分配
使用拍卖全球最近的邻居分配
描述
[[
returns a table of作业
,,,,unsignedrows
,,,,非分配的collumns
] =分配递减(costmatrix
,,,,costofnonassignment
)作业
of detections to tracks derived based on the forward/reverse auction algorithm. The auction algorithm finds a suboptimal solution to the global nearest neighbor (GNN) assignment problem by minimizing the total cost of assignment. While suboptimal, the auction algorithm is faster than the Munkres algorithm for large GNN assignment problems, for example, when there are more than 50 rows and columns in the cost matrix.
The cost of each potential assignment is contained in the cost matrix,costmatrix
。每个矩阵条目代表可能作业的成本。矩阵行表示轨道,列表示检测。所有可能的任务在成本矩阵中表示。成本越低,任务越有可能。每个轨道最多可以分配给一个检测,并且每个检测最多可以分配给一个轨道。如果行的数量大于列数,则某些轨道未分配。如果列数大于行的数量,则没有分配一些检测。您可以设置一个条目costmatrix
至Inf
至prohibit an assignment.
costofnonassignment
represents the cost of leaving tracks or detections unassigned. Higher values increase the likelihood that every existing object is assigned.
该功能返回一个未分配的曲目列表,unsignedrows
,以及未分配的检测列表,非分配的collumns
。
例子
输入参数
输出参数
References
[[1] Samuel S. Blackman and Popoli, R.现代跟踪系统的设计和分析。Artech House: Norwood, MA. 1999.