摘要
为克服同源激光图像灰度匹配的匹配速度慢、易被干扰等不足,建议一种基于NProd相关系数和人工免疫的灰度校准匹配方法。NProd相关系数作为相似性度量函数,具有匹配准确、鲁棒性好的优势。对人工免疫算法进行改进,引入动态种群变异操作进行并行搜索,提高了灰度匹配算法的匹配速度。仿真实验表明:该算法具有匹配精度高、鲁棒性强和实时性高的优点。
According to the characteristics of laser active image matching and the real-time requirement of aircraft guid- ance, a grayscale registration matching algorithm based on NProd correlation coefficient and artificial immune algorithm （AIA） is presented. The algorithm uses NProd correlation coefficient as similarity measure to ensure the high accuracy and robustness of matching algorithm. While the AIA based on dynamic population variation probability is taken to be searching strategy to improve the match speed of grayscale computation. The experiments show that： the algorithm has high matching accuracy, strong robustness and less matching time.
出处
《导航定位与授时》
2015年第1期36-41,共6页
Navigation Positioning and Timing
关键词
图像匹配
NProd相关系数
人工免疫算法
Image matching
NProd correlation coefficient
Artificial immune algorithm
