(14-15(01)) * << * >> * Russian * English * Content * All Issues

Information technology for analyzing the quality of the system for detecting objects in the images
E.I. Kolomiets, V.V. Sergeev

 PDF, 2214 kB

Pages: 166-178.

Full text of article: Russian language.

Abstract:
As a rule, the quality of the system for detecting objects in the images is determined by criteria that have the meaning of probabilities: errors in detecting objects, incorrect classification of the image, the system’s failure to achieve its purpose, etc. [1,2]. In situations of practical interest, the values of such quality indicators are rather small, and therefore their experimental measurement is very complicated. Additional difficulties are caused by the high dimension of the impact at the input of the detection system dealing with images. Below we describe the method of conducting statistical experiments with the system for detecting objects in images, which allows to reduce significantly the number of experiments needed to obtain the required accuracy of determining its quality characteristics as compared with the traditional methods [3,4]. We should note that all of the above results can be used to assess the probabilities of rare events that have a fairly arbitrary meaning.

Citation:
Kolomiets EI, Sergeev VV. Information technology for analyzing the quality of the system for detecting objects in the images. Computer Optics 1995; 14-15(1): 166-178.

References:
  1. Pratt WK, Digital Image Processing; Moscow: Mir Publisher; 1982; 1: 312; 2: 480.
  2. Yaroslavsky LP. Digital signal processing in optics and holography: Introduction to digital optics; Moscow: Radio i svyaz; 1987; 296.
  3. Ermakov SM, Mikhailov GA. Statistical Modeling; Moscow: Nauka; 1982; 296.
  4. Kleinen J. Statistical methods in simulation modeling; Moscow: Statistics; 1978; 1: 223.
  5. Cheng RCH. Variance Reduction Methods. Winter Simulation Conference. Washington D.C.: 1986; Proceedings; 60-68.
  6. Wilson JP. Variance Reduction Techniques for Digital Simulation. American Journal of Mathematical and Management Sciences; 1984; 3-4: 277-312.
  7. Kuzmin SZ, Kostina SS. Application of the significant-sample method to the study of signal-detection systems on the basis of simulation. News of universities. Radioehlektronika; 1984; 9: 8-14.
  8. Kolomiets EI, Sergeev VV. The method of significant sampling in the task of studying the procedure for detecting objects. Abstracts of the IV All-Union Conference "Mathematical Methods of Pattern Recognition"; Riga: 1989; 2: 69-70.
  9. Belousova MP, Kolomiets EI, Sergeev VV. Study of the detection system performance using simulation modeling. Problems of Informatics: Abstracts of the scientific-practical conference with international participation; Samara: 1991; 77-78.
  10. Stechkin SB, Subbotin YN. Splines in computational mathematics. - Moscow: Nauka; 1976; 248.
  11. Vainikko GM, Veretennikov AY. Iterative procedures in ill-posed problems; Moscow: Nauka; 1986; 184 .

© 2009, IPSI RAS
151, Molodogvardeiskaya str., Samara, 443001, Russia; E-mail: journal@computeroptics.ru ; Tel: +7 (846) 242-41-24 (Executive secretary), +7 (846) 332-56-22 (Issuing editor), Fax: +7 (846) 332-56-20