350 rub
Journal Science Intensive Technologies №1 for 2012 г.
Article in number:
WAY OF POLYMODELLING RECOGNITION OF OBJECTS UNDER THEIR IMAGES WITH APPLICATION OF NEURAL NETWORKS
Authors:
A. F. Ulasen, K. V. Kiselyov
Abstract:
Authors offer one decision of task recognition object on their image data in the article. Known what optimal decision for object recognition from image data is usage of neural network. However, dependence of external conditions, receive image data can be different quality, therefore it is necessary to use several neural networks. Because these conditions possess some indistinctness, that is also necessary to use elements of fuzzy logic. Main feature of this structure is presence of commutator (fuzzy classifier) wish allows to choose one of several network for recognition air object from image data at optical system. Training and selection internal parameters of neural classifier have made beforehand at the training selection, there are present image data of air objects with defined discrecity angle. After that probability and time of correct recognition are testing. There was tested possibility of identification classes air objects with prototype of model system recognition from mixed data images of objects and background noise which have different contrast. It is possible due to correct choose one neural network according to qualitative data and features of receiving images. Selection carried out with the fuzzy classifier (commutator). It operates on basis of intelligent database which before was created. Performance recognition evaluation will permit to increase probability and to decrease time for correct recognition of objects.
Pages: 31-35
References
  1. Головин С. А., Сизов Ю. Г. и др. Высокоточное оружие и борьба с ним. М.: ВПК. 1996.
  2. Сойфер В. А. Методы компьютерной обработки изображений. М.: Физматлит. 2003.
  3. Белозерский Л. А. Основы построения систем распознавания образов. Курс лекций. Донецк. ДНИИИ. 1997.
  4. Круглов В. В., Дли М. И., Голунов Р. Ю. Нечеткая логика и искусственные нейронные сети. М.: Горячая линия ?
    Телеком. 2001.
  5. Мелихов А. Н., Берштейн Л. С., Коровин С. Я. Ситуационные советующие системы с нечеткой логикой. М.: Наука. 1990.
  6. Борисов В. В., Круглов В. В., Федулов А. С. Нечеткие модели и сети. М.: Горячая линия - Телеком. 2007.
  7. Киселев К. В. Методика оценки эффективности полимодельной системы распознавания типов воздушных целей. М.: Оборонная техника. 2009.