一种快速的亚像素图像配准算法

A Fast Sub-pixel Image Registration Algorithm

图像超分辨率重建是在现有红外探测器基础上提升空间分辨率的一种有效方法。超分辨率图像重建是利用一组相互之间存在亚像素位移的低分辨率图像构造出一幅高分辨率的图像,快速、高精度估计图像间的位移是其关键技术之一。提出了一种用于超分辨率重建的亚像素配准算法,算法由特征检测、像素级配准和亚像素级配准三个处理过程组成。在特征检测过程,首先采用梯度算子对图像进行边缘检测,然后对边缘点进行角点预检测,排除非角点像素点,之后再进行Harris角点检测,大大减少了计算量;在像素级配准过程,用NCC算法进行像素级配准,用统计方法去除误匹配点对;在亚像素级配准过程,先对像素级匹配点的邻域进行插值放大,再进行亚像素匹配,误匹配点剔除,相对偏移量计算。对提出的算法进行了仿真实验,结果显示本算法的速度较类似算法速度有较大的提高。

The image super-resolution （SR） reconstruction is an effective method to enhance spatial resolution on the basis of existing infrared detectors. Image SR reconstruction is a process that obtains a high resolution image from a set of sub-pixel shifted low-resolution （LR） images. A rapid and high accuracy algorithm for estimating the shift between the images is one of the key technologies. In this paper, a fast sub-pixel registration algorithm is putted forward. The algorithm consists of three processes： feature detection, pixel-level registration and sub-pixel level registration. In the feature detection process, Firstly, we extract edges from the original images by Sobel operator. Secondly, preliminary detection of comer is used to remove the non-comer pixels. And finally, the Harris comers are extracted. In this way, the time for comer detecting is reduced greatly. In the pixel level registration process, we use NCC algorithm for registration, then statistical method is applied to remove false matching points. In the sub-pixel level registration process, we carry out interpolation in the neighborhood of the pixel-level match points, then the NCC matching and statistical method are used once again. Simulation experiment of the proposed algorithm is carried out, and the result shows that the speed of the proposed algorithm is significantly faster than the similar algorithms.