摘要
针对航拍图像载台随机运动严重,图像分辨率低、目标特征少的特点以及航拍图像电子稳像技术快速、准确的要求,提出基于SIFT算子和Kalman滤波器的航拍图像电子稳像算法。该算法的特点是:由于利用SIFT作为估计图像间运动矢量特征,因此运算速度快;对Kalman滤波状态方程改进,可准确估计有意运动参数;将图像镶嵌技术用于稳定后图像"无定义区"的重建,从而形成快速准确的航拍电子稳像图像。航拍图像实验表明,该算法运算准确性高、稳定性好、速度快。
With serious random movement of platform,the aerial images are low dpi and few features.In order to touch high demand of EIS,a new method proposed which is based on SIFT operator and Kalman filter.The algorithm is characterized by： using SIFT motion vectors between images to operation faster;improving the state equation of Kalman filter to estimate the motion parameters accurately;using the image mosaic technology for ＂undefined zone＂ reconstruction to get fast and accurate electronic image stabilization aerial images.The experiment result shows that this algorithm can stable the aerial image with high accuracy,good stability and speed.
出处
《火力与指挥控制》
CSCD
北大核心
2010年第10期47-50,共4页
Fire Control & Command Control
基金
国家自然科学基金(60634030)
国家自然科学基金资助项目(60702066)
