背景噪声影响下实景图像局部遮挡目标全轮廓提取算法

doi:10.13229/j.cnki.jdxbgxb.20240970

摘要/Abstract

摘要：

在实景图像中，当目标物体受其他物体或环境元素的遮挡时，在遮挡区域附近噪声会使得背景与目标物体的轮廓特征混淆，仅能分析可见部分的轮廓特征，推断出具有连续性和平滑性的被遮挡区域轮廓变得尤为困难，导致提取的前景目标全轮廓与原始目标边界之间出现偏差。为此，提出背景噪声影响下实景图像局部遮挡目标全轮廓提取算法。通过混合高斯背景差分方法构建实景图像背景模型，用于区分实景图像内的背景图像与前景目标图像，剔除背景噪声信息，获取前景目标图像。结合多分辨率方法与主动轮廓模型，并引入梯度矢量流外力场，提取前景目标图像中目标的轮廓；利用双弧插值算法，针对前景目标轮廓中被遮挡区域，通过平滑轮廓修复和角点轮廓修复，获取连续且平滑地前景目标全轮廓，确保修复后的轮廓与原始轮廓在视觉上保持一致，最小化与原始目标边界之间的差异。实验结果显示：本文算法对非背景元素的误判率始终保持在0.1%以下；轮廓提取过程中生成的轮廓曲线与目标边界高度吻合，且轮廓线条流畅连续；轮廓修复效果极其自然且逼真，误差值仅为0.02%。

关键词: 背景噪声, 实景图像, 局部遮挡, 双弧插值, 全轮廓提取, 轮廓修复

Abstract:

In the real image， when the target object is occluded by other objects or environmental elements， the noise near the occluded area will confuse the contour features of the background and the target object. It is particularly difficult to infer the contour of the occluded area with continuity and smoothness by analyzing only the contour features of the visible part. This results in a deviation between the extracted foreground target contour and the original target boundary. Therefore， a full-contour extraction algorithm for local occluding object in real image is proposed under the influence of background noise. The background model of the real image is constructed by the mixed Gaussian background difference method， which is used to distinguish the background image and the foreground target image in the real image， remove the background information， and obtain the foreground target image. Multi-resolution method is combined with ACM and GVF field is introduced to extract the contour of the foreground target image. The double-arc interpolation algorithm is used to obtain the full contour of the foreground target continuously and smoothly through smooth contour repair and corner contour repair， aiming at the blocked area in the foreground target contour. The repaired contour is visually consistent with the original contour， and the difference between the boundary and the original target is minimized. The experimental results show that the misjudgment rate for non-background elements is always below 0.1%； The contour curves generated in the process of contour extraction are highly consistent with the target boundary， and the contour lines are smooth and continuous. The contour restoration results are extremely natural and realistic， with an error value of only 0.02%.

Key words: background noise, realistic images, partial occlusion, double arc interpolation, full contour extraction, contour repair

中图分类号:

TP391

全蕾,陈杰. 背景噪声影响下实景图像局部遮挡目标全轮廓提取算法[J]. 吉林大学学报(工学版), 2025, 55(10): 3329-3336.

Lei QUAN,Jie CHEN. Full contour extraction algorithm for partially occluded targets in real-world images under influence of background noise[J]. Journal of Jilin University(Engineering and Technology Edition), 2025, 55(10): 3329-3336.

图/表 8

图1

表1

表2

图2

图3

图4

图5

表3

参考文献 15

[1]	庹武, 杜聪, 陈谦, 等. 基于计算机视觉与Canny算法的服装纸样轮廓提取[J]. 纺织学报, 2024, 45(5): 174-182.
	Wu Tuo, Du Cong, Chen Qian,et al.Clothing pattern contour extraction based on computer vision and Canny algorithm[J].Journal of Textile Research,2024,45(5):174-182.
[2]	王一超, 鲁芹, 吴孟伟. 结合前景轮廓提取的改进ViBe运动目标检测算法[J]. 微电子学与计算机, 2023, 40(8): 37-44.
	Wang Yi-chao, Lu Qin, Wu Meng-wei.Accurate and efficient moving object detection with ViBe and visual foreground contour extractor[J].Microelectronics & Computer, 2023, 40(8): 37-44.
[3]	王鑫刚, 田军委, 于亚琳, 等. 改进Canny算法的红外人脸图像边缘轮廓提取[J]. 应用光学, 2023, 44(1): 61-70.
	Wang Xin-gang, Tian Jun-wei, Yu Ya-lin,et al.Edge contour extraction of infrared face image based on improved Canny algorithm[J].Journal of Applied Optics,2023,44(1):61-70.
[4]	程彭阳子, 刘萍. 基于动态轮廓曲线的运动图像特征识别研究[J]. 计算机仿真, 2023, 40(7): 163-167.
	Yang-zi Chengpeng, Liu Ping. Research on moving image feature recognition based on dynamic contour curve[J].Computer Simulation,2023,40(7):163-167.
[5]	Chakraborty S, Shaikh S H, Chakrabarti A,et al.Quantum image edge extraction based on classical robinson operator[J].Multimedia Tools and Applications, 2022, 81(23): 33459-33481.
[6]	杨卫东, 叶长彬, 陈正林, 等. 基于Snake算法的声呐图像轮廓提取方法[J]. 压电与声光, 2023, 45(5): 752-758.
	Yang Wei-dong, Ye Chang-bin, Chen Zheng-lin,et al. Image contour extraction method based on Snake algorithm[J].Piezoelectrics & Acoustooptics,2023,45(5): 752-758.
[7]	徐旸, 史金光, 郑子玙, 等. 融合图像修复的遮挡目标检测算法[J]. 电光与控制, 2023, 30(1): 21-28, 86.
	Xu Yang, Shi Jin-guang, Zheng Zi-yu,et al.Occluded target detection algorithm with image restoration[J].Electronics Optics & Control, 2023, 30(1): 21-28, 86.
[8]	李阿标, 郭浩, 戚畅, 等. 复杂背景下遥感图像密集目标检测[J]. 计算机工程与应用, 2023, 59(8): 247-253.
	Li A-biao, Guo Hao, Qi Chang,et al.Dense object detection in remote sensing images under complex background[J].Computer Engineering and Applications, 2023, 59(8): 247-253.
[9]	娄联堂, 姚松. 基于局部轮廓曲线的图像配准方法[J].中南民族大学学报: 自然科学版, 2023, 42(1): 128-133.
	Lou Lian-tang, Yao Song.Image registration method based on local contour curve[J].Journal of South-Central University for Nationalities(Natural Science Edition), 2023, 42(1): 128-133.
[10]	段琳锋, 侯新国, 胡致远. NSCT轮廓与主方向一致性红外与可见光图像配准[J]. 电光与控制, 2022, 29(6): 1-5.
	Duan Lin-feng, Hou Xin-guo, Hu Zhi-yuan.NSCT Contour and main direction consistency infrared and visible image registration[J].Electronics Optics & Control, 2022, 29(6):1-5.
[11]	熊丽丽, 冯丽佳, 苑柳青. 基于深度卷积神经网络的激光雷达图像轮廓线提取方法[J]. 激光杂志,2023, 44(10): 94-99.
	Xiong Li-li, Feng Li-jia, Yuan Liu-qing.LiDAR Image contour extraction method based on deep convolution neural network[J]. Laser Journal, 2023, 44(10):94-99.
[12]	游小荣, 李淑芳, 熊宗志. 基于卷积姿态机和GrabCut的服装图像轮廓分割方法[J]. 毛纺科技, 2022, 50(5): 86-90.
	You Xiao-rong, Li Shu-fang, Xiong Zong-zhi.Garment image contour segmentation method based on convolution pose machine and GrabCut[J].Wool Textile Journal, 2022, 50(5): 86-90.
[13]	冯丽露, 康耀龙. 立体图像任意剖面轮廓线提取方法仿真研究[J]. 计算机仿真, 2022, 39(8): 239-242, 285.
	Feng Li-lu, Kang Yao-long.Simulation study of arbitrary profile contour extraction method[J]. Computer Simulation, 2022, 39(8): 239-242, 285.
[14]	姚昱, 郑洋, 程子燚, 等. 基于Scharr算子的计算鬼成像边缘提取技术[J]. 激光与光电子学进展, 2024, 61(10): 137-144.
	Yao Yu, Zheng Yang, Cheng Zi-yi.Edge detection technology based on computational ghost imaging using Scharr operator[J].Laser & Optoelectronics Progress, 2024, 61(10): 137-144.
[15]	赵亚威, 郑伟, 于洋, 等. 基于轮廓分段特征描述的遮挡目标识别算法[J]. 计算机应用研究, 2022, 39(2): 633-636.
	Zhao Ya-wei, Zheng Wei, Yu Yang.Occluded object recognition algorithm based on contour segmented feature description[J]. Application Research of Computers, 2022, 39(2): 633-636.

Metrics

Viewed

Full text

Abstract

Cited

Shared

Discussed

参数	数值
图像数量/幅	800
帧数/Hz	300
帧率/fps	25
分辨率/像素	500×300
采样时间/h	4
采样平均间隔时间/s	5
图像灰度平均值	0.5
图像饱和度	-26

图像帧数/帧	情况1	情况2
5	0.038	0.042
10	0.049	0.065
15	0.061	0.072
20	0.053	0.062
25	0.058	0.059
30	0.047	0.048
35	0.064	0.069
40	0.071	0.075

实景图像样本数/幅	平均相对误差/%	ViBe算法	Canny算法
100	0.02	2.35	2.03
200	1.70	3.65	2.45
300	2.31	3.25	2.36
400	2.10	3.45	2.56
500	1.89	3.28	2.31
600	1.87	3.64	2.58
700	2.03	2.95	2.59
800	2.12	2.89	2.37