Milena Jevtić

The influence of the CAPTCHA types to its solving times

In this paper, we present a survey of different CAPTCHA types. Then we address the problem of the CAPTCHA userability. In the experiment, the Internet user response to solve different types of CAPTCHA is tested. The obtained results are given and statistically processed. They are discussed leading to the conclusion different types of CAPTCHA usability and suitability for Internet users.

Solving a combined economic emission dispatch problem using adaptive wind driven optimization

In this paper, the adaptive wind driven optimization (AWDO) algorithm is applied for solving the combined economic emission dispatch (CEED) problem. AWDO is one of the newest hybrid algorithms, which optimizes the selection of coefficients at each iteration, eliminating the need for tuning the coefficients. The evaluation of AWDO performances is carried out on the standard IEEE 30-bus test system with 6 generating units and with various cost curve natures. The results of AWDO use with the test system are compared against the results of use of 3 algorithms: the moth swarm algorithm, firefly algorithm, and hybrid particle swarm optimization and gravitational search algorithm, which were proposed in recent literature for solving this problem. The present paper shows that AWDO gives an accurate and effective solution of the CEED problem and outperforms the other tested algorithms.A miniaturized lowpass filter (LPF) with ultrasharp response, good figure of merit (FOM), and simple structure is proposed. The designed structure is fabricated and measured with a 2.9 GHz cut-off frequency. The proposed LPF consists of high impedance lines that are loaded by three similar resonators and two uniform suppressing cells. The filter size is only 0.08~

Experimental Comparisons of Metaheuristic Algorithms in Solving Combined Economic Emission Dispatch Problem Using Parametric and Non-Parametric Tests

\times

Classification of German Scripts by Adjacent Local Binary Pattern Analysis of the Coded Text

~0.23~

Discrimination between Serbian and Slovenian language by texture analysis

\lambda

Moth Swarm Algorithm for Solving Combined Economic and Emission Dispatch Problem

g, which indicates small circuit size. The proposed design exhibits good features, such as an ultrasharp roll-off rate of about 1609 dB/GHz and a good FOM of 162,820. The insertion loss of the proposed LPF is less than 0.14 dB in the passband. With these excellent obtained specifications and flat group delay, this structure can be used in wireless antennae.

Document Image Coding and Clustering for Script Discrimination.

ABSTRACT In this paper, the parametric and non-parametric statistical tests are applied for comparisons of metaheuristic algorithms’ (MAs) behavior in solving Combined Economic Emission Dispatch (CEED) problem. In the last years, in many published papers, a large number of MAs have been proposed to solve CEED problem of different dimensions consisting of different objective functions. In this paper, the statistical tests are applied over samples of results obtained for eight objective functions of CEED problem using the four MAs: Firefly Algorithm, Moth Swarm Algorithm, Adaptive Wind Driven Optimization and Particle Swarm Optimization-Gravitational Search Algorithm. The standard IEEE 30-bus six-generator test system is used. The statistical tests are applied over results obtained for each function and over results obtained for all eight functions simultaneously. The analysis of the results of statistical tests over a single function shows that one MA statistically behaves differently for different functions and one MA is not the best for each function of CEED problem. Therefore, more MAs are more acceptable than one MA for solving a specified CEED problem. However, the analysis of the results of statistical tests over all functions simultaneously shows that all four MAs statistically behave in the same way.

Controlling the thermal environment of underground power cables adjacent to heating pipeline using the pavement surface radiation properties

The paper proposes a script classification method which is based on textural analysis of the script types. In the first stage, each letter is coded by the equivalent script type, which is defined by its baseline position. Obtained coded text is subjected to the adjacent local binary pattern analysis to extract the features. The result shows the diversity of the extracted features between scripts, which makes the feature classification easier. It is the basis for decision-making process of the script identification by automatic classification. The proposed method is tested on an example of synthetic and historical German printed documents written in Antiqua and Fraktur scripts. The experiment shows very positive results, which proved the correctness of the proposed algorithm.

Modelling the thermal effect of solar radiation on the ampacity of a low voltage underground cable

The paper proposed the method for the language identification according to texture analysis. First, the algorithm encrypted the text given in different languages to the cipher based on the baseline status of each script element in text. Then, the cipher was subjected to the texture analysis. The aim of this analysis was the extraction of texture features. Obtained texture features showed significant diversity between languages. Hence, they were suitable for creating the discrimination and identification criteria between languages. The proposed method was tested on Serbian and Slovenian documents from custom oriented database. The experiments gave encouraging results.