Home | Ukrainian |
UDC 537.81:620.1:621.6
R. M. Dzhala, B. Ya. Verbenets, B. I. Horon, O. I. Senyuk
ANTIINTERFERENCE DETERMINATION
OF UNDERGROUND PIPELINE PLACEMENT
Karpenko Physico-Mechanical Institute of the NAS of Ukraine
E-mail: dzhala@ipm.lviv.ua
The protection against interference with methods and means of determining the location of underground
cables and pipelines and the main stages of the development of induction track finders has been analyzed.
Theoretically, the informative features of the distribution of the magnetic field differential current
conductors for the choice of the difference magnetosensor of small-scale track finders protected from
external field are investigated.
It is found that for the difference magnetosensors orthogonal to the base, the profiling function
has a maximum signal over the current conductor. For the difference magnetosensors parallel to the base
there is a clear minimum of the signal over the current conductor between the two maxima located at
a distance equal to the depth of the current conductor. Such differential magnet sensors are suitable
for determining the location and depth of the underground pipeline, with the removal of the influence
of an external almost homogeneous obstacle-causing magnetic field.
Keywords: magnetic field, current conductor, profiling functions, difference signs,
pipeline, obstacles, trace detector.
Ð. Ì. Äæàëà, Á. ß. Âåðáåíåöü, Á. ². Ãîðîí, Î.². Ñåíþê
ÇÀÂÀÄÎÑÒ²ÉÊÅ ÂÈÇÍÀ×ÅÍÍß ÐÎÇ̲ÙÅÍÍß
ϲÄÇÅÌÍÎÃÎ ÒÐÓÁÎÏÐÎÂÎÄÓ
Ô³çèêî-ìåõàí³÷íèé ³íñòèòóò ³ì. Ã. Â. Êàðïåíêà ÍÀÍ Óêðà¿íè, Ëüâ³â
Ïðîàíàë³çîâàíî çàõèùåí³ñòü â³ä çàâàä ìåòîä³â ³ çàñîá³â âèçíà÷åííÿ ðîçì³ùåííÿ ï³äçåìíèõ êàáåë³â ³ òðóáîïðîâîä³â òà ðîçâèòîê
³íäóêö³éíèõ òðàñîøóêà÷³â. Òåîðåòè÷íî äîñë³äæåíî ³íôîðìàòèâí³ îçíàêè ðîçïîä³ëó äèôåðåíö³àë³â ìàãí³òíîãî ïîëÿ ñòðóìîïðîâîäó
äëÿ âèáîðó ð³çíèöåâèõ ìàãí³òîñïðèéìà÷³â ìàëîãàáàðèòíèõ çàâàäîçàõèùåíèõ òðàñîøóêà÷³â.
Âèÿâëåíî, ùî äëÿ ð³çíèöåâèõ îðòîãîíàëüíèõ áàç³ ìàãí³òîñïðèéìà÷³â ôóíêö³ÿ ïðîô³ëþâàííÿ ìàº
ìàêñèìóì ñèãíàëó íàä ñòðóìîïðîâîäîì. Äëÿ ð³çíèöåâèõ ïàðàëåëüíèõ áàç³ ìàãí³òîñïðèéìà÷³â º ÷³òêèé ì³í³ìóì ñèãíàëó
íàä ñòðóìîïðîâîäîì ì³æ äâîìà ìàêñèìóìàìè, â³ääàëåíèìè îäèí â³ä îäíîãî íà â³äñòàíü, ð³âíó ãëèáèí³ çàëÿãàííÿ ñòðóìîïðîâîäó.
Òàê³ ð³çíèöåâ³ ìàãí³òîñïðèéìà÷³ ïðèäàòí³ äëÿ âèçíà÷åííÿ ÿê ì³ñöÿ, òàê ³ ãëèáèíè çàëÿãàííÿ ï³äçåìíîãî òðóáîïðîâîäó ç âèëó÷åííÿì
âïëèâó ñòîðîííüîãî ìàéæå îäíîð³äíîãî çàâàäîíåñó÷îãî ìàãí³òíîãî ïîëÿ.
Këþ÷îâ³ ñëîâà: ìàãí³òíå ïîëå, ñòóìîïðîâ³ä, ôóíêö³¿ ïðîô³ëþâàííÿ, ð³çíèöåâ³ îçíàêè, òðóáîïðîâ³ä, çàâàäè, òðàñîøóêà÷.
UDC 621.391:519.22
². Ì. Javorskyj, O. Y. Dzeryn, R. Ì. Yuzefovych
COVARIANCE LSM-ANALYSIS
OF BIPERIODIC NONSTATIONARY VIBRATION SIGNALS
H. V. Karpenko Physico-Mechanical Institute of the NAS of Ukraine, Lviv;
Telecommunication Institute of University of Technology and Life Science, Bydgoszcz, Poland;
Lviv Polytechnic National University
E-mail: dzeryn_oksana@ukr.net
The estimators of covariance function of biperiodically correlated random processes – mathematical models
of vibration signals with binary stochastic recurrence, obtained with using the least squares method (LSM),
are analyzed. It was shown that these estimators are unbiased and consistent under the condition of correlation
relationships decaying with the bias rise. The main LSM-estimator advantage over the component estimator is the
absence of leakage effects, which can cause significant errors of covariance characteristics estimation when
combination frequencies have close values. Formulae obtained in this paper for statistic characteristics
of LSM-estimator give an opportunity to calculate processing errors for specific signal types and also
compare them with the errors of component estimation.
Key words: biperiodic nonstationary vibration signal, covariance function estimator,
least squares method, asymptotical unbiasedness and consistency, leakage.
². Ì. ßâîðñüêèé, Î. Þ. Äçåðèí, Ð. Ì. Þçåôîâè÷
ÊÎÐÅËßÖ²ÉÍÈÉ ÌÍÊ-ÀÍÀ˲Ç
Á²ÏÅвÎÄÈ×ÍÎ ÍÅÑÒÀÖ²ÎÍÀÐÍÈÕ Â²ÁÐÀÖ²ÉÍÈÕ ÑÈÃÍÀ˲Â
Ô³çèêî-ìåõàí³÷íèé ³íñòèòóò ³ì. Ã. Â. Êàðïåíêà ÍÀÍ Óêðà¿íè, Ëüâ³â;
²íñòèòóò òåëåêîìóí³êàö³¿ Òåõíîëîã³÷íî-ïðèðîäíè÷îãî óí³âåðñèòåòó, Áèäãîù, Ïîëüùà;
Íàö³îíàëüíèé óí³âåðñèòåò “Ëüâ³âñüêà ïîë³òåõí³êà”
Ïðîàíàë³çîâàíî îö³íêè êîðåëÿö³éíî¿ ôóíêö³¿ á³ïåð³îäè÷íî êîðåëüîâàíèõ âèïàäêîâèõ ïðîöåñ³â – ìàòåìàòè÷íèõ ìîäåëåé
â³áðàö³éíèõ ñèãíàë³â ç ïîäâ³éíîþ ñòîõàñòè÷íîþ ïîâòîðþâàí³ñòþ, ÿê³ çíàõîäÿòü ìåòîäîì íàéìåíøèõ êâàäðàò³â.
Äîâåäåíî, ùî òàê³ îö³íêè º àñèìïòèîòè÷íî íåçì³ùåíèìè é ñëóøíèìè çà óìîâè çàíèêàííÿ êîðåëÿö³éíèõ çâ’ÿçê³â
ç ðîñòîì çñóâó. Ïîêàçàíî, ùî öèì ìåòîäîì ìîæíà óíèêíóòè ñèñòåìàòè÷íèõ ïîõèáîê îö³íþâàííÿ, ïîâ’ÿçàíèõ ç åôåêòîì ïðîñî÷óâàííÿ.
Êëþ÷îâ³ ñëîâà: á³ïåð³îäè÷íî êîðåëüîâàíî âèïàäêîâ³ ïðîöåñè, îö³íêà êîðåëÿö³éíî¿ ôóíêö³¿,
ìåòîä íàéìåíøèõ êâàäðàò³â, àñèìïòîòè÷íà íåçì³ùåí³ñòü ³ ñëóøí³ñòü, ïðîñî÷óâàííÿ.
UDC 004.932
T. S. Mandziy, ². B. Ivasenko
SEGMENTATION OF CORROSION DAMAGE IMAGES WITH UNKNOWN BACKGROUND BY ENERGY MINIMIZATION
Karpenko Physico-Mechanical Institute of the NAS of Ukraine, Lviv
E-mail: ivasenko@ipm.lviv.ua
The problem of rusted areas segmentation on painted constructions is considered.
The rust percentage on the coatings can be computed using digital image processing methods.
Segmentation method by energy minimization using graph cut is applied. The results demonstrate that
the proposed approach can be effectively used for rust segmentation.
Keywords: image processing, segmentation, rust, graph cuts.
Ò. Ñ. Ìàíäç³é, ². Á. ²âàñåíêî
ÑÅÃÌÅÍÒÀÖ²ß ÇÎÁÐÀÆÅÍÜ ÊÎÐÎDzÉÍÈÕ ÏÎØÊÎÄÆÅÍÜ
ÍÀ ÍŲÄÎÌÎÌÓ ÔÎͲ ÌÅÒÎÄÎÌ Ì²Í²Ì²ÇÀÖ²¯ ÅÍÅÐò¯
Ô³çèêî-ìåõàí³÷íèé ³íñòèòóò ³ì. Ã. Â. Êàðïåíêà ÍÀÍ Óêðà¿íè, Ëüâ³â
Ðîçãëÿíóòî çàäà÷ó ñåãìåíòàö³¿ êîðîäîâàíèõ ä³ëÿíîê ïîâåðõîíü íà ôàðáîâàíèõ êîíñòðóêö³ÿõ. ³äñîòîê ³ðæàâ³ííÿ
íà ïîêðèòòÿõ îá÷èñëåíî ìåòîäàìè îáðîáêè öèôðîâèõ çîáðàæåíü. Çàñòîñîâàíî ìåòîä ñåãìåíòàö³¿ çà äîïîìîãîþ ì³í³ì³çàö³¿
åíåð㳿 íà îñíîâ³ ðîçð³çó ãðàôà. Âèÿâëåíî, ùî ïðîïîíîâàíèé ï³äõ³ä åôåêòèâíèé äëÿ ñåãìåíòàö³¿ ³ðæ³.
Êëþ÷îâ³ ñëîâà: îáðîáêà çîáðàæåíü, ñåãìåíòàö³ÿ, ðîçð³ç ãðàô³â.
UDC 551.568.85
A. S. Rubel, V. V. Lukin
ASSESMENT OF VISUAL QUALITY OF DENOISED IMAGES
National Aerospace University, Kharkiv
E-mail: rubel.andrew@gmail.com, lukin@ai.kharkov.com
The problem of image denoising is considered from the viewpoint of visual quality of filtered images.
Special experiments with a large number of observers have been carried out to determine probability that
a denoised image is preferable compared to the corresponding original noisy one. It has been found that there
are many practical cases when observers prefer noisy images. This usually happens if an image is highly textural,
noise has either quite low or too high intensity, and a used filter performs not efficiently. It has been also
shown that modern metrics, even those that take into account peculiarities of human vision system, often perform
not adequately. The cases when it is really worth to carry out image denoising are considered.
Keywords: image denoising, visual quality, experimental assessment.
A. Ñ. Ðóáåëü, Â. Â. Ëóê³í
ÎÖ²ÍÊÀ ²ÇÓÀËÜÍί ßÊÎÑÒ² ÇÎÁÐÀÆÅÍÜ Ï²ÑËß Ô²ËÜÒÐÀÖ²¯
Íàö³îíàëüíèé àåðîêîñì³÷íèé óí³âåðñèòåò, Õàðê³â
Ô³ëüòðàö³þ çîáðàæåíü ðîçãëÿíóòî ÿê â³çóàëüíó ÿê³ñòü îáðîáëåíèõ çîáðàæåíü. Âèêîíàíî åêñïåðèìåíòè
ç³ çàëó÷åííÿì âåëèêî¿ ê³ëüêîñò³ âîëîíòåð³â, ùîá âèçíà÷èòè éìîâ³ðí³ñòü òîãî, ùî îáðîáëåíå çîáðàæåííÿ
êðàùå çà â³äïîâ³äíå ïåðâèííå. Âèÿâëåíî, ùî º áàãàòî ñèòóàö³é, êîëè ëþäèíà â³ääຠïåðåâàãó ïåðâèííîìó çîáðàæåííþ,
ùî ñïîòâîðåíå øóìîì. Çîêðåìà, ÿêùî çîáðàæåííÿ äóæå òåêñòóðíå, øóì àáî äóæå ñëàáêèé, àáî çàíàäòî ³íòåíñèâíèé,
à ô³ëüòð ïðàöþº íååôåêòèâíî. Òàêîæ ïîêàçàíî, ùî ñó÷àñí³ ì³ðè, íàâ³òü â³çóàëüíî¿ ÿêîñò³, ÷àñòî ïðàöþþòü íåàäåêâàòíî.
Ðîçãëÿíóòî âèïàäêè çàñòîñîâíîñò³ ô³ëüòðàö³¿.
Këþ÷îâ³ ñëîâà: ô³ëüòðàö³ÿ çîáðàæåíü, â³çóàëüíà ÿê³ñòü, åêñïåðèìåíòàëüíå îö³íþâàííÿ.
UDC 551.568.85
I. V. Stasyshyn, Y. M. Kotsiuba, L. I. Muravsky, T. I. Voronyak
RETRIEVING THE SURFACE RELIEF COMPONENTS USING PHASE-SHIFTING INTERFEROMETRY AND GAUSSIAN FILTER
Karpenko Physico-Mechanical Institute of the NAS of Ukraine, Lviv
E-mail: ihorgo@hotmail.com
It is known that for the determination of mechanical, corrosive and tribological parameters,
such terms as “roughness” and “waviness” are often used. Filtering in the frequency domain
is used to extract these components from the total relief. An approach to determining the optimum value of the
cut-off frequency for 2D Gaussian filter is proposed to obtain the surface relief by three-step phase-shifting
interferometry with an arbitrary phase shift of the reference beam.
Keywords: 2D filtering; interferometry; surface relief; roughness; waviness.
². Â. Ñòàñèøèí, Þ. Ì. Êîöþáà, Ë. ². Ìóðàâñüêèé, Ò. ². Âîðîíÿê
²ÄÍÎÂËÅÍÍß ÊÎÌÏÎÍÅÍҲ ÐÅËܪÔÓ ÏÎÂÅÐÕͲ
ÇÀ ÄÎÏÎÌÎÃÎÞ ÔÀÇÎÇÑÓÂÍί ²ÍÒÅÐÔÅÐÎÌÅÒв¯
ÒÀ ÃÀÓÑѲÂÑÜÊÎÃÎ Ô²ËÜÒÐÀ
Ô³çèêî-ìåõàí³÷íèé ³íñòèòóò ³ì. Ã. Â. Êàðïåíêà ÍÀÍ Óêðà¿íè, Ëüâ³â
Äëÿ â³äíîâëåííÿ êîìïîíåíò³â øîðñòêîñò³ òà õâèëÿñòîñò³ ç³ çàãàëüíîãî ðåëüºôó âèêîðèñòîâóþòü ô³ëüòðàö³þ â ÷àñòîòí³é îáëàñò³.
Çàïðîïîíîâàíî ï³äõ³ä äëÿ âèçíà÷åííÿ îïòèìàëüíîãî çíà÷åííÿ ÷àñòîòè â³äñ³÷êè äëÿ 2D-ãàóññ³âñüêîãî ô³ëüòðà, ùîá îòðèìàòè ðåëüºô
ïîâåðõí³ çà äîïîìîãîþ òðèêðîêîâî¿ ôàçîçñóâíî¿ ³íòåðôåðîìåò𳿠ç äîâ³ëüíèì çñóâîì ôàçè îïîðíîãî ïó÷êà.
Ðîçðàõîâàí³ òàê êîìïîíåíòè ïîâåðõí³ ïðèäàòí³ äëÿ âèçíà÷åííÿ ìåõàí³÷íèõ, êîðîç³éíèõ òà òðèáîëîã³÷íèõ ïàðàìåòð³â ìàòåð³àëó.
Êëþ÷îâ³ ñëîâà: 2D ô³ëüòðàö³ÿ; ³íòðåôåðîìåòð³ÿ; ðåëüºô ïîâåðõí³; øîðñòê³ñòü; õâèëÿñò³ñòü.
UDC 004.832.2 : 004.853
D. H. Dosyn
PERTINENCE EVALUATION SYSTEM ARCHITECTURE ON A BASIS
OF LEARNING ONTOLOGY WITH PLANNING IN A CERTAIN DOMAIN
Lviv Polytechnic National University
E-mail: dmytro.h.dosyn@lpnu.ua
The method of text document pertinence estimation is proposed. It is based on agent approach, expected
value of perfect information analysis, hierarchical task network structure of a knowledge base and
automated planning algorithms. Use of Markov decision process approach allows us to estimate expected
utility of the strategy built in the framework of agent knowledge base with the aim to evaluate gain
of expected utility caused by account of information extracted from the text document. For this purposes
the text document is considered as a message with a two-part structure which should help us to supplement
information contained in this document by relevant context information from the knowledge base.
Keywords: pertinence evaluation, ontology learning, automated planning, hierarchical task network,
expected value of perfect information.
Ä. Ã. Äîñèí
ÀÐÕ²ÒÅÊÒÓÐÀ ÑÈÑÒÅÌÈ ÎÖ²ÍÞÂÀÍÍß ÏÅÐÒÈÍÅÍÒÍÎÑÒ²,
ÙÎ ÁÀÇÓªÒÜÑß ÍÀ ÍÀÂ×ÀÍͲ ÎÍÒÎËÎò¯ ÏËÀÍÓÂÀÍÍß
Ó ÂÈÁÐÀÍ²É ÏÐÅÄÌÅÒÍ²É ÎÁËÀÑÒ²
Íàö³îíàëüíèé óí³âåðñèòåò “Ëüâ³âñüêà ïîë³òåõí³êà”
Çàïðîïîíîâàíî ìåòîä îö³íþâàííÿ ïåðòèíåíòíîñò³ òåêñòîâèõ äîêóìåíò³â, ÿêèé áàçóºòüñÿ íà àãåíòíîìó
ï³äõîä³, àíàë³ç³ î÷³êóâàíî¿ âåëè÷èíè äîñêîíàëî¿ ³íôîðìàö³¿, ñòðóêòóð³ áàçè çíàíü ó ôîðìàò³ ³ºðàðõ³÷íî¿
ìåðåæ³ çàäà÷ òà àëãîðèòì³â àâòîìàòè÷íîãî ïëàíóâàííÿ. Ìàðê³âñüêà ìîäåëü ïðèéíÿòòÿ ð³øåíü äຠçìîãó
îá÷èñëþâàòè î÷³êóâàíó êîðèñí³ñòü ñòðàòå㳿, ïîáóäîâàíî¿ çàñîáàìè áàçè çíàíü àãåíòà, äëÿ îö³íêè ïðèðîñòó êîðèñíîñò³,
çóìîâëåíîãî âðàõóâàííÿì ³íôîðìàö³¿, îòðèìàíî¿ ç òåêñòîâîãî äîêóìåíòà. Äëÿ öüîãî òåêñòîâèé äîêóìåíò ðîçãëÿíóòî ÿê ïîâ³äîìëåííÿ
ç äâîõåëåìåíòíîþ ñòðóêòóðîþ, ùî äîïîìàãຠäîïîâíþâàòè ³íôîðìàö³þ, ÿêà ì³ñòèòüñÿ â öüîìó äîêóìåíò³,
â³äïîâ³äíîþ êîíòåêñòíîþ ³íôîðìàö³ºþ ç áàçè çíàíü.
Këþ÷îâ³ ñëîâà: îö³íþâàííÿ ïåðòèíåíòíîñò³, íàâ÷àííÿ îíòîëî㳿, àâòîìàòè÷íå ïëàíóâàííÿ,
³ºðàðõ³÷íà ìåðåæà çàäà÷, î÷³êóâàíà âåëè÷èíà äîñêîíàëî¿ ³íôîðìàö³¿.
UDC 004.93.14
A. L. Yerokhin, O. V. Zolotukhin
FUZZY PROBABILISTIC NEURAL NETWORK
IN DOCUMENT CLASSIFICATION TASKS
Kharkiv National University of Radio Electronics
E-mail: andriy.yerokhin@nure.ua, oleg.zolotukhin@nure.ua
The modification of a probabilistic neural network based on adding fuzziness is investigated.
This article discusses the architecture and learning algorithm of fuzzy probabilistic
neural network that can classity on-line text documents.
Keywords: fuzzy neural network, text document classification.
À. Ë. ªðîõ³í, Î. Â. Çîëîòóõ³í
ÍÅײÒÊÀ ²ÌβÐͲÑÍÀ ÍÅÉÐÎÍÍÀ ÌÅÐÅÆÀ
 ÇÀÄÀ×ÀÕ ÊËÀÑÈÔ²ÊÀÖ²¯ ÄÎÊÓÌÅÍÒ²Â
Õàðê³âñüêèé íàö³îíàëüíèé óí³âåðñèòåò ðàä³îåëåêòðîí³êè
Äîñë³äæåíî ìîäèô³êàö³þ éìîâ³ðí³ñíî¿ íåéðîííî¿ ìåðåæ³ íà îñíîâ³ ââåäåííÿ íå÷³òêîñò³.
Ðîçãëÿíóòî àðõ³òåêòóðó òà àëãîðèòì íàâ÷àííÿ íå÷³òêî¿ éìîâ³ðí³ñíî¿ íåéðîííî¿ ìåðåæ³,
ÿêà ìîæå êëàñèô³êóâàòè òåêñòîâ³ äîêóìåíòè ó ðåæèì³ ðåàëüíîãî ÷àñó.
Êëþ÷îâ³ ñëîâà: íå÷³òêà íåéðîííà ìåðåæà, òåêñòîâèé äîêóìåíò, êëàñèô³êàö³ÿ.