process specified by means of the UML activity diagrams. Both, manufacturing process and logic controller behaviour can be specified as well.

The present paper proposes a new investigation area in automata theory - n-parallel jumping finite automata. These automata further extend recently presented jumping finite automata that are focused on discontinuous reading. The proposed modification uses multiple reading heads that work in parallel and can discontinuously read from the input in several places at once. We also define the more restricted version of these automata which only allows jumping to the right. This restricted version is then further studied, compared with n-parallel right linear grammars, and several of its properties are derived.

We are introducing Scalable Distributed Two-Layer Datastore. The system that is an efficient solution while storing relatively big multimedia files. In the article we are focusing on storing high-resolution photos. We are introducing some of the key implementation concepts as well as the careful evaluation. We are comparing our solution with two of the most recognizable data storing systems: MongoDB and Memcached.

Temporal database approach is one of the most significant sphere of data processing, the basic part is exact time management, changes and progress monitoring over the time. Conventional database does not support time management, historical data can be processed only partially by difficult and time consuming transformations. Object level principle uses state granularity, thus the whole state is updated, if any attribute value is changed. This approach can be inappropriate, if the frequency and granularity of the changes is not the same and even time varying. This paper deals with object level and attribute level temporal architecture. Usually, data are shared, however, not all data should be accessible to all users, and therefore concept of index definition using contexts is proposed. In terms of efficiency, indexes are restructured using pointer mapping, hybrid context trees are defined.

Recent prognoses about the future of Cloud Computing, Internet of Things and Internet Services show growing demand for an efficient processing of huge amounts of data within strict time limits. First of all, a real-time data store is necessary to fulfill that requirement. One of the most promising architecture that is able to efficiently store large volumes of data in distributed environment is SDDS (Scalable Distributed Data Structure). In this paper we present SDDS LH_RT, an architecture that is suitable for real-time cloud applications. We assume that deadlines, defining the data validity, are associated with real time requests. In the data store a real-time scheduling strategy is applied to determine the order of processing the requests. Experimental results shows that our approach significantly improves the storage Quality-of-service in a real-time cloud environment.

The paper deals with detection of structures in source codes employing statistical classification. To enhance source code perception by development tools like code editors, modeling tools and source code repositories, various methods of patterns classification are proposed and tested. To be able to apply classification algorithms, well-defined feature space is required. Thus, such a feature space is presented and tested. Sub-models search is carried out by a genetic algorithm to select the optimal feature space subset without deterioration of a classification system. The results show that with standard classification algorithms like k-NN or Perceptron, accuracy of 0.8 can be achieved.

There are several different approaches to teaching programming, based on programming styles. A concept “first the object-oriented style, then the other one” known as “object-first” is currently being promoted by a number of technical colleges. The reasons originate from the practical area. Also the teaching of programming at the College of Polytechnics Jihlava (COPJ) is being switched from the “structural-then-object” style of programming to the “object-first” style. After the second run, the results achieved by students do not seem (very) good. This can be confirmed from the courses’ feedback. It seems that the students are puzzled and their skills are poorer. Therefore, we decided to examine the results achieved by students and their opinion on the “object-first” style especially. This survey was carried out after the completion of the course with “object-first” teaching and at the beginning of the course with “structural” teaching. We are interested in skills in object-oriented programming and also in structural programming, but especially skills at the beginning of study at COPJ and the type of completed high school. We addressed our students attending the course of “structural programming”. The third run of this teaching approach started. In this paper we introduce the first survey results. Even though the number of respondents is not big, the statistic results are significant within the College.

The primary idea of the IP Fast Reroute (IPFRR) technology is to reduce the network recovery time after a link or router failure within an ISP network. The key feature of existing IPFRR mechanisms for reaching low recovery time is the usage of pre-computed alternative backup paths. These alternative backup paths have to be pre-calculated before a network failure will occur.

The calculation of the alternative backup path utilizes the specific information about destination networks, and thus most of existing IPFRR mechanisms are dependent on the distance-vector routing protocols (RIP, EIGRP). Other IPFRR mechanisms requires an additional information about the network topology, and therefore strongly depend on the usage of link-state routing protocols (OSPF, IS-IS). The paper is focusing on the analysis of existing IPFRR mechanisms and is identifying and presenting theirs primary problematic areas.

A method of synthesis of software for low-power real-time embedded systems is presented in this paper. A function of the system is specified in the form of the task graph, then it is implemented using embedded processors with low-power and high-performance cores. The power consumption is minimized using the developmental genetic programming. The optimization is based on finding the makespan, satisfying all real-time constraints, for which the power consumption is as low as possible. We present experimental results, obtained for real-life examples and for some standard benchmarks. The results show that our method gives better solutions than makespans obtained using existing methods.

One of the best IDE for introduction courses of object oriented programming is the BlueJ IDE. This IDE was developed with respect to needs of the absolute beginners. It offers almost everything what we need for teaching according the Architecture First methodology. However later, when the students go over to some professional IDE, they lose many of the BlueJ excellent features, especially the ability to design the program and its architecture in the interactive mode. The paper shows, how the BlueJ IDE was modified to work as fully functional plugin for NetBeans IDE and which new possibilities we obtain with it.

The paper presents the designed algorithm, which is able to integrate of inertial sensor data into control algorithm. Autonomous operation of the mobile system requires reliable measurement of its position. Sources of such data are various; most commonly used is global satellite navigational system. However, this technique can be used only outdoors. For navigation inside building, under metal roof or underground only inertial or contact methods are available. This article analyzes possibilities of deployment of the inertial navigation in the control of the wheeled mobile platform. Experimental platform uses inertial measurement unit x-IMU manufactured by x-IO Technologies. According to our experiments inertial navigation can be reliably used only in fusion with other absolute sensors (odometers, magnetometers).

The Consortium for IT Software Quality (IT-CISQ) standard claims to provide valid measures as well as an assessment method that are suitable for properly measuring software quality. We implemented the measures for maintainability as specified by the IT-CISQ standard and wanted to find out whether both – the IT-CISQ defined assessment method and the IT-CISQ measures – are suitable for determining the maintainability of object-oriented systems. We identified a reference study that classifies the maintainability of eight open-source Java projects. This study follows a comprehensive measurement and assessment process and therefore can be used for validating the IT-CISQ approach. Due to the missing consideration of project size metrics, the IT-CISQ assessment method is not capable of properly determining the quality of projects. Even considering size metrics does not substantially enhance the result, which is an indicator that the measures proposed by IT-CISQ do not properly measure maintainability. Finally, our benchmarking approach was applied. It sets the measurements in relation to 26 projects that constitute the benchmark base. Despite a lack of statistical significance, the benchmarking results show a better correlation with the ranking published by the reference study. As our benchmarking approach is well validated, we can conclude that the measures proposed by IT-CISQ have to be considerably enhanced, i.e., additional measures have to be added to be able to determine the maintainability of object-oriented software projects.

Software requirements analysis is a critical phase in the software development life cycle. It is usually carried out using one of the following: use cases, user stories or business scenarios. Use cases, user stories and business scenarios thus become inputs for the object-oriented analysis (OOA) of application software development. Most translating and mapping of software requirements from text to objects and classes creates problems with software acceptance. There is also a practical issue with object-oriented analysis, namely, that OOA and object-oriented design (OOD) both operate with the same objects and classes, and, in practice, it has not yet been possible to separate them from one another. The goal of this article is to provide a solution that will simplify software requirements analysis and separate such analysis from design, in order to give architects, developers and testers the ability to work independently through a contract (interface) that integrates their work.

Modeling protean objects, i.e. objects adapting their structure and behavior dynamically with respect to changeable environment, may be challenging in traditional object oriented languages. While some dynamic languages make the implementation of changeable behavior of objects possible by injecting code into the objects at run-time, their lack of an in-depth compile-time code analysis makes the resulting code fragile. The concept of object metamorphism (OM) targets the gap between the two language types by introducing a declarative modeling of protean objects. Such a model, which is validated at compile-time, defines all possible compositions of a given object from smaller parts represented by traits. The actual composition used to instantiate the object is chosen dynamically with respect to the current environment. The result of this research will provide the theoretical concept of OM along with a proof-of-concept adoption of OM to Scala.

Animations can help students to comprehend computer science algorithms. Previous experiments, mentioned in this paper, show that interactivity is very important in educational animations. In this contribution we also describe three categories of algorithm animations with different views, and we introduce our interactive card animations that belong to the first group (animations with conceptual views). These card animations of sorting algorithms were used in our experiment, where first-year computer science students were asked to fill out a pre-test, use the animations, and fill out a post-test. In the third part of the paper we discuss the obtained results, which proved that the interactive card animations can help students to understand the essential aspects of different sorting algorithms. Finally, we draw conclusions and introduce our future plans.

In this paper, we introduce an incremental approach to semantic clustering, designed for software visualization, inspired by behavior of fire ant colony. Our technique focus on identification of equally sized but natural clusters that provides better hindsight of software system structure for development participants. We also address performance issues of existing approaches by maintaining similarities based on global weights incrementally, using subspaces and covariance matrix. Effectivity of visualization is improved by representing multiple documents with precise medoid approximation.

The article summarizes activities and results regarding pattern recognition for the JEM-EUSO experiment done by Slovak group of JEM-EUSO collaboration. The activities include estimation of trigger probability of false positives and reconstruction of simulated UHECR showers in UV background using Euso Simulation and Analysis Framework (ESAF). The Hough transform-based techniques are presented as methods to find UHCER showers in the JEM-EUSO detector recorded data. Additionally, the article describes structure and data flow of the framework.