Video measurements and a correlation stereo indentification problem solving using the gradient approach and pyramids of stereopairs

Аuthors

Knyaz V. A.¹, Ilya V. Krupennikov I. I.², Sharonov A. V.^*

1. State Institute of Aviation Systems, 7, Victorenko str., Moscow, 125319, Russia
2. Moscow Aviation Institute (National Research University), 4, Volokolamskoe shosse, Moscow, А-80, GSP-3, 125993, Russia

*e-mail: anatoly.sharonov@yandex.ru

Abstract

Full time needed for stereo identification is intolerably large in problems of stereopair-based video measurement aimed to motion control for dynamic systems. An algorithm is suggested to overcome this situation. The algorithm uses some gradient approach and an image pyramid to perform stereo identification for pixels obtained from two cameras with mutually known orientation parameters. The algorithm is implemented on experimental facility equipped with CCD cameras. The resolution is 640×480 pixels and the frame rate is 30 frames per second for the cameras.