结合雷达目标特征的红外成像目标识别

Target recognition using IR image and radar features

由于单传感器在导弹的目标识别和跟踪方面存在的固有缺陷，近年来，数据融合技术在这一方面得到广泛地应用和研究。文中提出了在双模（ 雷达，红外） 数据融合方案中，结合雷达目标特征的红外成像目标识别方法，并详细讨论了所涉及的图像处理技术（ 区域分割，骨架提取） 和人工神经网络分类技术。最后给出了实验结果和结论。

Because of the shortcoming of target recognition and tracking using only one sensor, data fusion is applied and researched widely in recent years. In this paper, a new way of target recognition using IR image and radar signal is presented, and the processing way to extract features from IR image is discussed in detail. Two different ways to recognize the target according to the distance are used. When the distance is far, the target is some pixels on IR image, some rules are used to recognize the target. When the target is near, it becomes a region on IR image, the shape feature is used to recognize the target. And ANN is used to combine two sets of features getting from IR image and radar. Finally, the result of experiment and conclusion has been given.