摘要
针对单一目标图像目标识别结果的不确定性问题,分析了不确定性推理相关理论方法,通过多传感器获取目标图像,并将DS证据推理方法应用于多传感器目标识别结果的决策融合环节。试验中基于单一目标影像目标识别过程中的模型匹配相关系数计算基本概率赋值及不确定性,最终通过DS证据组合规则得出融合识别结果,从而降低单一目标图像目标识别的不确定性。
Aiming at the uncertainty problem of target recognition based on single image, the relevant theories are analyzed, and DS evidential reasoning theory is confirmed to be used in the step of decision fusion based on target recognition results from multiple sensors. In experiment, Basic Probability Assignment （BPA） and uncertainty are calculated based on the matching degree with target model. Finally, the results of simulation experiment show that the uncertainty of target recognition result is reduced effectively.
出处
《无线电通信技术》
2013年第1期70-72,共3页
Radio Communications Technology
