混合高斯模型和帧间差分相融合的自适应背景模型

提出了运动目标检测中背景动态建模的一种方法。该方法是在Stauffer等人提出的自适应混合高斯背景模型基础上,为每个像素构建混合高斯背景模型,通过融入帧间差分把每帧中的图像区分为背景区域、背景显露区域和运动物体区域。相对于背景区域,背景显露区中的像素点将以大的更新率更新背景模型,使得长时间停滞物体由背景变成运动前景时,被遮挡的背景显露区被快速恢复。与Stauffer等人提出的方法不同的是,物体运动区不再构建新的高斯分布加入到混合高斯分布模型中,减弱了慢速运动物体对背景的影响。实验结果表明,在有诸多不确定性因素的序列视频中构建的背景有较好的自适应性,能迅速响应实际场景的变化。

Adaptive Background Modeling Based on Mixture Gaussian Model and Frame Subtraction

In this paper a dynamic background modeling approach for moving objects detection is proposed. This model is based on mixture Gaussian model suggested by Stauffer et al. It constructs a mixture Gaussians Model for each pixel. In sequence frames subtracting the model classify the pixels in each frame into background area,uncovered background area and moving objection area. In order to quick restore the background covered by stagnated objects when they move again,the model set the update rate in uncovered background area larger than which in background area. Compare to the Stauffer's model,our model moving objection area no longer creates new Gaussian distribution,so it can avoid classifying slow-moving objects to the background.The experimental resultal indicate that our model has preferable adaptive performance to the scene with many uncertain factors,and correspondence quickly.