论文部分内容阅读
针对无人机拍摄的影像偏角大、投影差明显的问题,提出一种基于影像分割与匹配特征的无人机影像变化检测方法。该方法基于匹配的特征点和分割的单元,以配准误差为缓冲半径进行相关运算,并提出了双向互相关方法来抑制影像分割不一致对变化检测结果的影响。实验结果表明,该方法提高了无人机影像变化检测的精度,对无人机影像由于大倾角所带来的配准误差问题有较好的容忍度,并削弱了无人机影像的投影差对于变化检测的影响。
Aiming at the problem that the image of the UAV is large and the difference of the projection is obvious, this paper proposes a method of UAV image change detection based on the image segmentation and matching features. Based on the matched feature points and the partitioned units, the method performs the correlation operation with the registration error as the buffering radius and proposes a bidirectional cross-correlation method to suppress the influence of the inconsistency of image segmentation on the change detection result. Experimental results show that this method improves the precision of UAV image detection and has better tolerance to the registration error of UAV image due to large dip angle and weakens the projection difference of UAV image Impact on change detection.