红外小目标的背景预测与分割

Background Prediction and Segmentation of Infrared Small Targets

根据红外小目标图像特性的分析,对不同的背景预测方法进行了比较,得出了不同方法的适用条件,并且对不同的门限分割方法进行了介绍。最后采用先均值后形态学Top-Hat滤波,最大绝对对比度自适应门限的方法进行分割,实验表明该方法是有效的。

According to the characteristics of small targets of infrared images, different background prediction methods are compared. Applicable conditions for different methods are obtained and different approaches of threshold segmentation are introduced. Finally after a mean filter and Top - Hat filter, the largest absolute contrast gradient adaptive threshold is used to perform segmentation,and experiments show that this method is effective.