Open AccessUsing Television Image Fragments of a Machine Vision for Verifying Noise Immunity of an Extended Object Velocity Measurement
Author(s) -
Rinat R. Diyazitdinov,
Nikolay N. Vasin
Publication year - 2022
Publication title -
trudy učebnyh zavedenij svâzi
Language(s) - English
Resource type - Journals
eISSN - 2712-8830
pISSN - 1813-324X
DOI - 10.31854/1813-324x-2022-8-1-34-40
Subject(s) - computer vision , measure (data warehouse) , automation , computer science , frame (networking) , reliability (semiconductor) , artificial intelligence , noise (video) , object (grammar) , motion (physics) , image (mathematics) , simulation , engineering , data mining , physics , telecommunications , mechanical engineering , power (physics) , quantum mechanics
Modern diagnostic systems used to control railway infrastructure are equipped with technical vision systems. In addition to video recording, these systems perform recognition tasks and measure parameters necessary for the automation of technological processes. One of the existing tasks is to measure the velocity of long objects. Measuring velocity is necessary for slowing down a carriage on a gravity yard, for the formation of an image of extended objects (separate wagons or convoys) that cannot fit into the frame as a whole (the image is formed from fragments of different frames). The article describes the procedure of verifying fragments of images used to measure the velocity of extended objects, which increases noise immunity. The verification procedure improved the existing algorithm based on comparing two adjacent frames to calculate the speed of motion, thus increasing the reliability of measurements.