红外技术, 2017, 39 (6): 529, 网络出版: 2017-07-07
基于图像配准的无人机目标精确定位算法
UAV Reconnaissance Images Accurate Targeting Method Based on Image Registration
无人机侦察图像 目标定位 特征点检测 侧抑制竞争 图像配准 UAV reconnaissance image target orientation feature point detection lateral inhibition competition image registration
摘要
为了满足“察打一体化”无人机精确打击目标的要求, 针对现有目标定位算法的准确率低、实时性差的特点, 本文提出了一种基于图像配准的无人机目标精确定位算法。该算法主要分为两个阶段: 特征点检测阶段和目标精确定位阶段。首先采用基于侧抑制竞争的特征点检测算法, 实现局部图像中目标特征点的检测工作; 然后利用图像局部信息的配准算法, 完成目标所处的局部区域图像的精确配准工作, 最终实现了目标的高精度定位。实验结果显示该算法的定位精度可以达到 0.21 m, 能够满足精确作战时的目标情报保障需求。
Abstract
In order to meet the needs of the integrated reconnaissance/strike UAV precision attack, and to overcome low accuracy and worse real-time performance of these existing target localization algorithms, this paper proposes a UAV accurate positioning algorithm based on image registration, which consists of two parts: feature point detection and target accurate positioning. First, the feature point detection algo-rithm based on lateral inhibition competition is used to detect the local image feature points; and then the image registration algorithm of local information is applied to complete the work of local image precise registration, ultimately achieving the goal of high-precision positioning of the targets. The experiments show that the positioning accuracy of this algorithm can reach 0.21 m, which can meet the target informa-tion security needs during combat.
杨帅, 程红, 李婷, 赵鹤. 基于图像配准的无人机目标精确定位算法[J]. 红外技术, 2017, 39(6): 529. YANG Shuai, CHENG Hong, LI Ting, ZHAO He. UAV Reconnaissance Images Accurate Targeting Method Based on Image Registration[J]. Infrared Technology, 2017, 39(6): 529.