中文摘要：

提出了一种基于角点检测和灰度差分不变量区域统计直方图的局部特征描述子相结合的快速匹配算法。通过建立较少图层的阶梯金字塔，利用Shi-Tomasi算法提取图层中数量适中的强角点。采用具有几何意义的低阶灰度差分不变量描述图像特征，改进了传统算法对噪声及角点位置偏差敏感、计算复杂等问题。实验结果表明算法在保证具有旋转、尺度缩放、光照及小视角变化、模糊等性能不变的同时，整体匹配所耗时间以及正确匹配率均优于SIFT和SURF算法。

英文摘要：

A fast matching algorithm is proposed based on corner detection and region histogram of Gray-value Differential Invariants （GDI） local feature descriptor. Establish less layer pyramid and use Shi-Tomasi algorithm to extract a reasonable number of corners for the layers. Use low-level gray scale differential invariants with geometric meaning to describe image features which improve the traditional algorithm, such as sensitive to noise and corner position deviation, complex calculation and so on. Experimental results demonstrate that the proposed algorithm has the invariance for rotation, scale scaling, illumination changes, smaller viewpoint changes, blur and so on, and the time consuming and the correct matching rate are superior to SIFT and SURF algorithm.