混合高斯背景更新算法的改进和实现

Optimization and Implementation of Mixed Gauss Background Updating Algorithm

混合高斯背景更新算法在处理红外目标检测中,与其他算法相比有着明显的优势。然而传统的混合高斯背景更新算法仍然存在着不足,针对传统算法的不足,提出了有效的改进方法,使混合高斯背景更新算法的处理分为兴趣区和非兴趣区进行处理,在不同的区域采取不同的高斯分布个数和权值更新速率进行处理,同时通过改变不同的背景更新速率可以处理背景发生突变的情况和前景目标静止的情况,有效地提高了混合高斯背景更新算法检测红外目标的能力,仿真实验取得了较好的效果。

Compared with other algorithms,the mixed Gauss background updating algorithm has obvious advantages in infrared target detection.Aiming at the shortcomings of traditional methods,an improved method was proposed,for which,the processing of the mixed Gauss background updating algorithm is divided into areas of interest and non-interest,and different numbers of Gauss distribution and weight updating rates are applied in different areas.Meanwhile,by adjusting the different background updating rate,it can deal with the situation of background mutation and static foreground objects,so the ability of infrared targets detection is effectively improved,and simulation experiments achieve good results.