多特征融合的异视角目标关联算法

目标关联在协同多目标探测中具有重要意义，受视角变换等因素的影响，传统目标关联算法在异视角目标观测情况下效果较差。本文提出了一种基于拓扑特征与颜色特征融合的无人机协同侦查目标关联算法，通过利用无人机提供的位置、姿态及其搭载传感器获得的多维特征数据，提取出目标的拓扑特征及颜色特征，并通过D-S证据理论融合多维特征，完成对异视角目标群的关联。实验表明，这种算法框架能够有效帮助无人机完成对目标群的关联任务。

Target association from different perspectives based on multi-feature fusion

Target association has great significance in collaborative multi-target detection. As traditional target association algorithms are affected by factors such as changes in view angle, they are less effective when observing targets from different perspectives. This paper proposes a target association algorithm for collaborative reconnaissance by unmanned aerial vehicles (UAVs) based on the fusion of topological and color features. Using position, attitude, and multi-dimensional feature data obtained by the sensors in the UAV, the topological and color features of the target are extracted and the multi-dimensional features are then fused based on the Dempster–Shafer evidence theory to complete the association of target groups from different perspectives. Experiments show that this algorithm framework can effectively help UAVs to complete the task of target group association.