Abstract. The paper is devoted to the implementation of logic-objective approach to the solving of a polyhedron contour images (in particular partially covered images) recognition problem in a complex scene represented on the display screen. A way of predicate value calculation for representation the display screen is described in the paper. Examples of a program run constructing descriptions of both separate pictures and classes of objects are presented. For recognition of partially covered objects on the complex scene the concept of partial deducibility is used. Additionally the certainty level of the correct recognition is calculated.

Keywords: artificial intelligence, pattern recognition, predicate calculus.

ACM Classification Keywords: I.2.4 ARTIFICIAL INTELLIGENCE Knowledge Representation Formalisms and Methods – Predicate logic.

Link:

SOFTWARE FOR THE RECOGNITION OF POLYHEDRON CONTOUR IMAGES IN THE FRAMEWORK OF LOGIC-OBJECTIVE RECOGNITION SYSTEM

Natalya Bondar, Tatiana Kosovskaya

http://www.foibg.com/ijima/vol01/ijima01-2-p09.pdf