基于Matlab的SIFT和SURF算法在无人机影像配准中的对比研究

Contrast and research of SIFT and SURF algorithm in UAV image registration based on Matlab

无人机影像配准是无人机影像拼接中最关键的技术问题,首先介绍了SIFT和SURF两种无人机影像配准常用算法,并进行了比较,然后选取无人机岩鹰UX-1000对某地区的航拍影像基于Matlab进行实验,对两种算法配准的速度和精度进行分析研究。实验表明,两种算法在无人机影像配准过程中都可以抑制和抵消一定的干扰,但SIFT算法的精度更高,适合用于对精度要求高的工程,SURF算法的速度更优,适合用于对速度要求高的工程。

The UAV image registration was the key technology of UAV image fusion,firstly,the two kind of UAV image registration algorithms were introduced and compared,SIFT algorithm and SURF algorithm,and then,the aerial images of a region obtained by the Yanying UX-1000 UAV was selected to experiment based on Matlab,the speed and accuracy of the two algorithms were analyzed and studied. Experimental results showed that the two algorithms could suppress and cancel some interference in the process of UAV image registration,but the SIFT algorithm was better accurate and suitable for the requirements of high precision engineering,the speed of SURF algorithm was better and suitable for engineering with high speed requirement.