Choice of recognition attributes when searching hidden objects

Authors

  • Александр Николаевич Логунов East Ukraine Volodymyr Dahl National University, Molodizhny bl., 20а, Lugansk, Ukraine, 91034, Ukraine
  • Галина Леонидовна Логунова Scientific Research and Project Designing Institute Iskra, Zveineka st,. 145-c, Lugansk, Ukraine, 91033, Ukraine

DOI:

https://doi.org/10.15587/1729-4061.2012.4622

Keywords:

control examination, y-radiation, spectrum, classification.

Abstract

During the inspection of transport facilities, various special services apply portable radioisotopic devices. Gamma radiation, dispersed backwards by the unit under the test, carries information about the internal structure of the object and thus, gives the possibility to detect hidden beetles. As the improvement of technical capa-bilities of the device for the account of the activity growth of the radiation source is impossible on security grounds, one of the basic ways of control effectiveness in-crease is the improvement of the algorithm of data processing. The hidden objects search method was suggested, based on the analysis of image, formed from energy spectral curve of registered gamma radiation, dispersed backwards. The problem of classification of spectral images was solved by the image discrimination method. The target of the work is the formation of the most informative system of classification at-tributes. The attributes are chosen according to the metric criterion of informativity – standardized Euclidian distance. The morphometric attributes, necessary for classifi-cation, were calculated for spectrograms, obtained experimentally. The spectrum dif-ferential for the obstruction with and without a beetle is estimated by the standard-ized Euclidian distance. The moment attributes of various procedures, which influ-ence most on the standardized Euclidian distance, were determined. The usage of the chosen moment attributes for image spectrum recognition helps to reveal hidden bee-tles, placed behind the steel 6 mm obstacles.

Author Biographies

Александр Николаевич Логунов, East Ukraine Volodymyr Dahl National University, Molodizhny bl., 20а, Lugansk, Ukraine, 91034

Assistant Chair of electronic industry equipment

Галина Леонидовна Логунова, Scientific Research and Project Designing Institute Iskra, Zveineka st,. 145-c, Lugansk, Ukraine, 91033

Senior Scientist Scientific Research laboratories of the specialized technologies

References

  1. Buster K910B Contraband Detector Specifications. [Электронный ресурс]. - Режим доступа: http://www.sasrad.com/products/buster /specifications.php
  2. Детектор изменения плотности (Детектор "контрабанды") ДИП-А01М [Электронный ресурс]. - Режим доступа: http://aspect.dubna.ru/new/page. php?page=381
  3. Polimaster Устройство поиска неоднородностей УПН-РМ1401М-П (Детектор контрабанды) [Электронный ресурс]. - Режим доступа: http://www.polimaster.ru/products/ contraband_detector/pm1401t/
  4. НИПКИ «Искра». Детектор скрытых пустот «РОСЬ 4М». [Электронный ресурс]. - Режим доступа: http://iskra.lugansk.ua/index.php?lang=ru&page =sci-tech-prod/contraband
  5. Логунова Г.Л. Аспекты безопасности радиоизотопных приборов в технологиях досмотрового контроля транспортных средств / Г.Л. Логунова // Науковий вісник Укр НДПІБ – № 2 (24 ), 2011. – С. 87-93.
  6. Логунова Г.Л. Моделирование параметров радиоизотопных приборов досмотрового контроля / Г.Л. Логунова // Искусственный интеллект: Научно-теоретический журнал. – Донецк. – 2011. – С. 299-302.
  7. Анисимов Б. В. Распознавание и цифровая обработка изображений: [Учеб. пособие для студентов вузов] / Б. В. Анисимов, В.Ф. Курганов, В.К. Зло-бин. – М.: Высш. шк., 1983. – 295 с.
  8. Методы компьютерной обработки изображений / Под ред. В.А. Сойфера. – 2-е изд., испр. – М.: Физматлит, 2003. – 784 с.

Published

2012-10-09

How to Cite

Логунов, А. Н., & Логунова, Г. Л. (2012). Choice of recognition attributes when searching hidden objects. Eastern-European Journal of Enterprise Technologies, 5(9(59), 21–25. https://doi.org/10.15587/1729-4061.2012.4622

Issue

Section

Information and controlling system