一种改进的航拍图像序列自动排序算法

An improved automatic sorting algorithm for aerial image sequences

为实现无人机航拍图像中图像序列自动排序,提出了一种基于相位相关法改进的图像序列自动排序算法。该算法利用对数极坐标的方式来表示图像间的平移、旋转、尺度缩放的关系,并利用最大相关度准则以及峰值坐标判断相邻图像的位置关系。实验结果表明,此算法能有效地解决全景图像拼接中序列图像混乱的问题,避免了人工干预,增强了算法的应用范围,具有很强的实用价值。

In order to realize the automatic image sequence sequencing in UAV aerial images, an improved image sequence automatic sorting algorithm based on phase correlation is proposed. The algorithm uses the logarithmic polar coordinates to represent the relationship between the image translation, rotation, scaling, and use the maximum correlation criterion and the peak coordinates to determine the positional relationship between adjacent images. Experimental results show that this algorithm can effectively solve the problem of chaotic image sequence in panorama image mosaic, avoids human intervention and enhances the application range of the algorithm, and has a strong practical value.