[1]梁 磊,刘 怀,梁秦嘉,等.灰度序列图像中结合互相关法与梯度特征的阴影去除算法[J].南京师范大学学报(工程技术版),2019,19(02):059.[doi:10.3969/j.issn.1672-1292.2019.02.008]
　Liang Lei,Liu Huai,Liang Qinjia,et al.Shadow Removal Algorithm by Combining Cross-correlationand Gradient Feature in Gray Sequence Images[J].Journal of Nanjing Normal University(Engineering and Technology),2019,19(02):059.[doi:10.3969/j.issn.1672-1292.2019.02.008]
点击复制

灰度序列图像中结合互相关法与梯度特征的阴影去除算法
分享到：

南京师范大学学报（工程技术版）[ISSN:1006-6977/CN:61-1281/TN]

卷:
19卷
期数:
2019年02期
页码:
059
栏目:
计算机与信息工程
出版日期:
2019-06-30

文章信息/Info

Title:
Shadow Removal Algorithm by Combining Cross-correlationand Gradient Feature in Gray Sequence Images
文章编号:
1672-1292(2019)02-0059-09
作者:
梁 磊; 刘 怀; 梁秦嘉; 董春燕
南京师范大学电气与自动化工程学院,江苏 南京 210023
Author(s):
Liang Lei; Liu Huai; Liang Qinjia; Dong Chunyan
School of Electrical and Automation Engineering,Nanjing Normal University,Nanjing 210023,China
关键词:
阴影去除; 灰度序列图像; 灰度比值梯度; 归一化互相关
Keywords:
shadow removal; gray sequence images; gradient of grayscale ratio; normalized cross correlation(NCC)
分类号:
TP391
DOI:
10.3969/j.issn.1672-1292.2019.02.008
文献标志码:
A
摘要:
在视频监控场景下的目标检测中,运动的阴影会对所检测目标的准确性造成不利影响. 为了去除运动阴影的干扰,提出了一种结合灰度比值的四方向梯度与归一化互相关(NCC)特征的阴影去除算法. 首先在灰度序列图像中计算前景与其对应背景灰度的比值; 其次根据阴影区域的相邻像素灰度比值变化改变很小,通过计算灰度比值的四方向梯度均值来判断阴影并加以去除; 最后为了避免运动目标被误去除,考虑到运动目标与背景的相关性差异远远大于阴影与背景之间的相关性差异,结合归一化互相关特征来保留目标,以确保运动目标的完整性. 定性和定量的实验结果分析表明,该算法在阴影去除率和保持目标完整性方面优于其他阴影去除算法.
Abstract:
Moving shadows have adverse effects on the accuracy of the detected target while detecting moving target in video surveillance scene. In order to eliminate the interference of the moving shadows,a shadow removal algorithm is proposed by combining the four-direction gradient of the grayscale ratio and the feature of the normalized cross correlation(NCC)in this paper. Firstly,the grayscale ratio of the foreground pixel to its corresponding background pixel is calculated in gray sequence images. Secondly,because the grayscale ratio among a pixel’s neighborhood in shadow region has a small change,the calculated four-direction gradient mean can be applied to detect and remove shadow. Finally,in order to avoid that the moving target is removed as shadow and ensure the integrity of the detected target,the feature of the NCC is employed to keep the integrity of moving target because difference of correlation between the moving target and its corresponding background is much larger than that between shadow and its corresponding background. The qualitative and quantitative analysis of experimental results show that the algorithm presented in this paper is superior to other similar shadow removal algorithms in shadow removal ratio and keeping the integrity of the detected target.

参考文献/References:

[1] SANIN A,SANDERSON C,LOVELL B C. Shadow detection:a survey and comparative evaluation of recent methods[J]. Pattern recognition,2012,45(4):1684-1695.
[2]WANG Y,GONG N,GU X. Research on method for moving shadow detection[C]//International Conference on Intelligent Systems and Knowledge Engineering. Nanjing:IEEE,2017:1-4.
[3]段志刚,屈靓琼,田建东,等. 基于正交分解的室外光照阴影检测[J]. 光学学报,2016,36(8):201-209.
DUAN Z G,QU L Q,TIAN J D,et al. Outdoor illumination shadow detection based on orthogonal decomposition[J]. Acta optica sinica,2016,36(8):201-209.(in Chinese)
[4]MO N,ZHU R,YAN L,et al. Deshadowing of urban airborne imagery based on object-oriented automatic shadow detection and regional matching compensation[J]. IEEE journal of selected topics in applied earth observations and remote sensing,2018,11(2):585-605.
[5]GUO J M,HSIA C H,LIU Y F,et al. Fast background subtraction based on a multilayer codebook model for moving object detection[J]. IEEE transactions on circuits and systems for video technology,2013,23(10):1809-1821.
[6]钟小芳,周浩,高志山,等. 基于码本模型的运动阴影去除算法[J]. 计算机工程,2017,43(8):266-271.
ZHONG X F,ZHOU H,GAO Z S,et al. Moving shadow removal algorithm based on codebook model[J]. Computer engineering,2017,43(8):266-271.(in Chinese)
[7]QU L,TIAN J,FAN H,et al. Evaluation of shadow features[J]. Iet computer vision,2018,12(1):95-103.
[8]王燕玲,李广伦,林晓. 复杂动态环境下运动目标自动检测算法[J]. 系统仿真学报,2015,27(4):715-722.
WANG Y L,LI G L,LIN X. Method for auto-detection of tracking moving objects in complicated dynamic environment[J]. Journal of system simulation,2015,27(4):715-722.(in Chinese)
[9]KAR A,DEB K. Moving cast shadow detection and removal from video based on HSV color space[C]//International Conference on Electrical Engineering and Information Communication Technology. Dhaka:IEEE,2015:1-6.
[10]JIANG K,LI A H,CUI Z G,et al. Adaptive shadow detection using global texture and sampling deduction[J]. Iet computer vision,2013,7(2):115-122.
[11]韩超,邓甲昊,邹金慧,等. 基于差分均值背景提取和矩阵分区目标检测算法的研究[J]. 北京理工大学学报,2012,32(12):1247-1251,1257.
HAN C,DENG J H,ZOU J H,et al. Background extraction based on differential mean method and shadow detection using matrix subregion partition[J]. Transactions of Beijing institute of technology,2012,32(12):1247-1251,1257.(in Chinese)
[12]韩延彬,郭晓鹏,魏延文,等. RGB和HSI颜色空间的一种改进的阴影消除算法[J]. 智能系统学报,2015,10(5):769-774.
HAN Y B,GUO X P,WEI Y W,et al. An improved shadow removal algorithm based on RGB and HSI color spaces[J]. CAAI transactions on intelligent systems(TIS),2015,10(5):769-774.(in Chinese)
[13]FAROU B,ROUABHIA H,SERIDI H,et al. Novel approach for detection and removal of moving cast shadows based on RGB,HSV and YUV color spaces[J]. Computing and informatics,2017,36(4):837-856.
[14]RUSSELL M,ZOU J J,FANG G. Real-time vehicle shadow detection[J]. Electronics letters,2015,51(16):1253-1255.
[15]RUSSELL M,ZOU J J,FANG G,et al. Feature-based image patch classification for moving shadow detection[J]. IEEE transactions on circuits and systems for video technology,2017,14(8):1-15.
[16]刘艳丽,吴彧,陈祥祥,等. 室外移动视点视频的在线阴影边缘检测[J]. 计算机辅助设计与图形学学报,2018,30(10):1827-1834.
LIU Y L,WU Y,CHEN X X,et al. Online detection of outdoor shadow edges from live videos under moving viewpoints[J]. Journal of computer-aided design and computer graphics,2018,30(10):1827-1834.(in Chinese)
[17]CHEN Z,ZHAO Y,HUANG X,et al. An improved shadow removal algorithm based on gradient amendment[C]//International conference on signal processing. Hangzhou:IEEE,2014:1190-1194.
[18]戴璐平,刘海英,郑宽磊. 结合局部二元图特征的运动目标阴影抑制方法[J]. 华中科技大学学报(自然科学版),2016,44(10):119-122.
DAI L P,LIU H Y,ZHENG K L. Shadow suppression method for moving object based on the characteristics of binary pattern[J]. Journal of Huazhong university of science and technology(nature science edition),2016,44(10):119-122.(in Chinese)
[19]XU M,ZHU J,Lü P,et al. Learning-based shadow recognition and removal from monochromatic natural images[J]. IEEE transactions on image processing,2017,26(12):5811-5824.
[20]袁博,阮秋琦,安高云. 改进的自适应灰度视频序列阴影检测方法[J]. 信号处理,2014(11):1370-1374.
YUAN B,RUAN Q Q,AN G Y. Improved adaptive shadow detection approach in grayscale video sequences[J]. Journal of signal processing,2014(11):1370-1374.(in Chinese)
[21]韩延祥,张志胜,郝飞,等. 灰度序列图像中基于纹理特征的移动阴影检测[J]. 光学精密工程,2013,21(11):2931-2942.
HAN Y X,ZHANG Z S,HAO F,et al. Shadow detection based on texture features in gray sequence images[J]. Optics and precision engineering,2013,21(11):2931-2942.(in Chinese)
[22]邢藏菊,温兰兰,何苏勤. TLD视频目标跟踪器快速匹配的研究[J]. 小型微型计算机系统,2015,36(5):1113-1116.
XING C J,WEN L L,HE S Q. Research on fast matching based TLD video target tracking[J]. Journal of Chinese computer systems,2015,36(5):1113-1116.(in Chinese)
[23]武明虎,宋冉冉,刘敏. 结合HSV与纹理特征的视频阴影消除算法[J]. 中国图象图形学报,2017,22(10):1373-1380.
WU M H,SONG R R,LIU M. Video shadow elimination algorithm by combining HSV with texture features[J]. Journal of image and graphics,2017,22(10):1373-1380.(in Chinese).

备注/Memo

备注/Memo:
收稿日期:2018-10-19.
基金项目:国家自然科学基金(61603194).
通讯联系人:刘怀,博士,副教授,研究方向:数字图像处理、实时控制系统. E-mail:liuhuai@njnu.edu.cn
更新日期/Last Update: 2019-06-30