首页 > 论文 > 半导体光电 > 40卷 > 4期(pp:564-570)

基于全卷积神经网络和动态自适应区域生长法的红外图像目标分割方法

Infrared Image Target Segmentation Algorithm Based on Full Convolutional Neural Network and Dynamic Adaptive Region Growth

  • 摘要
  • 论文信息
  • 参考文献
  • 被引情况
  • PDF全文
分享:

摘要

复杂背景下的红外图像往往由于噪声较多、背景区域重叠、目标与背景对比度较差等因素, 在对目标区域分割时会造成过分割或欠分割。针对此现象, 提出了一种将全卷积神经网络和动态自适应区域生长法相结合的红外分割算法。首先利用全卷积神经网络对目标区域在像素级别进行特征提取, 通过神经网络强大的自学习能力获得目标区域的粗分割结果; 然后根据粗分割结果, 对其取外接最小面积矩形框, 并根据矩形框位置在原始图像上确定目标区域, 并以此矩形区域进行动态自适应区域生长, 形成第二次分割结果。最后融合全卷积网络(FCN)的粗分割结果和区域生长分割结果, 实现目标区域的最终分割和提取。仿真实验表明, 该方法能有效利用FCN对红外图像复杂背景的消除能力, 而区域生长法对分割细节的敏感也同时弥补了FCN分割精度的不足, 取得了较好的分割效果。

Abstract

Infrared images in complex backgrounds tend to be over-segmented or under-segmented when segmenting the target region due to factors such as high noise, overlapping background regions, and poor target and background contrast. Aiming at this phenomenon, it proposes an infrared segmentation algorithm combining the full convolutional neural network and the dynamic adaptive region growing method. Firstly, the full-convolution neural network is used to extract the feature of the target region at the pixel level, and the coarse segmentation result of the target region is obtained by the powerful self-learning ability of the neural network. Then, according to the result of the coarse segmentation, the minimum area rectangular frame is taken outside, and according to the position of the rectangular frame, the target area is determined on the original image, and the dynamic adaptive area growth is performed by using the rectangular area to form a second segmentation result. Finally, the rough segmentation results and regional growth results of the FCN are combined to achieve the final segmentation and extraction of the target region. The simulation experiments show that the method can effectively utilize the FCN to eliminate the complex background of infrared images, and the sensitivity of the region growing method to the segmentation details also compensates for the lack of FCN segmentation accuracy and achieves a good segmentation effect.

广告组1 - 空间光调制器+DMD
补充资料

中图分类号:TP391

DOI:10.16818/j.issn1001-5868.2019.04.023

所属栏目:光电技术及应用

基金项目:山西水利职业技术学院课题项目(GH-17141).

收稿日期:2019-02-27

修改稿日期:--

网络出版日期:--

作者单位    点击查看

任志淼:山西水利职业技术学院, 太原 030027

联系人作者:任志淼(sqjgeo@sina.com)

备注:任志淼(1971-), 女, 山西省平遥人, 硕士, 讲师, 研究方向为控制理论与控制工程。

【1】Yen J C. A new criterion for automatic multilevel thresholding[J]. IEEE Trans. Image Process, 1995, 4(3): 370-378.

【2】Huang L K, Wang M J. Image thresholding by minimizing the measures of fuzziness[J]. Pattern Recognition, 1995, 28(1): 41-51.

【3】高 敏, 李怀胜, 周玉龙, 等. 背景约束的红外复杂背景下坦克目标分割方法[J]. 自动化学报, 2016, 42(3): 416-430.
Gao Min, Li Huaisheng, Zhou Yulong, et al. Tank segmentation under infrared complex background with background restriction[J]. Acta Automatica Sinica, 2016, 42(3): 416-430.

【4】龙建武, 申铉京, 臧 慧, 等. 高斯尺度空间下估计背景的自适应阈值分割算法[J]. 自动化学报, 2014, 40(8): 1773-1782.
Long Jianwu, Shen Xuanjing, Zang Hui, et al. An adaptive thresholding algorithm by background estimation in Gaussian scale space[J]. Acta Automatica Sinica, 2014, 40(8): 1773-1782.

【5】Adams R, Bischof L. Seeded region growing[J]. IEEE Trans. on Pattern Analysis & Machine Intelligence, 2002, 16(6): 641-647.

【6】李启翮, 罗予频, 萧德云. 基于流向标量场与快速扫掠法的图像分割[J]. 自动化学报, 2008, 34(8): 993-996.
Li Qihe, Luo Yupin, Xiao Deyun. Flow direction scalar field and fast scanning method based image segmentation[J]. Acta Automatica Sinica, 2008, 34(8): 993-996.

【7】孙超男, 易 芹, 崔 丽. 小波变换结合模糊聚类在示温漆彩色图像分割中的应用[C]// 全国智能CAD与数字娱乐学术会议, 2012.
Sun Chaonan, Yi Qin, Cui Li. Application of wavelet transform combined with fuzzy clustering in color image segmentation of temperature indicating paint[C]// National Academic Conf. on Intelligent CAD and Digital Entertainment, 2012.

【8】肖春霞, 初 雨, 张 青. 高斯混合函数区域匹配引导的Level Set纹理图像分割[J]. 计算机学报, 2010(7): 1295-1304.
Xiao Chunxia, Chu Yu, Zhang Qing. Texture image segmentation using level set function evolved by Gaussian mixture model[J]. Chinese J. of Computers, 2010(7): 1295-1304.

【9】张迎春, 郭 禾. 基于粗糙集和新能量公式的水平集图像分割[J]. 自动化学报, 2015, 41(11): 1913-1925.
Zhang Yingchun, Guo He. Level set image segmentation based on rough set and new energy formula[J]. Acta Automatica Sinica, 2015, 41(11): 1913-1925.

【10】宋艳涛, 纪则轩, 孙权森. 基于图像片马尔科夫随机场的脑MR图像分割算法[J]. 自动化学报, 2014, 40(8): 1754-1763.
Song Yantao, Ji Zexuan, Sun Quansen. Brain MR image segmentation algorithm based on Markov random field with image patch[J]. Acta Automatica Sinica, 2014, 40(8): 1754-1763.

【11】Krizhevsky A, Sutskever I, Hinton G. ImageNet classification with deep convolutional neural networks[C]// NIPS. Curran Associates Inc., 2012.

【12】Bai X, Wang W. Saliency-SVM: An automatic approach for image segmentation[J]. Neurocomputing, 2014, 136: 243-255.

【13】Long J, Shelhamer E, Darrell T. Fully convolutional networks for semantic segmentation[J]. IEEE Trans. on Pattern Analysis & Machine Intelligence, 2014, 39(4): 640-651.

【14】Chen L C, Papandreou G, Kokkinos I, et al. Semantic image segmentation with deep convolutional nets and fully connected CRFs[J]. Computer Science, 2014(4): 357-361.

【15】Kamnitsas K, Ledig C, Newcombe V F J, et al. Efficient multi-scale 3D CNN with fully connected CRF for accurate brain lesion segmentation[J]. Medical Image Analysis, 2016: S1361841516301839.

【16】Malarvel M, Sethumadhavan G , Bhagi P C R, et al. Region growing based segmentation with automatic seed selection using threshold techniques on X-radiography images[C]// 2016 IEEE Inter. Conf. on Computational Intelligence and Computing Research (ICCIC), 2017.

【17】翟伟明, 胡成文, 张伟宏, 等. 基于动态自适应体素生长的肺部CT图像3维分割算法[J]. 中国图象图形学报, 2005, 10(10): 1269-1274.
Zhai Weiming, Hu Chengwen, Zhang Weihong, et al. A dynamic adaptive 3D voxel-growing segmentation algorithm for pulmonary CT images[J]. J. of Image and Graphics, 2005, 10(10): 1269-1274.

【18】宋 红, 王 勇, 黄小川, 等. 基于动态自适应区域生长的肝脏CT图像肿瘤分割算法[J]. 北京理工大学学报, 2014, 34(1): 72-76.
Song Hong, Wang Yong, Huang Xiaochuan, et al. A dynamic adaptive region growing segmentation algorithm for tumor of liver CT images[J]. Trans. of Beijing Institute of Technol., 2014, 34(1): 72-76.

【19】Davis J W, Keck M A. A two-stage template approach to person detection in thermal imagery[C]// Proc. of the 7th IEEE Workshops on Applications of Computer Vision, 2005: 364-369.

引用该论文

REN Zhimiao. Infrared Image Target Segmentation Algorithm Based on Full Convolutional Neural Network and Dynamic Adaptive Region Growth[J]. Semiconductor Optoelectronics, 2019, 40(4): 564-570

任志淼. 基于全卷积神经网络和动态自适应区域生长法的红外图像目标分割方法[J]. 半导体光电, 2019, 40(4): 564-570

您的浏览器不支持PDF插件,请使用最新的(Chrome/Fire Fox等)浏览器.或者您还可以点击此处下载该论文PDF