结合SIFT和Harris算法的小型无人机影像拼接方法

Automatic Mosaic of Small Unmanned Aerial Vehicle Images Based on SIFT and Harris Algorithm

考虑到小型无人机影像的成像过程具有视点离散、视角变化有规律等特点,选择拼接参考基准影像,利用无人机的GPS/IMU参数信息计算图像相交区域来减小图像匹配范围;综合利用Harris特征点提取算法和SIFT特征向量计算方法,进行特征点的提取和特征向量的计算,并用PCA算法进行降维处理;在特征匹配过程中,采用最临近(NN)方法和BBF算法提高匹配速度,应用PROSAC特征点提纯和仿射变换整体平差算法提高匹配的精度;最后利用光度对准和加权平均算法进行光度差异消除,实现了无人机影像的拼接。实验结果表明了文章中的图像拼接方法在准确性、效率方面具有优势。

Due to the disperse and regular characteristics of view points and the view angle of UAV Images, this paper first preconditions the image data, combines the Harris feature points with SIFT feature vectors, extracts Harris feature points, calculates the characteristics radius of feature points and SIFT feature vector, and uses PCA （Principal Component Analysis） to reduce the dimension of SIFT feature vectors. And then the most close method （NN） is used to feature matching, the BBF algorithm is applied to search the nearest neighbor feature for improving the matching speed, the PROSAC algorithm is used to purify initial feature point matching pairs, and motion model parame- ters are calculated. Finally, in order to smooth and improve the result image, photometric registration and weighting-average blending are used. The results of the experiment prove that such algorithm is good in efficiency and accuracy.