DEGREE:一种基于Delaunay三角的任意群目标外形识别方法OA北大核心CSTPCD

Compared with the single or even multiple targets,the group targets exhibit complex and time-varying structure,making the group shape estimation and evaluation quite challenging.This paper proposes a data-driven multi-sensor target group shape modeling and recognition approach to arbitrary shape estimation for group targets,and a group target shape fitting evaluation metric.The proposed approach consists of three parts.Firstly,the information flooding method is used to realize the collection and dissemination of the target informa-tion in the field of view by strongly connected sensors.Secondly,a density peak clustering method is utilized to cluster the data set.Finally,an improved Delaunay triangular network algorithm is used to fit the shape of group targets.The proposed group shape fitting evaluation metric can quantitatively evaluate the accuracy of any group target shape estimate.The effectiveness and reliability of the proposed algorithm are verified in comparison with the classic target shape fitting methods such as the hypersurface and random matrices.