超像素在机场跑道图像分割中的应用研究

摘要

The research on SAR image segmentation of airport runway is great significance. In this paper, image segmentation of the airport targets was realized, taking reducing the initial point number of participate in iteration as the breakthrough point to accelerate the optimization process of the Meanshift calculation, dividing the original image into several super-pixels with the good homogeneity, and then setting up the appropriate mathematical model. The experimental results show that this method can improve the speed of the segmentation and at the same time can effectively guarantee the accuracy and integrity of airport target segmentation.%机场跑道分割在SAR图像处理研究中有着重要意义。本文通过减少参与迭代的初始点数量为切入点来对均值漂移过程进行加速，将原始图像分割成若干同质性良好的超像素，再建立适当的数学模型，实现对机场目标的分割。实验表明，该方法在提高分割速度的同时，有效提高了分割结果的精确性和完整性。