中文说明：资源描述 本代码提供了一种基于SURF特征点的人脸识别方法，可以实现人脸的高准确率匹配的同时，还可以高准确率识别具有任意旋转角度的人脸图像。SURF（加速的具有鲁棒性特征）算法是基于SIFT（尺度不变特征变换）算法发展而来，对图像的尺度和旋转具有不变性，且对光照变化的图像匹配具有很好的鲁棒性。我们发现把SURF算法用于人脸识别，可以很好解决目前大多数算法，如LBP，LGBP等对具有一定旋转角度的人脸图片识别率偏低的情况，并可以达到很高的识别率。当然，这依赖于人脸图像的SURF特征点的一个良好的匹配。从人脸图像抽取的SURF特征点后，为了提高SURF算法在人脸识别方面的性能，优化SURF特征点的匹配，将是这篇论文的重点。 首先本系统基于SURF算法实现人脸识别，识别率可以高达95%，。由于SURF算法比较成熟，在这里不做详细的介绍，具体可参考文献。这里简单说一下算法的实现过程：SURF算法应用于人脸识别的思路：首先是两张人脸图片各自抽取出来的特征点进行初匹配，接着介绍如何用HOUGH变换优化初匹配的结果得到候选匹配，然后介绍仿射参数计算，进一步优化匹配的结果。最后介绍两张人脸图片相似度的计算公式。
English Description：
Application background this code provides a face recognition method based on SURF feature points, which can achieve high accuracy of face matching, but also can identify the face image with arbitrary rotation angle. SURF (speed up feature) algorithm is developed based on SIFT (scale invariant feature transform) algorithm, which is invariant to image scale and rotation, and has good robustness to image matching. We found that the SURF algorithm for face recognition, can be a good solution to the current majority of algorithms, such as LBP, LGBP, and so on have a certain rotation angle of face image recognition rate is low, and can achieve a high recognition rate. Of course, this relies on a good matching of the SURF feature points of the face image. In order to improve the performance of SURF algorithm in face recognition, the matching of SURF feature points is the key point of this paper in order to improve the performance of SURF algorithm.&nbs