论文部分内容阅读
采用粒子滤波跟踪变背景下红外目标时,粒子的规模和模板的更新方法对粒子滤波是非常重要的,针对上述问题本文提出了一种基于动态粒子和自适应多模板更新策略的粒子滤波方法,首先通过动态去除权值较小的粒子,不仅防止了粒子的退化,而且提高了算法性能,其次,算法保留两个模板,动态保留下的粒子与两个模板同时比较,对匹配程度最大的模板进行更新,这样可以有效降低跟丢率。最后我们采用红外飞机目标视频检验本文算法,实验结果表明在变背景下,本文算法有效而可靠。
Particle filter is an important part of particle filter when using particle filter to track infrared targets with variable background. Particle filter method based on dynamic particle and adaptive multi-template updating strategy is proposed in this paper. First, by removing particles with smaller weights dynamically, not only the particle degradation is prevented, but also the performance of the algorithm is improved. Secondly, the algorithm preserves two templates, the particles under dynamic reservation are compared with the two templates at the same time, Update, so as to effectively reduce the loss rate. Finally, we use infrared aircraft target video test algorithm in this paper, the experimental results show that the variable background, the algorithm is effective and reliable.