摘要
基于图像序列分析的图像制导极大地提高了成像武器的跟踪精度。给出了一种基于可变模板技术和遗传退火算法的图像序列目标跟踪算法。根据目标的物理特征构造出可变模板,定义合适的能量函数,采用遗传退火算法进行优化搜索。实验结果表明,该算法能够精确检测并跟踪目标,具有很快的收敛性和较高的抗噪声性。
The image guidance using image sequences analysis improves the shoot accuracy of missiles greatly. In this paper, we present a new method of automatic object recognition and tracking based on deformable template and genetic annealing algorithm. After the constructing object template and defining energy function according to their features, we use genetic annealing algorithm to search for the optimal solution. Experiment results show the method can be applied to detect and track the object efficiently and accurately.
出处
《国防科技大学学报》
EI
CAS
CSCD
北大核心
2003年第5期80-84,共5页
Journal of National University of Defense Technology
关键词
目标跟踪
图像序列
可变模板
能量函数
遗传退火算法
object tracking
image sequences
deformable template
energy function
genetic annealing algorithm
