基于SSDA的图像匹配跟踪算法  

Image Matching Pursuit Algorithm Based on SSDA

作　　者：王文胜 翟尚礼[2] 白俊奇[2] 

机构地区：[1]空军驻沪宁地区军事代表室,南京210007 [2]中国电子科技集团公司第二十八研究所,南京210007

出　　处：《指挥信息系统与技术》2016年第6期77-80,共4页Command Information System and Technology

摘　　要：概述了相关匹配跟踪的研究现状,介绍了序贯相似性检测算法(SSDA)流程及存在的问题,提出了基于SSDA的图像匹配跟踪算法。该算法结合运动轨迹模型,预测了目标在下一帧的位置,并对运动目标进行搜索及跟踪状态估计和维护。试验结果表明,该算法可提高恶劣条件下图像跟踪适应能力,提升了整个图像跟踪的稳定性。The research status of the correlation matching pursuit is summarized. The process and the existed problems of the similarity sequential detection algorithm （SSDA） are introduced. The image matching pursuit algorithm based on SSDA is proposed. Combined with the motion trajectory model, the algorithm predicts the target positions in the next frame, searches the moving target, and estimates and maintains the tracking state. Experimental results show that the algorithm can improve the image tracking ability under harsh conditions, thus improving the stability of the image tracking.

关 键 词：匹配跟踪 轨迹预测 卡尔曼滤波器 序贯相似性检测算法 

分 类 号：TP302.1[自动化与计算机技术—计算机系统结构] TP319[自动化与计算机技术—计算机科学与技术]