Mitko Kostov

Mitko Kostov
"St. Kliment Ohridski" University - Bitola · Faculty of Technical Sciences Bitola

PhD

About

59
Publications
9,376
Reads
How we measure 'reads'
A 'read' is counted each time someone views a publication summary (such as the title, abstract, and list of authors), clicks on a figure, or views or downloads the full-text. Learn more
40
Citations
Additional affiliations
June 2018 - present
June 2013 - June 2018
University "St. Kliment Ohridski" - Bitola
Position
  • Professor (Associate)
June 2008 - June 2013
University "St. Kliment Ohridski" - Bitola
Position
  • Professor (Assistant)

Publications

Publications (59)
Article
Full-text available
In this paper an approach of real time measurement and analysis of magnetic field disturbances is described. Multiresolution based analysis appears as useful tool for feature extraction and identification of different signal phenomena. As a sensing element, a magentoresistive sensor is utilized together with virtual instrumentation. Experimental re...
Article
Full-text available
The paper gives an analysis of automated guided vehicles navigation systems. Two types of guidance are analysed and simulated by using LEGO Mindstorm NXT package – fixed-path guidance through line following as the oldest type and vision-based guidance as a low-cost guidance system that utilizes mobile phone camera for tracking marks and detecting o...
Article
Full-text available
In this paper we present the advantage of non-uniform over uniform threshold wavelet shrinkage denoising method, applied on noisy signals with signal dependent noise. We illustrate our results by comparing the noise energy after using the both filtration methods on the same set of artificially noise contaminated images. The experiments are made wit...
Conference Paper
Full-text available
In this paper an application of adaptive sampling algorithm is considered. Adaptive sampling is utilized as an efficient tool for records reduction of SCADA Datalog Database. Decreasing the number of records of Datalog File significantly increase the data-logging period. Proposed procedure is tested on commercially installed data-logging system and...
Conference Paper
Full-text available
This paper presents a method for signal denoising in the case of varying noise proportional to the local signal intensity. The signals are processed in a wavelet domain with a non-uniform threshold adjusted to the noise level.
Conference Paper
Full-text available
The paper presents a Hybrid Electrical Vehicle modeling based on multi-physics approach. The integrated simulation model of a series-parallel HEV is developed including the electric drive system, PI controller, vehicle load model, and gear box. The system step response is simulated and analyzed. The simulation results confirm the validity of the mo...
Conference Paper
Full-text available
Electrical power and energy balance is equality between total electrical power and energy input and output in/from power system. Electrical energy input in power system is consisted of several categories: generation of thermal, hydro and wind power plants, generation of power plants on distribution network (dispersed generation (DG)) and input of e...
Conference Paper
Full-text available
This paper considers open source software for designing mechatronic systems. Thereby, emphasis is placed on Scilab/Xcos. Through simple example the paper presents the way how to formalise dynamic system with known physical and mathematical model in this environment.
Conference Paper
Full-text available
The goal of this paper is to investigate the correlation between power system load and air temperature in Republic of Macedonia. Power system load hourly values (8760 per year) for year 2014 and 2015 are used as input data. In absence of hourly values for air temperature in Republic of Macedonia, data from internet is used for air temperature in ci...
Conference Paper
Full-text available
The Wavelet transform is described as a toolfor creating an image pseudo-hash in order to enable content-based image retrieval. The images aredecomposed in few levels and the most important wavelet detail coefficients are selected to compose the pseudo-hash. When searching for digital images in large database, the pseudo-hashes of the images are co...
Article
Full-text available
In this paper we consider a specific way of organizing 1D signals or 2D image databases, such that a more efficient storage and faster querying is achieved. A multiresolution technique of data processing is used in order of saving the most significant processed data.
Conference Paper
Full-text available
Dispersed generation (DG) brings together wide range of technologies for electricity production. Most frequently used electric machines for electricity production at each technology that belongs to DG group are: synchronous generators, asynchronous generators (AG) and power converters. This paper deals with transient stability of AG. Typical model...
Conference Paper
Full-text available
The Short-Time Fourier transform is described as a detection method of power supply disturbances. Compared to the standard Fourier transform, it provides time data in frequency domain. The most common power disturbances are listed and three simulation models are analyzed to describe the advantages and drawbacks of the Short-Time Fourier transform.
Article
In this paper an approach of real time measurement and analysis of magnetic field disturbances is described. Multiresolution based analysis appears as useful tool for feature extraction and identification of different signal phenomena. As a sensing element, a magentoresistive sensor is utilized together with virtual instrumentation. Experimental re...
Conference Paper
Full-text available
The paper discusses the wavelet transform as a tool for detection of power disturbances. On contrary of Fourier transform, wavelet transform provides both time and frequency localization. Compared to short-time Fourier transform, wavelets provide better frequency localization.
Data
The process of maintaining ICT infrastructure in digital companies is very important for the quality and safety of e-business processes. E-maintenance as a synonym for the maintenance of ICT infrastructure includes a wide range of activities that must be considered from several aspects.This paper provides an overview of each item, which is an impor...
Conference Paper
The process of maintaining ICT infrastructure in digital companies is very important for the quality and safety of e-business processes. E-maintenance as a synonym for the maintenance of ICT infrastructure includes a wide range of activities that must be considered from several aspects.This paper provides an overview of each item, which is an impor...
Article
Full-text available
In this paper we propose a procedure for discrete wavelet transform coefficients estimation on the images with missing samples. This procedure extends application of multiresolutionally based algorithm for non-uniformly sampled signal reconstruction. Estimation efficiency is tested on proprietary chosen images and results are presented graphically.
Conference Paper
One of the most challenging and important problems in robot programming is cognitive robot mapping. To solve this problem it is necessary to define and detect problem space. This to find a complete set of possible states, generated by exploring all the possible states or moves, which may, or not lead from a given start state to supposed goal state....
Conference Paper
One of the most challenging and important problems in robot programming is cognitive robot mapping. To solve this problem it is necessary to define and detect problem space. This means to find a complete set of possible states, generated by exploring all the possible states or moves, which may, or not lead from a given start state to supposed goal...
Article
Full-text available
In this paper a low cost virtual test bed for educational purposes is developed and tested. Hardware-In-the-Loop (HIL) design concept as an experimental part of power electronics is considered. Software simulations lead to develop control of the studied system. Before a real-time implementation of the control, HIL simulations could be a very useful...
Article
Full-text available
In this paper an algorithm for fast querying in images database is presented. It uses multiresolution technique, all the images are decomposed in few levels and the most important wavelet detail coefficients are selected to compose images pseudo-hashes. When searching for an image-query in the database, the pseudo-hashes of the images are considere...
Conference Paper
Full-text available
In this paper a low cost virtual test bed for educa-tional purposes is developed and tested. Hardware-In-the-Loop (HIL) design concept as an experimental part of power electronics is considered. Software simulations lead to develop control of the studied system. Before a real-time implementation of the control, HIL simulations could be a very usefu...
Conference Paper
Full-text available
In this paper we present ways of organizing images into a database in order to be able to make fast querying. Multiresolution technique is used and the most important wavelet detail coefficients are selected to compose signal/images pseudo-hashes. When searching for an image-query in the database, the pseudo-hashes of the images are considered inst...
Conference Paper
Full-text available
In this paper we present an algorithm for fast querying in database with images. It uses multiresolution technique, all the images are decomposed in few levels and the most important wavelet coefficients are used for calculation of pseudohash. When searching for an image-query in the database, pseudohash is calculated from the image-query and used...
Conference Paper
Full-text available
In this paper we present a technique for thresholding wavelet coefficients based on multiscale product analysis. We apply experiments both on noise-free and noisy signals; noise is signal dependent. Experimental results show that the noise is effectively removed.
Conference Paper
Full-text available
In this paper we compare the near-perfect reconstruction (NPR) quadrature mirror filter (QMF) design, versus perfect reconstruction (PR)-QMF design in terms of the flexibility of the design and the design filter characteristics, and their efficiency for denoising of the signals based on wavelet shrinkage technique.
Conference Paper
Full-text available
In this paper we present the advantage of non-uniform over uniform threshold wavelet shrinkage denoising method, applied on noisy signals with signal dependent noise. We illustrate our results by comparing the noise energy after using the both filtration methods on the same set of artificially noise contaminated images. The experiments are made wit...
Conference Paper
Full-text available
This paper presents experimental results obtained using a method that we propose for signal denoising. Noisy signals are processed in a discrete wavelet transform domain with a non-uniform threshold adjusted to the noise level.
Conference Paper
Full-text available
This paper presents our proposal on signal estimation in a case of varying noise proportional to the local intensity of the signal. Noisy signals are processed in a discrete wavelet transform domain with linear phase NPR-QMF filters with a cut-off frequency adjusted to the signal frequency spectrum. The algorithm based on multiscale products tends...
Conference Paper
Full-text available
In this paper we present an application and its interface for pre-processing of chest region NM images. It works in the Matlab environment and uses Matlab user defined functions. The software processes the images in the discrete wavelet transform domain. It uses QMF filters with linear phase designed to achieve both good image decomposition and nea...
Conference Paper
Full-text available
In this paper we present our approach on pre-processing of the heart region dynamic NM images. The presented method combines a discrete wavelet transform (DWT) realized via linear phase near perfect reconstruction (NPR) quadrature mirror filters (QMF) with a specific strategy for selecting the filters cut-off frequency and the filtering into the ti...
Conference Paper
Full-text available
In the paper we present our approach on pre-processing of NM heart-region images. The proposed method combines Discrete Wavelet Transform realized via near perfect reconstruction QMF bank with a specific strategy for selecting an appropriate threshold. The performance of the proposed method is demonstrated on real NM images.
Conference Paper
Full-text available
In the paper we present our approach on pre-processing of chest-region NM images. The proposed method combines Discrete Wavelet Transform realized via QMF filters with a specific strategy for selecting an appropriate threshold. Due to the signal-dependence of the Poisson noise, the Anscombe variance-stabilizing transformation is applied. As a resul...
Conference Paper
Full-text available
This paper presents some experimental results of NM image filtering by using several types of wavelets. For this purpose, one set of wavelets that belong to different wavelet families is chosen. Based on these experiments, the Coiflet 5 wavelet is identified as the best wavelet. The aim of the paper is identifying the best wavelet for filtering of...
Conference Paper
Full-text available
In this paper we present our approach on pre-processing chest region dynamical NM images which enables anatomical data extraction of the vena cava and the heart. The aim of the method is developing sophisticated diagnostic software that could automatically offer the optimal positions and the shapes of the regions of interest needed for the heart st...
Conference Paper
Full-text available
In this paper we present our approach on pre-processing chest region dynamical NM images which enables anatomical data extraction of the vena cava and the heart. The aim of the method is developing sophisticated diagnostic software that could automatically offer the optimal positions and the shapes of the regions of interest needed for the heart st...
Conference Paper
Full-text available
In this paper an approach on pre-processing of dynamic nuclear medicine images is proposed. The method is based on conventional threshold level detection, applied on adequately pre-processed set of dynamic images. A combination of autocorrelation low-pass filtering technique and wavelet transform-domain processing is used. The performance of the pr...
Conference Paper
Full-text available
In this paper an approach for pre-processing of dynamic nuclear images is proposed. The method is based on conventional threshold level detection, applied on adequately pre-processed set of dynamic images. A few raw and a few pre-processed nuclear medicine images are provided.
Conference Paper
Full-text available
This paper considers the constraints imposed on the values of the design parameters in the efficient method proposed by Hedge and Shenoi. We have examined the relationships between the filter parameters in order to make correct specifications of the permissible values of the design parameters.
Conference Paper
Full-text available
This paper considers the practical implementation of the wavelet transform. An algorithm for estimation of the signal disturbances is proposed. The results are compared with the results obtained by classical filtering.
Conference Paper
Full-text available
The iterative algorithme of Chen and Lee for the design of quasiequiripple QMF banks in the frequency domain is modified; the weighting function incorporated in the objective function is updated at the end of each iteration. As a result, the number of iteration and the number of floating point operations are reduced significantly.
Conference Paper
Full-text available
A modification of the Chen and Lee's algorithm for the design of quadrature mirror filter banks in the frequency domain is presented. It concerns updating the frequency response weighting function and results in significantly reduced number of iterations.

Network

Cited By

Projects

Projects (2)
Project
NM images
Project
PARENCHIMA - Magnetic Resonance Imaging Biomarkers for Chronic Kidney Disease (COST action CA16103) The overall aim of PARENCHIMA is to eliminate the main barriers to the broader study, commercial exploitation and clinical use of renal MRI biomarkers. PARENCHIMA will coordinate the research of leading European groups in this area to: (1) improve the reproducibility and standardisation of renal MRI biomarkers; (2) increase their availability by developing an open-access toolbox with software and data; (3) demonstrate biological validity and clinical utility in a prospective multicentre clinical study. For additional information please visit the project homepage at www.renalmri.org or follow us on the social media Linkedin: renalMRI.org - PARENCHIMA Research Community (https://www.linkedin.com/feed/#/groups/8448307) Facebook: #renalMRI (https://www.facebook.com/renalmri/) Twitter: @renalMRI (https://twitter.com/renalmri) This project is a COST Action (CA16103), supported by COST (European Cooperation in Science and Technology). www.cost.eu.