UDC 621.396.96
ALGORITHM OF TYPICAL OBJECTS RECOGNITION ON URBAN ENVIRONMENT IMAGES AIMING AT UNCERTAINTY ZONES REDUCTION FOR MOBILE ROBOTS POSITIONING
S. N. Kirillov, PhD (technical sciences), full professor, Head of the Department, RSREU, Ryazan; This email address is being protected from spambots. You need JavaScript enabled to view it.
V. M. Berdnikov, PhD (technical sciences), associate professor, RSREU, Ryazan; This email address is being protected from spambots. You need JavaScript enabled to view it.
E. V. Akopov, engineer, RSREU, Ryazan; This email address is being protected from spambots. You need JavaScript enabled to view it.
An algorithm for recognizing typical objects in the area image in order to reduce uncertainty zone in the positioning that uses a digital map and information about the area relief contours is developed. The proposed algorithm includes selecting interest regions in the images by detecting the boundaries of several species followed by the refinement of required objects (industrial chimneys) availability by measuring angular dimensions and calculating the ratio between them. An experimental study of the algorithm resulting in a value of type I error probability in the picture is detected, containing typical object, amounted to 14%, while the value of the probability of type II error is within 5%, and the recognition of typical objects of this kind will 8-22% reduce uncertainty area in average.
Key words: videopositioning, videonavigation, urban area, digital map, area relief contours.