﻿ 改进混合高斯模型的电网检修人员行为检测
 计算机系统应用  2020, Vol. 29 Issue (10): 242-247 PDF

1. 中国科学院大学, 北京 100049;
2. 中国科学院 沈阳计算技术研究所, 沈阳 110168

Behavior Detection of Grid Maintenance Personnel Based on Improved Mixed Gaussian Model
JIA Yu-Wei1,2, WANG Han-Jun2
1. University of Chinese Academy of Sciences, Beijing 100049, China;
2. Shenyang Institute of Computing Technology, Chinese Academy of Sciences, Shenyang 110168, China
Abstract: Aiming at moving object detection, the current algorithm has certain applicability and limitations, as well as incomplete detection information. Based on the inter-frame difference method and hybrid Gaussian model, an improved hybrid Gaussian model target detection algorithm is proposed. To solve the problem of incomplete background contours of moving targets caused by the inter-frame difference method. This method is based on the traditional mixed Gaussian model, within a certain number of frames, checks the weights of all Gaussian distributions, deletes the Gaussian distributions that meet the conditions, and finally obtains a moving target with a clearer outline. The experimental results show that the algorithm in this study fully considers the influence of background characters on moving target monitoring. During the experiment, real grid data is used. The result shows that the proposed algorithm is 3.37% more accurate than other algorithms, with better accuracy and better adaptability to the environment.
Key words: moving target detection     mixed Gaussian model     inter-frame difference method

1 基于视频内容的运动检测算法 1.1 帧间差分法

 图 1 帧间差分法基本原理图

 ${D_m}(x, y) = |{F_m}(x, y) - {F_m}_{ - 1}(x, y)|$ (1)

Bm(x, y)是Dm(x, y)进行阙值化后的二值化图像, 阙值大小为T. 当差分图像Dm(x, y)中的像素点的灰度值大于阙值T时, 就认为该像素点是运动目标, 相反, 如果灰度值小于阙值T, 则认为该像素点是背景图像, 具体公式如下:

 ${B_m}(x,y) = \left\{ {\begin{array}{*{20}{c}} 1&{{D_m}(x,y) > T}\\ 0&{{D_m}(x,y) < T} \end{array}} \right.$ (2)

1.2 混合高斯模型

 $\left\{ {{X_i}(x, y)|1 \le i \le t} \right\} = \left\{ {{X_1},{X_2},\cdots,{X_t}} \right\}$ (3)

k个高斯分布线性加成组成一个高斯混合模型 $\rho$ , 如图2所示.

 图 2 混合高斯模型结构示意图

 $p = \sum\limits_{i = 1}^k {{\varphi _{i,t}}{\rho _{i,t}}(X,{\mu _{i,t}},{\Sigma _{i,t}})}$ (4)

 ${\rho _i}(X,{\mu _{i,t}},{\Sigma _{i,t}}) = \frac{1}{{{{(2\pi )}^{\frac{N}{2}}}|{\Sigma _{i,t}}{|^{\frac{1}{2}}}}}{e^{ - \frac{1}{2}{{(X - {\mu _{i,t}})}^T}{\sum _{i,t}}^{ - 1}(X - {\mu _{i,t}})}}$ (5)

 $\left\{ \begin{array}{l} {\varphi _{i,t}} = (1 - \alpha ) + \alpha R \\ {\mu _{i,t}} = (1 - \beta ){\mu _{i,t - 1}} + \beta {X_i} \\ \;\;\;\;\;\;=(1 - \beta )\sigma _{i,t - 1}^2 + \beta {({X_i} - {\mu _{i,t}})^T}({X_i} - {\mu _{i,t}}) \\ \beta = \alpha /{\varphi _i} \\ \end{array} \right.$ (6)

 $\delta = \arg {\min _b}\left( {\sum\limits_{i = 1}^b {{\varphi _{i,t}}} \ge T} \right)$ (7)

2 改进的混合高斯模型

 $\alpha = \left\{ {\begin{array}{*{20}{c}} {\dfrac{1}{{2f}}}&{f\le{T_0}}\\ {\dfrac{1}{{2{T_0}}}}&{f > {T_0}} \end{array}} \right.$ (8)

 图 3 背景前景判断图

 ${T_0} = \sum\limits_{i = 1}^k {{\varphi _{i,t}}|{\sum _{i,t}}|} + {\mu _{i,t}}$ (9)

 图 4 改进方法图

3 实验设计与结果分析 3.1 实验设计

 图 6 室内检测运动目标原图

 图 7 室内运动目标检测图

 图 8 改进前算法检测结果

 图 9 改进后算法检测结

 图 5 整体算法流程图

3.2 实验结果对比

4 实验分析

5 结论

 [1] 王晓强. 网络视频监控系统现状与发展. 现代工业经济和信息化, 2019, 9(8): 112-113. [2] 胡觉晖, 李一民, 潘晓露. 改进的光流法用于车辆识别与跟踪. 科学技术与工程, 2010, 10(23): 5814-5817. DOI:10.3969/j.issn.1671-1815.2010.23.052 [3] 朱明旱, 罗大庸. 基于帧间差分背景模型的运动物体检测与跟踪. 计算机测量与控制, 2006, 14(8): 1004-1006, 1009. DOI:10.3321/j.issn:1671-4598.2006.08.009 [4] 王辉, 于立君, 孙蓉, 等. 融入帧间差分法的深度学习目标识别仿真研究. 实验技术与管理, 2019, 36(12): 178-181, 190. [5] 陈媛, 胡娜, 余秋月. 融合背景差分法和帧间差分法的运动目标检测. 现代计算机, 2019(34): 50-53. [6] 屈晶晶, 辛云宏. 连续帧间差分与背景差分相融合的运动目标检测方法. 光子学报, 2014, 43(7): 213-220. [7] 郭碧, 丁春平. 基于帧间差分累积的铁路限界异物检测提取算法. 铁道标准设计, 2019, 63(9): 153-158. [8] 李珣, 南恺恺, 景军锋. 基于改进混合高斯模型的车辆多目标检测方法. 西安工程大学学报, 2017, 31(6): 795-802. [9] 杜鹃, 吴芬芬. 高斯混合模型的运动目标检测与跟踪算法. 南京理工大学学报, 2017, 41(1): 41-46. [10] Xie CY, Hong RX, Wang J, et al. Vehicle target detection based on improved mixed gaussian modeling. Proceedings of 2018 International Conference on Computer Science, Electronics and Communication Engineering (CSECE 2018). Paris, France. 2018. 208–211. [11] 於正强, 潘赟, 宦若虹. 一种结合帧差法和混合高斯的运动检测算法. 计算机应用与软件, 2015, 32(4): 129-132, 270. DOI:10.3969/j.issn.1000-386x.2015.04.031 [12] 李明, 赵勋杰. 改进的基于高斯混合模型的运动目标检测算法. 计算机工程与应用, 2011, 47(8): 204-206. DOI:10.3778/j.issn.1002-8331.2011.08.060