自适应复杂背景干扰的运动目标检测算法

王东京; 张宝辉; 陈弘原; 王润宇; 吴杰; 吴旭东

红外技术, 2017, 39 (11): 1024, 网络出版: 2017-11-27

自适应复杂背景干扰的运动目标检测算法

Moving-target Detection Algorithm Adapting Complex Background Interference

论文大纲

王东京 ¹张宝辉 ²陈弘原 ¹王润宇 ²吴杰 ²吴旭东 ²

作者单位

¹ 南京理工大学电子工程与光电技术学院，江苏南京 210094

² 北方夜视科技集团有限公司南京研发中心，江苏南京 211106

运动目标检测复杂背景模型更新自适应阈值 moving target detection complex background updating model adaptive threshold

摘要

基于视频图像的运动目标检测，是根据目标的像素特征来判别出相对于背景运动的目标，当图像背景动态变化时，将难以区分背景和运动目标的像素特征，易造成检测错误。复杂背景下的运动目标检测是一大难点，目前主流的运动目标检测算法在背景灰暗、水面波动、气流颤动等复杂背景干扰下，难以准确地检测出运动目标。针对上述问题，提出一种自适应复杂背景干扰的运动目标检测算法，采用新的前景判断和背景模型更新方法，同时设计了一种创新型自适应阈值更新方法，当视频背景变化时，自动更新阈值。该算法增强了对复杂背景、镜头抖动的抗干扰能力，通过各种视频测试，背景点检测正确率达到0.9958，前景点检测正确率达到0.8012，极大提高了前景检测率，而且该算法满足高实时性要求，对复杂背景下的运动目标检测有显著效果。

Abstract

Video-based moving-object detection is according to the pixel characteristics of the target to determine the target that is moving relative to the background. When the background of the image changes dynamically, it will be difficult to distinguish the pixel characteristics between the background and the moving target, which is prone to detection error. Moving target detection in a complex background has been a challenging problem until now. The mainstream algorithm of moving target detection still cannot accurately detect moving targets under complex backgrounds such as dark background, water surface fluctuation, and airflow quiver. Aiming at the above problems, a novel moving-target detection algorithm was proposed for adapting to complex background interference. The algorithm adopted a new standard of classifying the foreground and a new method of updating the background model. At the same time, a creative method was put forward to update the threshold, which would automatically adjust the size of the threshold according to the change of background. The algorithm enhanced the ability to resist the interference of the complex background and camera shake. According to various video tests, the accuracy rate of background detection reached 0.9958.On the other hand, the accuracy rate of foreground detection reached 0.8012. Obviously, it is a great progress on foreground detection. Besides, the novel algorithm has high work-efficiency and has a significant effect on the extraction of moving objects in a complex background.

参考文献

[1] 杨智雄,余春超,严敏,等.基于特征融合的粒子滤波红外目标跟踪算法[J].红外技术,2016,38(3):211-217.

YANG Zhixiong, YU Chunchao, YAN Min, et al. Particle Filter Infrared Target Tracking Algorithm Based on Feature Fusion[J]. Infrared Technology, 2016, 38(3): 211-217.

[2] 李倩倩,刘彦隆.先验信息光流法在运动目标检测中的应用[J].火力与指挥控制,2015(10):156-160.

LI Qianqian, LIU Yanlong. Based on Optical Flow Method with Priori Information in the Application of Detecting Moving Target Tracking[J]. Fire Control & Command Control, 2015(10):156-160.

[3] Shen Y, Hu W, Yang M, et al. Real-time and Robust Compressive Background Subtraction for Embedded Camera Networks[J]. IEEE Transactions on Mobile Computing, 2016, 15(2): 406-418.

[4] Hu X, Zheng J. An Improved Moving Object Detection Algorithm Based on Gaussian Mixture Models[J]. Open Journal of Applied Sciences, 2016, 6(7): 449-456.

[5] Zeng Z. Moving Object Extraction Using the Improved Adaptive Gaussian Mixture Model and Shadow Detection Model[J]. Journal of Information & Computational Science, 2015, 12(14): 5515-5522.

[6] Hofmann M, Tiefenbacher P, Rigoll G. Background segmentation with feedback: The Pixel-Based Adaptive Segmenter[C]//Computer Vision and Pattern Recognition Workshops. IEEE, 2012: 38-43.

[7] Wang H, Suter D. Background Subtraction Based on a Robust Consensus Method[C]//International Conference on Pattern Recognition, IEEE, 2006: 223-226.

[8] Barnich O, Van D M. ViBe: a universal background subtraction algorithm for video sequences[J]. IEEE Transactions on Image Processing A Publication of the IEEE Signal Processing Society, 2011, 20(6): 1709-1724.

[9] 孙丰,秦开怀,孙伟,等.一种针对移动相机的实时视频背景减除算法[J].计算机辅助设计与图形学学报,2016,28(4):572-578.

Sun Feng, Qin Kaihuai, Sun Wei, et al. A Real-Time Background Subtraction Algorithm for Freely Moving Cameras[J]. Journal of Computer-Aided Design & Computer Graphics, 2016, 28(4): 572-578.

[10] Zivkovic Z, Heijden F V D. Efficient adaptive density estimation per image pixel for the task of background subtraction[J]. Elsevier, 2006, 27(7): 773-780.

[11] 解婷,陈忠,马荣毅.一种基于PGF、BEMD和局部逆熵的新型红外小目标检测方法[J].红外与毫米波学报,2017,36(1):92-101.

Xie Ting, Chen Zhong, Ma Rongyi. A novel method for infrared small target detection based on PGF, BEMD and LIE[J]. J. Infrared Millim. Waves, 2017, 36(1): 92-101.

[12] N. Goyette, P.-M. Jodoin, F. Porikli, et al. Change detection.net: A new change detection benchmark dataset[C]//Proc. IEEE Workshop on Change Detection (CDW-2012) at CVPR-2012, 2012: 16-21.

王东京, 张宝辉, 陈弘原, 王润宇, 吴杰, 吴旭东. 自适应复杂背景干扰的运动目标检测算法[J]. 红外技术, 2017, 39(11): 1024. WANG Dongjing, ZHANG Baohui, CHEN Hongyuan, WANG Runyu, WU Jie, WU Xudong. Moving-target Detection Algorithm Adapting Complex Background Interference[J]. Infrared Technology, 2017, 39(11): 1024.

自适应复杂背景干扰的运动目标检测算法

关于本站 Cookie 的使用提示

全站搜索

自适应复杂背景干扰的运动目标检测算法

相关论文

相关资讯

关于本站 Cookie 的使用提示

全站搜索