Abstract
An algorithm for tracking objects in a videostream based on the use of a hierarchical Bayesian network is proposed. A specific feature of the algorithm proposed is the use of multi-dimensional scaling, which made possible to significantly reduce the network training time. The algorithm is resistant to temporary monitored object disappearances. It is able to monitor several objects against a complex background and well parallelized.
DOI
10.31144/bncc.cs.2542-1972.2018.n42.p67-75
File
Tarkov.pdf422.59 KB
Pages
67-75