基于对目标识别特点的认识，提出了可用于坦克目标识别的层次化信息融合算法。对高分辨率毫米波非相干雷达、毫米波辐射计及红外辐射计等目标识别概率进行的融合实验表明，该算法不仅可融合不同层次的信息，而且具有较强的处理信息不确定性的能力和容错能力，极大地提高了目标识别概率。 Based on the recognition of the characteristics of target recognition, a hierarchical information fusion algorithm that can be used to identify tank targets is proposed. The fusion experiments on the target recognition probability of high resolution millimeter wave incoherent radar, millimeter wave radiometer and infrared radiometer show that this algorithm not only can fuse different levels of information but also has strong ability to deal with information uncertainty And fault tolerance, which greatly improves the probability of target recognition.