Przemysław Mazurek
West Pomeranian University of Technology
Network
Latest external collaboration on country level. Dive into details by clicking on the dots.
Publication
Featured researches published by Przemysław Mazurek.
IP&C | 2011
Przemysław Mazurek
In the paper Track–Before–Detect (TBD) algorithm for tracking a low–signal objects with amplitude modulated signal is proposed. Direct application of TBD algorithms is not sufficient for such case due to accumulative approach. This signal has a zero mean value and cannot be processed directly. The proposed algorithm is based on applications of two different TBD algorithms. The first is the directional IIR filter that works as a velocity filter as a part of the noncoherent demodulator. The second TBD algorithm is the recurrent Spatio–Temporal TBD that support trajectory switching using Markov’s matrix. Numerical experiments (Monte Carlo tests) for point target are used for verification of the proposed solution.
International Journal of Applied Mathematics and Computer Science | 2014
Przemysław Mazurek; Dorota Oszutowsk A.-M. Azurek
Abstract The Slit Island Method (SIM) is a technique for the estimation of the fractal dimension of an object by determining the area- perimeter relations for successive slits. The SIM could be applied for image analysis of irregular grayscale objects and their classification using the fractal dimension. It is known that this technique is not functional in some cases. It is emphasized in this paper that for specific objects a negative or an infinite fractal dimension could be obtained. The transformation of the input image data from unipolar to bipolar gives a possibility of reformulated image analysis using the Ising model context. The polynomial approximation of the obtained area-perimeter curve allows object classification. The proposed technique is applied to the images of cervical cell nuclei (Papanicolaou smears) for the preclassification of the correct and atypical cells.
IP&C | 2013
Dorota Oszutowska–Mazurek; Przemysław Mazurek; Kinga Sycz; Grażyna Waker–Wójciuk
The aim of cervical screening is detection of potentially pre–cancerous and cancerous changes in the endocervical canal of female reproductive system. There are many Fractal Dimension (FD) that are important for the computer assisted analysis of the texture of cell nuclei. The best method is the algorithm that process all pixels of the cell nuclei, not limited by the constraints. Variogram algorithm is proposed for cell nuclei classification (correct and atypical cell classes). Obtained results shows that variogram and cell nuclei size are important for the automatic classification of cell nuclei.
IP&C | 2011
Przemysław Mazurek
Track–Before–Detect (TBD) algorithms are used in object tracking applications. TBD algorithms are used for tracking of a low–SNR objects. Selection of the measurement space is the one of the most important factors that significantly influences on the tracking results. In this paper a few measurement spaces are compared using Monte Carlo numerical tests. The best obtained results are for the angle–only measurement space that is signal shape oriented.
international conference on methods and models in automation and robotics | 2014
Przemysław Mazurek
Line following robots requires image acquisition and processing algorithms for the determination of the line trajectory. The Viterbi algorithm is proposed for the estimation of the line trajectory in this paper. The robustness of this algorithm is verified using the Monte Carlo approach for two distortions types: additive Gaussian noise and false lines sets. The results show possibilities of reliable estimation, even if the real line is hidden in the noise, without the application of additional estimators based on the previous measurements.
IP&C | 2014
Przemysław Mazurek
The detection of local regular patterns and dependent values in heavy noised signal is proposed in this paper. The moving window approach allows computation of the maximal autocovariance of signal. The differences are emphasized using Spatio–Temporal Track–Before–Detect algorithm so tracking such objects is possible. The possibilities of this technique are shown in a few illustrative examples and discussed. The detection of weak signals hidden in the background noise is emphasized.
international conference on transport systems telematics | 2010
Krzysztof Okarma; Przemysław Mazurek
One of the main tasks of the statistical traffic analysis is its rating due to the size, type or number of axles. A typical method for measuring the volume of traffic along with the initial classification is based on data derived from inductive sensors and load cells. The possibilities of such a system are however limited, therefore in recent years a great interest in machine vision systems can be observed. An interesting image analysis technique that allows a rapid classification of the types of vehicles observed from the side view is the shape analysis. It can be applied for binary images, for which the values of shape descriptors such as e.g. Feret’s diameter can be calculated, as well as some additional quantities such as the center of gravity determined for greyscale images. The article presents the results of the shape analysis obtained for different types of vehicles observed from the camera placed beside the road.
international conference on transport systems telematics | 2011
Krzysztof Okarma; Przemysław Mazurek
The capabilities of the vehicle tracking systems based on data acquisition from a single camera with a linear processing characteristic are limited by the range of light radiation acquired at a given moment. Taking into account the fact that different vehicles present on a single video frame can be either underexposed, and overexposed (especially in the case of a highly reflective surface in the sunlight), the possibilities of distinguishing between them are limited. The paper presents a method of vehicle tracking with simultaneous acquisition of video frames with different exposure parameters. Aside from changing a single camera exposure it is also possible to use a double-camera variant, especially for distant objects, for which the shift on both images acquired simultaneously is negligible. The proposed method uses a modified Track-Before-Detect technique with additional data fusion in order to obtain the High Definition Range images used for the tracking of vehicles represented on the images as the objects with varying brightness and size.
international conference on methods and models in automation and robotics | 2009
R. Krupiński; Przemysław Mazurek
Abstract In the paper the optimization of median filters (single and two cascade configurations) for improving the electrooculogram and blinking signals separation using the synthetic model of measured biopotentials is considered. The obtained separate electrooculogram and blinking signals can be used for improving motion capture systems and reduce the number of markers around eyes. The achieved parameters are validated on the real electrooculography (EOG) data.
IP&C | 2015
Adam Piórkowski; Przemysław Mazurek; Jolanta Gronkowska-Serafin
Assessment of regularity based on quantitativemeasurements is important for biological structures analysis. The most important is the analysis of the cells’ structures. In this paper a few assessment methods are compared: hexagonality, coefficient of variation of sizes, and average coefficient of variation of cells’ sides lengths (the new proposed by the authors). This evaluation is based on synthetically generated image datasets for unbiased Monte Carlo analysis. The cell structures are time– evolved using optimization approach starting from irregular to regular.