月球表面图像的SIFT特征提取与匹配

摘要

在分析月球表面不同尺度、不同谱段图像特点的基础上,本文以Visual C++和OpenCV为开发平台,对月球表面多光谱图像数据和多尺度图像对数据进行特征点提取和匹配,并验证算法的有效性.实验结果表明,改进的SIFT特征具有旋转、平移、尺度缩放、亮度变化和视角变换的不变性,能较好地完成月球表面各种图像的特征提取和匹配.%On the basis of analyzing multi-spectral and different-scale lunar surface image feature points, this paper extracts feature point and realizes stereo matching of the multispectral lunar surface image data and multi-scale image pairs data, and verifies the effectiveness of the proposed algorithm with Visual C ++ and OpenCV development platform. Results show that improved SIFT feature is invariant to rotation, scale, intensity roughness, preserves stability of view variations, and completes feature extraction and matching of lunar surface image better.