文中给出一种新的点匹配方法，这种方法对从传感器图像中抽取的点与从参考图像抽取的图像点进行匹配。点匹配算法要求从两个点集之间得到仿射变换，这种变换使得一个点集到另一个点集的距离最小。在这里一种新的距离度量———Ｈａｕｓｄｏｒｆ距离作为点匹配的距离度量（相似性度量）得到应用，给出了在匹配度量下图像匹配的结果，说明这种方法有很强的抗干扰性和容错性。 In this paper, a new point matching method is proposed, which matches the points extracted from the sensor image with those extracted from the reference image. The point matching algorithm requires affine transformations between two point sets that minimize the distance from one point set to another. Here, a new distance metric, Hausdorff distance, is used as the distance metric (similarity measure) for point matching, the result of image matching under matching metric is given, which shows that this method has strong anti-interference And fault tolerance.