融合多重信息的图像局部不变特征描述

Image Local Invariant Feature Description Fusing Multiple Information

以SIFT(Scale Invariant Feature Transform)算法为基础,提出了一种将局部二进制模式(Local Binary Patterns,LBP)描述子和全局上下文(Global Context)信息相融合的图像局部不变特征描述算法,增强了SIFT算法的仿射不变性,以及对处于图像相似区域的特征辨别能力。在特征检测阶段,通过迭代变换,使得SIFT特征点收敛到仿射不变点;在特征描述阶段,为每个特征点计算主方向,分别计算特征点的LBP描述子和全局上下文信息。实验结果表明,提出的局部不变特征描述子对图像仿射、尺度和旋转、光照等变换均具有良好的不变性。

A method for image local invariant feature description based on SIFT（ Scale Invariant Feature Transform） is proposed,which fuses LBP（ local Binary Pattern） descriptor with global context information.This algorithm improves the affine invariance of the initial SIFT algorithm and the ability of distinguishing features where local regions are similar. In the feature detection step,through iterative transformation,the initial feature point derived from the SIFT converges to the affine invariant point. In the feature description step,the dominant orientation is computed for each feature point.Then the LBP descriptor and global context are calculated respectively.Feature matching experiment shows that the image local invariant feature descriptor present in this paper is invariant to image affine,scaling,rotation,illumination changes and so on.