一种基于自适应背景划分的点目标检测算法

摘要

A novel single-frame detection algorithm based on Adaptive Background-Division is proposed for air point-target tracking under the environment including complex ground scene. Firstly, the gray variance of each row of the image is found. And then, the Ratio of Divided Background Variance (RDBV), representing difference of gray distribution uniformity, is calculated to detect the boundary of sky background adaptively.Finally, the single-frame detection algorithm based on spatial filter is performed. Experiment results indicate that this method can reduce the number of potential targets greatly without undetected targets, and it is applicable for both dark and light background.%针对含有部分复杂地面背景的空中点目标跟踪环境提出一种基于自适应背景划分(Adaptive Background-Division)的单帧检测算法.先求出图像的行灰度方差,计算表示灰度分布均匀性差异的划分背景方差比(RDBV),自适应地检测出天空背景边界以屏蔽地面背景;再进行基于空域滤波的单帧检测算法.实验结果表明该算法在保证无漏检的情况下大大降低了潜目标数量,而且对于暗背景和亮背景都同样适用.