Improved ViBe algorithm based on adaptive threshold and dynamic update factor
-
摘要:
针对传统ViBe算法不能及时反映场景变化，动态场景适应性差等问题，提出一种改进的ViBe算法。改进内容包括：采用随机选取背景样本和24邻域法获取初始背景，可以加速“鬼影”消融；结合大津法（OTSU）和均匀性度量法的平均自适应阈值计算方法，可以提高算法对树叶晃动、水波纹和光照变化等环境的适应性，最大限度保留有效像素；更新阶段引入自适应更新因子，可以有效减少被误判的概率，从而增强算法的鲁棒性；最后通过形态学处理和滤波使目标更加完整。采用标准数据集视频对改进算法进行了测试和对比分析，改进算法相对于KDE算法、GMM算法和传统ViBe算法各项指标均有大幅度提高，精确度分别提高30.44%、40.72%和20.95%，错分比分别降低了43.28%、40.59%和29.43%。
Abstract:Aiming at the problems that the traditional visual background extractor (ViBe) algorithm cannot reflect the scene changes in time and has poor adaptability to dynamic scenes, an improved ViBe algorithm was proposed by using randomly selected background samples and 24 neighborhood method to obtain the initial background, which could accelerate the "ghost" ablation. The average adaptive threshold calculation method was adopted to improve algorithm adaptability to external dynamic environment and illumination changes in combination with OTSU method and uniformity measurement method, which retained effective pixels to the greatest extent. In the update phase, the adaptive update factor was introduced, which could effectively reduce the misjudgment probability, so as to enhance algorithm robustness. Finally, the target was more complete through morphological processing and filtering. The standard dataset video was applied to test and compare the improved algorithm. Compared with kernel density estimation (KDE) algorithm, Gaussian mixed model (GMM) algorithm and traditional ViBe algorithm, the indexes of the improved algorithm were greatly improved. The accuracy is improved by 30.44%, 40.72% and 20.95%, respectively and the percentage of wrong classifications is reduced by 43.28%, 40.59% and 29.43%, respectively.
-
-
![]()
图 1 ViBe算法中不同背景建模算法目标检测效果对比图
Figure 1. Comparison images of target detection effects in ViBe algorithm with different background modeling algorithms
![]()
图 2 ViBe算法中不同前景和背景判断阈值目标检测效果对比图
Figure 2. Comparison images of target detection effects in ViBe algorithm with different foreground and background judgment thresholds
![]()
图 3 ViBe算法中不同背景更新因子目标检测效果对比图
Figure 3. Comparison images of target detection effects in ViBe algorithm with different background update factors
![]()
图 5 4种算法在6类场景下目标检测效果对比图
Figure 5. Comparison images of target detection effects of four algorithms in six types of scenes
表 1 4种算法在6类场景下测试指标平均值
Table 1 Average values of test indexes of four algorithms in six types of scenes
算法 Rprecision Rrecall PPWC F RFPR RFNR KDE 0.702626667 0.506002 3.4639 0.548503 0.015443 0.493998 GMM 0.6513 0.362367 3.3073 0.446828 0.006733 0.637633 VIBE 0.757776667 0.567875 2.7843 0.639273 0.009717 0.432125 本文算法 0.916511667 0.604228 1.9648 0.720945 0.00253 0.395772 
下载: 导出CSV
-
[1] 杨丹, 戴芳. 运动目标检测的ViBe算法改进[J]. 中国图象图形学报,2018,23(12):1813-1828. doi: 10.11834/jig.180304 YANG Dan, DAI Fang. Improved ViBe algorithm for detection of moving objects[J]. Journal of Image and Graphics,2018,23(12):1813-1828. doi: 10.11834/jig.180304
[2] 王伟, 郭中华, 兰旭婷. 一种基于阴影检测的运动目标分割改进算法[J]. 计算机与网络,2021,47(9):67-71. doi: 10.3969/j.issn.1008-1739.2021.09.056 WANG Wei, GUO Zhonghua, LAN Xuting. An improved algorithm for moving target segmentation based on shadow detection[J]. Computer & Network,2021,47(9):67-71. doi: 10.3969/j.issn.1008-1739.2021.09.056
[3] 唐悦, 吴戈. 基于灰度投影运动估计的ViBe改进算法[J]. 长春理工大学学报(自然科学版),2021,44(1):95-101. TANG Yue, WU Ge. Improvement of ViBe algorithm based on gray projection motion estimation[J]. Journal of Changchun University of Science and Technology (Natural Science Edition),2021,44(1):95-101.
[4] SHAHBAZ A, HARIYONO J, JO K H. Evaluation of background subtraction algorithms for video surveillance[C]//2015 21st Korea-Japan Joint Workshop on Frontiers of Computer Vision. USA: IEEE, 2015: 1-4.
[5] SINGLA N. Motion detection base on frame difference method[J]. International Journal of Information & Computation Technology,2014,4(15):1559-1565.
[6] LIU X, ZHAO G, YAO J, et al. Background subtraction based on low-rank and structured sparse decomposition[J]. IEEE Transactions on Image Processing,2015,24(8):2502-2514. doi: 10.1109/TIP.2015.2419084
[7] 瞿中, 刘帅, 刘妍. 融合时域信息的自适应ViBe算法[J]. 计算机工程与设计,2019,40(3):782-787. QU Zhong, LIU Shuai, LIU Yan. Self-adaptive ViBe algorithm with time domain information[J]. Computer Engineering and Design,2019,40(3):782-787.
[8] 李善超, 车国霖, 张果, 等. 基于自适应ViBe的运动目标检测方法及其应用[J]. 数据通讯,2021(1):48-54. LI Shanchao, CHE Guolin, ZHANG Guo, et al. Moving target detection method based on adaptive ViBe and its application[J]. Data Communications,2021(1):48-54.
[9] 杨恒, 王超, 姜文涛, 等. 基于随机背景建模的目标检测算法[J]. 应用光学,2015,36(6):880-887. doi: 10.5768/JAO201536.0602001 YANG Heng, WANG Chao, JIANG Wentao, et al. Object detection based on randomized background modeling algorithm[J]. Journal of Applied Optics,2015,36(6):880-887. doi: 10.5768/JAO201536.0602001
[10] 戴航, 郭秀娟. 背景减除法在移动对象信息提取中的应用研究[J]. 吉林建筑大学学报,2020,37(4):81-84. doi: 10.3969/j.issn.1009-0185.2020.04.016 DAI Hang, GUO Xiujuan. Research on background substraction in moving target information extraction[J]. Journal of Jilin Jianzhu University,2020,37(4):81-84. doi: 10.3969/j.issn.1009-0185.2020.04.016
[11] 陈皓月, 钱钧, 姜文涛, 等. 一种基于粒子群优化的高斯混合灰度图像增强算法[J]. 应用光学,2017,38(4):592-598. CHEN Haoyue, QIAN Jun, JIANG Wentao, et al. Gaussian mixture grayscale image enhancement algorithm based on particle swarm optimization[J]. Journal of Applied Optics,2017,38(4):592-598.
[12] ELGAMMAL A M , HARWOOD D, DAVIS L S. Non-parametric model for background subtraction[C]//Proceeding of the European Conference on Computer Vision. Dublin: Springer, 2000: 751-767.
[13] BARNICH O, DROOGENBROECK M V. ViBE: A powerful random technique to estimate the background in video sequences[C]//Proceedings of 2009 IEEE International Conference on Acoustics, Speech and Signal Processing. USA: IEEE, 2009: 945-948.
[14] 王嘉俊, 段先华. 改进Canny算子在水面目标边缘检测中的研究[J]. 计算机时代,2020(1):35-38. WANG Jiajun, DUAN Xianhua. Research on improved Canny operator in edge detection of the target on water surface[J]. Computer Era,2020(1):35-38.
[15] 仝甄, 顾桂梅, 余晓宁. 弓网燃弧可见光图像分割算法[J]. 激光与光电子学进展,2020,57(16):202-208. TONG Zhen, GU Guimei, YU Xiaoning. Visible light image segmentation algorithm of pantograph-catenary arcing[J]. Laser & Optoelectronics Progress,2020,57(16):202-208.
[16] GOYETTE N, JODOIN P M, F PORIKLI, et al. Changedetection. net: a new change detection benchmark dataset[C]//Proceedings of IEEE Computer Society Conference on Computer Vision and Pattern Recognition Workshops. USA: IEEE, 2012: 1-8.
[17] LIU Ling, CHAI Guohua, QU Zhong. Moving target detection based on improved ghost suppression and adaptive visual background extraction[J]. Journal of Central South University,2021,28(3):747-759. doi: 10.1007/s11771-021-4642-9
-
期刊类型引用(7)
1. 王凯，李得睿，向升，程斌. 基于光斑投影3D-DIC的动态液面波高场测量方法研究. 力学学报. 2023(10): 2427-2438 . 
百度学术
2. 杜连续，金永. 小波变换轮廓术测量精度影响因素的研究. 机械与电子. 2022(03): 13-16 . 百度学术
3. 张申华，杨延西. 非静态物体的光栅图像投影3D测量方法. 电子测量与仪器学报. 2022(08): 158-166 . 百度学术
4. 夏桂书，吴虹星，魏永超，武兴焜. 旋转状态下的航空发动机叶片形变测量. 中国测试. 2022(12): 40-44 . 百度学术
5. 李雪，陶曾杰，雷琳. 傅里叶变换轮廓术在通信原理课程的教学应用. 实验室研究与探索. 2022(11): 140-144 . 百度学术
6. 夏桂书，武兴焜，魏永超，吴虹星. FTP动态测量航空发动机叶片三维型面. 中国测试. 2021(03): 30-35 . 百度学术
7. 朱荣刚，周健杰，张敏涛，陈鹏. 基于傅里叶变换迭代的条纹延拓方法研究. 金陵科技学院学报. 2021(03): 22-27 . 百度学术
其他类型引用(15)
计量
- 文章访问数: 417
- HTML全文浏览量: 142
- PDF下载量: 41
- 被引次数: 22
陕公网安备 61011302001501号 