05运动估计(1)--背景建模

博客已迁移至自建网站,此博客已废弃
请移步至:https://blog.ours1984.top
欢迎大家访问。

相对运动的基本方式

• 相机静止,目标运动–背景提取(减除)
• 相机运动,目标静止–光流估计(全局运动)
• 相机和目标均静止–光流估计

混合高斯背景

根据经验，相机所拍摄到的图像中，人眼感受到的认为不变的背景，其灰度值随时间推移呈现出高斯分布

任何一种分布函数都可以看做是多个高斯分布的组合

混合高斯模型

背景图像的每一个像素分别用由K个高斯分布构成的混合高斯模型来建模

• $I$是输入的像素, $N$是混合高斯模型

• $w_q$是混合高斯模型中第 $q$个高斯分布的权值

• ${\mu }_q$和${\sigma }_q^2$分别表示混合高斯模型中第$q$个高斯分布的均值和方差

$$\begin{array}{rl}P(I)=& \sum _{q=1}^Q{w}_q{N}_q(I;{\mu }_q,{\sigma }_q^2);\\ N_q(I;{\mu }_q,{\sigma }_q^2)=& \frac{1}{\sqrt{2\pi }{\sigma }_q}{e}^{-\frac{(I-{\mu }_q{)}^2}{2{\sigma }_q^2}}\end{array}$$

• 建模任务即为求每一帧图像各个像素的权值、均质、方差，据此判断是否为背景，分离出前景

背景建模流程

1. 初始化模型

   • 选择高斯分布个数Q(通常3～7)和学习率$\alpha$(通常为0.01～0.1)

   • 输入第一帧图像，初始化权值以及各个高斯分布参数
     $$w_q(1)=\frac{1.0}{Q},{\mu }_q=I(1),{\sigma }_q=10(取一个较大值)$$

2. 匹配高斯分布

   • 获取第$k$帧图像，像素为$I(k)$

   • 判断像素是否匹配高斯分布
     $$|I(k)-{\mu }_q(k-1)|<2.5{\sigma }_q(k-1)$$

   • 优选多个匹配
     $$l=\{\begin{array}{ll}\arg \underset{q}{\min }[\frac{|I(k)-{\mu }_q(k-1)|}{{\sigma }_q(k-1)}]& 存在q满足上式\\ \arg \underset{q}{\min }(w_q)& 不存在q满足上式\end{array}$$

3. 更新高斯参数

   • 存在最优匹配
     $$\begin{array}{rl}{w}_q(k)=& \{\begin{array}{ll}(1-\alpha )w_q(k-1),& q\ne l\\ w_q(k-1),& q=l\end{array}\\ {\mu }_q(k)=& \{\begin{array}{ll}{\mu }_q(k-1),& q\ne l\\ (1-\rho ){\mu }_q(k-1)+\rho I(k),& q=l\end{array}\\ {\sigma }_q^2(k)=& \{\begin{array}{ll}{\sigma }_q^2(k-1),& q\ne l\\ (1-\rho ){\sigma }_q^2(k-1)+\rho [I(k)-{\mu }_q(k){]}^2,& q=l\end{array}\\ \rho =& \alpha N_q(I(k);{\mu }_q,{\sigma }_q^2)\end{array}$$

   • 不匹配
     w q ( k ) = { w q ( k − 1 ) , q ≠ l 0.5 min ⁡ q { w q ( k − 1 ) } , q = l μ q ( k ) = { μ q ( k − 1 ) , q ≠ l I ( k ) , q = l σ q 2 ( k ) = { σ q 2 ( k − 1 ) , q ≠ l 2 max ⁡ q { σ 2 ( k − 1 ) } q = l \begin{aligned} w_q(k)=&\begin{cases} w_q(k-1),&q\neq l\\ 0.5\min\limits_q\{w_q(k-1)\},&q=l \end{cases}\\ \mu_q(k)=&\begin{cases} \mu_q(k-1),&q\neq l\\ I(k),&q=l \end{cases}\\ \sigma_q^2(k)=&\begin{cases} \sigma_q^2(k-1),&q\neq l\\ 2\max\limits_q\{\sigma^2(k-1)\}&q=l \end{cases}\\ \end{aligned} wq​(k)=μq​(k)=σq2​(k)=​{wq​(k−1),0.5qmin​{wq​(k−1)},​q​=lq=l​{μq​(k−1),I(k),​q​=lq=l​{σq2​(k−1),2qmax​{σ2(k−1)}​q​=lq=l​​

   注意：一旦权值w发生改变，其他所有权值需要重新归一化

4. 确定背景模型

   • Q个高斯分布按照$\frac{w_q}{{\sigma }_q}$的值从大到小排序

   • 按如下公式选取前B个高斯分布作为背景像素模型
     $$B=\arg \underset{b}{\min }(\sum _{q=1}^b{w}_q>T)$$

   • T为预置的阈值(0.5~1)

5. 判断前景

   当前像素若在b个高斯分布内则为背景，否则为前景

openCV函数

/** @brief Base class for background/foreground segmentation. :

The class is only used to define the common interface for the whole family of background/foreground
segmentation algorithms.
 */
class CV_EXPORTS_W BackgroundSubtractor : public Algorithm
{
public:
    /** @brief Computes a foreground mask.

    @param image Next video frame.
    @param fgmask The output foreground mask as an 8-bit binary image.
    @param learningRate The value between 0 and 1 that indicates how fast the background model is
    learnt. Negative parameter value makes the algorithm to use some automatically chosen learning
    rate. 0 means that the background model is not updated at all, 1 means that the background model
    is completely reinitialized from the last frame.
     */
    CV_WRAP virtual void apply(InputArray image, OutputArray fgmask, double learningRate=-1) = 0;

    /** @brief Computes a background image.

    @param backgroundImage The output background image.

    @note Sometimes the background image can be very blurry, as it contain the average background
    statistics.
     */
    CV_WRAP virtual void getBackgroundImage(OutputArray backgroundImage) const = 0;
};

/** @brief Creates MOG2 Background Subtractor

@param history Length of the history.
@param varThreshold Threshold on the squared Mahalanobis distance between the pixel and the model
to decide whether a pixel is well described by the background model. This parameter does not
affect the background update.
@param detectShadows If true, the algorithm will detect shadows and mark them. It decreases the
speed a bit, so if you do not need this feature, set the parameter to false.
 */
CV_EXPORTS_W Ptr<BackgroundSubtractorMOG2>
    createBackgroundSubtractorMOG2(int history=500, double varThreshold=16,
                                   bool detectShadows=true);