属性测度及其在目标识别中的应用

为了获得一种训练快、精度高、计算量少的雷达目标识别方法 ,研究了属性测度和属性模式识别 ,及属性模式识别在目标识别中的应用 ,并将模糊聚类分析用于属性测度。经过研究发现 ,在不用指标的峰度时 ,可以达到很高的识别效果。与传统的神经网络法相比 ,属性模式识别法在训练速度上得到了大大的提高 ,并在识别效果上优于径向基神经网络法。以雷达目标识别为例证明了该算法的有效性。这为我们研究低分辨雷达提供了一种全新的方法。

Attribute measure and its application in target recognition

This paper considered the attribute measure, attribute model recognition and their application in target recognition. Fuzzy cluster analysis was used to obtain the attribute measure. The attribute model recognition could be trained faster than the traditional radial basis network. As an example, the method was found to be faster than the traditional radial basis network for radar target recognition with real radar data. The attribute model recognition approach also outperformed the radial basis network by recognizing all the targets more accurately in the test. A new method is provided for improving the performance of low resolution radar.