WAN Shu-ping. Multi-sensor fusion method on object-level based on deviation[J]. 2009, 28(1): 22-23+26. DOI: 10.13873/j.1000-97872009.01.018.
基于偏差的多传感器目标级融合方法
摘要
针对具有多个特征指标的目标识别问题
提出了一种新的模糊传感器数据融合方法。该方法由最大最小隶属度函数得到相对距离矩阵
通过求解总偏差最小的数学规划问题
确定各特征的权重
从而给出目标识别算法
提高了目标识别结果的客观性。应用实例验证了算法的有效性。
Abstract
Aimed at the object recognition problem with multiple characteristic indices
a new fusion method for the fuzzy sensor data is proposed.The method introduces the max-min membership degree function to define the relative distance matrix.By solving the mathematical programming of minimizing total deviations
the weights of characteristic indices are determined.Hence
the algorithm of object recognition is given.The method improves the objectivity of recognition result.The applied example proves that the method is effective.
Target recognition based on modified combination rule [J]. Chen Tianlu & Que Peiwen (Inst. of Automatic Detection, Shanghai Jiaotong Univ., Shanghai 200030, China). Journal of Systems Engineering and Electronics . 2006(02)
Multi-sources information fusion algorithm in airborne detection systems [J] . Yan Yang,Zhanrong Jing,Tian Gao,Huilong Wang. Journal of Systems Engineering and Electronics . 2007 (1)