TY - JOUR
T1 - Dual supporting matching for multi-view target association
AU - Zhang, Yang
AU - Wang, Zhizhen
AU - Xu, Yuan
AU - Zhang, Mingqing
AU - Zhu, Qunxiong
N1 - Publisher Copyright:
© 2025
PY - 2025/9
Y1 - 2025/9
N2 - Efficient target association algorithm can significantly improve the performance of multi-target tracking. However, how to quickly and accurately associate the same target in different cameras has become a key issue for multi-target tracking tasks across perspectives. In this paper, a cross-view multi-objective association algorithm based on epipolar constraint and graph topological association in two-view geometry is proposed. By exploring the graph network topology between two views, the consistency constraints of different perspectives in the same scene are constructed, so as to realize multi-objective association. In specific experiments, this paper carried out simulation experiments and real scene experiments. Compared with related algorithms, the algorithms proposed in this paper have achieved excellent performance.
AB - Efficient target association algorithm can significantly improve the performance of multi-target tracking. However, how to quickly and accurately associate the same target in different cameras has become a key issue for multi-target tracking tasks across perspectives. In this paper, a cross-view multi-objective association algorithm based on epipolar constraint and graph topological association in two-view geometry is proposed. By exploring the graph network topology between two views, the consistency constraints of different perspectives in the same scene are constructed, so as to realize multi-objective association. In specific experiments, this paper carried out simulation experiments and real scene experiments. Compared with related algorithms, the algorithms proposed in this paper have achieved excellent performance.
KW - Adjacent cross entropy
KW - Geometric constraint
KW - Graph matching
KW - Support degree
KW - Topological association
UR - https://www.scopus.com/pages/publications/105007904329
U2 - 10.1016/j.aei.2025.103485
DO - 10.1016/j.aei.2025.103485
M3 - Article
AN - SCOPUS:105007904329
SN - 1474-0346
VL - 67
JO - Advanced Engineering Informatics
JF - Advanced Engineering Informatics
M1 - 103485
ER -