TWO-THRESHOLD ALGORITHM FOR DETECTING POINT OBJECTS FROM STEREO IMAGES
V. S. Kirichuk, A. K. Shakenov
Institute of Automation and Electrometry. Siberian Branch, Russian Academy of Sciences, pr. Akademika Koptyuga 1, Novosibirsk, 630090 Russia
Keywords: search for dynamic objects, small-size objects, stereo vision, detection by stereo sequences
Abstract
This paper considers various approaches to improving the detection performance of point dynamic objects by using stereo vision under conditions where complete joint processing of a stereo pair of images is impossible. An improved triangulation algorithm with double thresholding is implemented. The possibility of forming marks from epipolar lines is studied for the case of a priori known shooting geometry. The situation of stereo observation of a dynamic object by moving cameras is simulated. The dependence of the detection probability on the level of false alarms is obtained experimentally for the proposed algorithms using model data.
|
