基于Harris的角点检测改进算法  

Improved Corner Detection Algorithm Based on Harris

作　　者：王民钢[1] 王超[1,2] 樊英平 杨尧[1] 

机构地区：[1]西北工业大学航天学院 [2]96263部队 [3]69079部队

出　　处：《计算机仿真》2012年第11期294-298,共5页Computer Simulation

摘　　要：研究红外成像目标跟踪精度优化问题,模板跟踪的累计误差会造成目标跟踪漂移,稳定的角点可以对模板进行校正并减少累积误差。为了实现对运动目标的精确跟踪,使其始终与目标的相关性最强,针对Harris算法检测尺度单一和鲁棒性不足的缺点,采用多分辨率的方法,提出了Harris与B-Spline尺度空间相结合的角点检测算法,并对改进算法进行了数字仿真。结果表明,提出的算法提高了Harris算子的性能及抗噪声性能,可去除错误检测点,并提高了角点检测精度,从而实现了在复杂背景条件下地面目标的准确识别和稳定跟踪。During the process of infrared target tracking, the error accumulation caused by the template content change will lead to target drift. Stable angle point can correct the template and reduce the accumulative error. In or- der to achieve the goal of precise tracking and make it always most relevant to the moving target in image, and for the disadvantages of Harris algorithm which include the single intrusion detection and the weak robustness, we made full use of the advantages of the resolution and put forward the Harris and B-spline scale space combined corner detection algorithm. The digital simulation was done based on this combined method. According to the research results, this combined methodology can enhance Harris operator＇s time performance and noise immunity ability; also, a great num- ber of false detection points can be wiped out, increasing the accuracy of comer detection substantially. Thereby,. the tracking accuracy and stability of the algorithm were verified under complex background conditions.

关 键 词：模板校正 尺度空间 角点检测 多分辨率 

分 类 号：TP391.9[自动化与计算机技术—计算机应用技术]