S.S. Sadykov - Dr. Sc. (Eng.), Professor, Vladimir State University named after A.&N. Stoletovs (Murom Branch). E-mail: sadykovss@yandex.ru A.V. Terekhin - Senior Lecturer, Department «Information Systems», Vladimir State University named after A.&N. Stoletovs (Murom Branch). E-mail: terehin_murom@mail.ru

The application of systems of automatic recognition (SAR) caused by the increased requirements for conformity of the goods to the quality standards. The quality of the finished product depends on the accuracy of the assembly. Serial production allowed the manual assembly of individual units of product, but in mass production in many cases, it is undesirable, and sometimes impossible. Existing SAR of three-dimensional objects using two main approaches. The first approach is to analyze the image of flat orthogonal projection of the object, which leads to errors in recognition when a pro-jection image of the different objects are identical. Systems based on this approach often do not find practical application in view of their narrow specialization. Fairly universal systems are built on the basis of the second approach. In these systems, detection is performed by using three-dimensional models. The advantage of using such etalon models is the appearance of the possibility of a more accurate comparison of the projection of the object with its etalon analogue. Reference models are created, for example, by using expensive sensors, that determines the distance of points on the surface of objects from the SAR and 2D lasers illuminating strip along the uniformly moving object, which leads to a dramatic increase in the adoption of decisions. The main disadvantages of these systems are their complexity, high cost and considerable time to make a decision. In this paper, we describe the structure of the system of automatic recognition of three-dimensional objects, which is based on the use of two video sensors, diagonal form features of six orthogonal projections of three-dimensional objects and the using three-dimensional models. The scheme describing the process of automatic recognition of three-dimensional objects and the results of research are presented.

 

1. Sensoren: [EHlektronnyjj resurs]. URL: http://www.sensoren.ru (data obrashhenija: 25.11.2013).
2. Sicksensorintelligence: [EHlektronnyjj resurs]. URL: http://www.sick-automation.ru/ (data obrashhenija: 25.11.2013).
3. Sensotec: [EHlektronnyjj resurs]. URL: http://sensotek.ru/ (data obrashhenija: 25.11.2013).
4. Terekhin A.VInnovacionnyjj podkhod k raspoznavaniju trekhmernykh obektov na promyshlennykh sborochnykh konvejjerakh s ispolzovaniem dvukh kamer // Sb. trudov Mezhdunar. nauchno-prakticheskojj konf. «Aktualnye problemy razvitija nauki i obrazovanija». V 7 ch. M.: AR-Konsalt. 2014. S. 44−45.
5. Terekhin A.V. Raspoznavanie neskolkikh ne nalozhennykh trekhmernykh obektov po dvum snimkam // Sb. trudov VI Vseros. mezhvuzovskojj nauchnojj konf. «Nauka i obrazovanie v razvitii promyshlennojj, socialnojj i ehkonomicheskojj sfer regionov Rossii». Regiony Rossii - 2014. Murom: filial VlGU. 2014. S. 415.
6. Terekhin A.V. Algoritm vychislenija diagonalnykh priznakov formy // Algoritmy, metody i sistemy obrabotki dannykh. 2012. № 22. S. 129−138.
7. Sadykov S.S., Terekhin A.V.Identificationofthree-dimensionalobjectsbycomputingestimatesbasedondiagonalfeaturesofformsandoctreemodels // The 11‑thInternationalConference «PatternRecognitionandImageAnalysis: NewInformationTechnologies» (PRIA-11-2013). Samara: IPSIRAS. 2013. V. 2. P. 721−723.
8. Sadykov S.S., Terekhin A.V., Zakharov K.S.Opredelenie diapazonov znachenijj priznakov formy ploskikh geometricheskikh figur pri ikh proizvolnom raspolozhenii v oblasti sceny // Trudy Mezhdunar. simpoziuma «Nadezhnost i kachestvo - 2013» / Pod red. N.K. JUrkova. Penza: Izd-vo PGU. T. 1. S. 343−345.
9. Terekhin A.V. Metod opisanija ehtalonov trekhmernykh obektov po forme ikh proekcijj i priznakam otverstijj // Algoritmy, metody i sistemy obrabotki dannykh. 2013. № 23. S. 65−71.
10. Terekhin A.V.Raspoznavanie obektov metodom vychislenija ocenok s ispolzovaniem diagonalnykh priznakov formy // Izvestija VUZov. Povolzhskijj region. Tekhnicheskie nauki. 2014. № 1. S. 17−25.
11. Terekhin A.V., Savicheva S.V. Algoritm formirovanija kosougolnojj proekcii trekhmernogo obekta po modeli okto-dereva // Algoritmy, metody i sistemy obrabotki dannykh. 2013. № 25. S. 74−81.