一种基于对比度阈值的改进SIFT算法

为了提高基于SIFT(Scale Invariant Feature Transform)图像匹配算法对于图像对比度变化的鲁棒性和算法效率,提出了一种具有可变系数的自适应对比度阈值SIFT算法。根据特征点局部邻域的灰度信息初步确定对比度阈值,同时根据当前特征点的多少确定对比度阈值系数的大小;特征点越多,系数越大,增大对比度阈值从而达到避免特征点数量过大的目的。实验结果表明,改进后的SIFT算法能够根据特征点邻域内的灰度分布情况,自动计算对比度阈值,明显增强了SIFT算法对于低对比度图像匹配的鲁棒性;同时最终提取出的特征点数量避免了过大,稳定在预定的区间内,算法效率提高了1倍多。

Improved SIFT algorithm based on the contrast threshold

In order to improve the SIFT （Scale Invariant Feature Transform）-based image matching algorithm in the robustness and algorithm efficiency for image contrast changes ,an adaptive contrast threshold SIFT algorithm with variable coefficient is designed in this paper.The algorithm Initially identified contrast threshold according to the gray level information of the local neighborhood ,and also determine the size of the coefficient according to the current number of feature points.The experimental results show that the improved SIFT algorithm can automatically calculates the contrast threshold,significantly enhanced the robustness of the SIFT algorithm for low-contrast image matching; the final feature points number is not too large, stabilized at a predetermined interval, so the algorithm efficiency is more than doubled.