基于夜间环境下的改进的AMBER背景减除算法

Improved AMBER Background Subtraction Algorithm Based on Night Environment

提出的改进算法在AMBER背景减除算法的基础上,将LBSP纹理特征引入模型,并依据空间一致性提出了"一点调整多点联动"的动态阈值调整机制。实验结果表明,提出的改进算法在changedetection.net数据库中night Videos子条目下的检测结果较原算法的精确性有了较大的提高。

In this paper,an improved algorithm based on AMBER background subtraction algorithm is introduced,and the LBSP texture is introduced into the model.According to the spatial consistency,the dynamic threshold adjustment mechanism is proposed.The experimental results indicated that the improved algorithm proposed in this paper behaves more accuracy than the original AMBER background subtraction algorithm in the night Videos subset of the changedetection.net database.