高级搜索

留言板

尊敬的读者、作者、审稿人, 关于本刊的投稿、审稿、编辑和出版的任何问题, 您可以本页添加留言。我们将尽快给您答复。谢谢您的支持!

姓名
邮箱
手机号码
标题
留言内容
验证码

一种新的图像超像素分割方法

廖苗 李阳 赵于前 刘毅志

廖苗, 李阳, 赵于前, 刘毅志. 一种新的图像超像素分割方法[J]. 电子与信息学报, 2020, 42(2): 364-370. doi: 10.11999/JEIT190111
引用本文: 廖苗, 李阳, 赵于前, 刘毅志. 一种新的图像超像素分割方法[J]. 电子与信息学报, 2020, 42(2): 364-370. doi: 10.11999/JEIT190111
Miao LIAO, Yang LI, Yuqian ZHAO, Yizhi LIU. A New Method for Image Superpixel Segmentation[J]. Journal of Electronics & Information Technology, 2020, 42(2): 364-370. doi: 10.11999/JEIT190111
Citation: Miao LIAO, Yang LI, Yuqian ZHAO, Yizhi LIU. A New Method for Image Superpixel Segmentation[J]. Journal of Electronics & Information Technology, 2020, 42(2): 364-370. doi: 10.11999/JEIT190111

一种新的图像超像素分割方法

doi: 10.11999/JEIT190111
基金项目: 国家自然科学基金(61702179, 61772555),湖南省自然科学基金(2017JJ3091),中国博士后科学基金(2018M632994),湖南省教育厅资助科研项目(17C0643)
详细信息
    作者简介:

    廖苗:女,1988年生,博士,讲师,硕士生导师,研究方向为数字图像处理、图像分割、模式识别

    李阳:女,1993年生,博士,研究方向为数字图像处理,图像分割

    赵于前:男,1973年生,博士,教授,博士生导师,研究方向为数字图像处理、模式识别、视频处理、信息安全等

    刘毅志:男,1973年生,博士,副教授,硕士生导师,研究方向为数字图像处理、多媒体内容分析与检索

    通讯作者:

    廖苗 liaomiaohi@163.com

  • 中图分类号: TP391.41

A New Method for Image Superpixel Segmentation

Funds: The National Natural Science Foundation of China (61702179, 61772555), The Hunan Provincial Natural Science Foundation of China (2017JJ3091), The Postdoctoral Science Foundation Funded Project of China (2018M632994), The Scientific Research Fund of Hunan Provincial Education Department (17C0643)
  • 摘要:

    针对现有超像素分割方法无法自动确定合适的超像素数目,以及难以有效贴合图像目标边界等问题,该文提出一种新的利用局部信息进行多层级简单线性迭代聚类的图像超像素分割方法。首先,运用基于局部信息的简单线性迭代聚类(LI-SLIC)对原始图像进行超像素初分割,然后,根据超像素的色彩标准差对其进行自适应多层级迭代分割,直至每个超像素块的色彩标准差小于预设阈值,最后,利用相邻超像素间的色彩差异对过分割的超像素进行合并。为验证方法的有效性,该文采用Berkeley, Pascal VOC和3Dircadb公共数据库作为实验数据集,并与其他多种超像素分割方法进行了比较。实验结果表明,该文提出的超像素分割方法能更精确贴合图像目标边界,有效抑制图像过分割和欠分割。

  • 图  1  本文算法流程图

    图  2  超像素多层级迭代分割示例

    图  3  Berkeley数据库的部分实验结果比较

    图  4  PASCAL VOC数据库的部分实验结果比较

    图  5  CT图像实验结果比较

    表  1  Berkeley数据库超像素分割结果评价(均值±标准差)

    方法BR(%)UE(%)超像素数目
    SLIC[9]83.17±9.856.17±3.16441±272
    ASLIC[9]69.62±12.287.69±3.64438±267
    SNIC[10]83.52±10.806.85±4.20468±311
    本文方法86.25±7.826.13±3.24443±243
    下载: 导出CSV

    表  2  Pascal VOC数据库超像素分割结果评价(均值±标准差)

    方法BR(%)UE(%)超像素数目
    SLIC[9]85.56±7.372.80±1.72423±239
    ASLIC[9]82.34±8.112.94±1.80418±236
    SNIC[10]85.55±7.372.86±1.73421±238
    本文方法87.95±7.202.57±1.75420±241
    下载: 导出CSV

    表  3  3Dircadb数据库超像素分割结果评价(均值±标准差)

    方法BR(%)UE(%)超像素数目
    SLIC[9]90.97±6.940.50±0.28526±78
    ASLIC[9]88.14±8.970.55±0.31469±70
    SNIC[10]90.17±7.500.64±0.42521±74
    本文方法93.82±8.060.41±0.34521±73
    下载: 导出CSV
  • WANG Murong, LIU Xiabi, GAO Yixuan, et al. Superpixel segmentation: A benchmark[J]. Signal Processing: Image Communication, 2017, 56: 28–39. doi: 10.1016/j.image.2017.04.007
    ZHOU Xianen, WANG Yaonan, ZHU Qing, et al. SSG: Superpixel segmentation and GrabCut-based salient object segmentation[J]. The Visual Computer, 2019, 35(3): 385–398. doi: 10.1007/s00371-018-1471-4
    SHAO Hong, YU Tianshu, XU Mengjia, et al. Image region duplication detection based on circular window expansion and phase correlation[J]. Forensic Science International, 2012, 222(1/3): 71–82.
    TIAN Zhiqiang, LIU Lizhi, ZHANG Zhenfeng, et al. Superpixel-based segmentation for 3D prostate MR images[J]. IEEE Transactions on Medical Imaging, 2016, 35(3): 791–801. doi: 10.1109/TMI.2015.2496296
    WANG Jun, LIU Weibin, XING Weiwei, et al. Visual object tracking with multi-scale superpixels and color-feature guided kernelized correlation filters[J]. Signal Processing: Image Communication, 2018, 63: 44–62. doi: 10.1016/j.image.2018.01.005
    LI Zhengqin and CHEN Jiansheng. Superpixel segmentation using linear spectral clustering[C]. IEEE Conference on Computer Vision and Pattern Recognition, Boston, USA, 2015: 1356–1363.
    SHI Jianbo and MALIK J. Normalized cuts and image segmentation[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2000, 22(8): 888–905. doi: 10.1109/34.868688
    GIRAUD R, TA V T, and PAPADAKIS N. Robust superpixels using color and contour features along linear path[J]. Computer Vision and Image Understanding, 2018, 170: 1–13. doi: 10.1016/j.cviu.2018.01.006
    ACHANTA R, SHAJI A, SMITH K, et al. SLIC superpixels compared to state-of-the-art superpixel methods[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2012, 34(11): 2274–2281. doi: 10.1109/TPAMI.2012.120
    ACHANTA R and SÜSSTRUNK S. Superpixels and polygons using simple non-iterative clustering[C]. IEEE Conference on Computer Vision and Pattern Recognition, Honolulu, USA, 2017: 4895–4904.
    VASQUEZ D and SCHARCANSKI J. An iterative approach for obtaining multi-scale superpixels based on stochastic graph contraction operations[J]. Expert Systems with Applications, 2018, 102: 57–69. doi: 10.1016/j.eswa.2018.02.027
    CONZE P H, NOBLET V, ROUSSEAU F, et al. Scale-adaptive supervoxel-based random forests for liver tumor segmentation in dynamic contrast-enhanced CT scans[J]. International Journal of Computer Assisted Radiology and Surgery, 2017, 12(2): 223–233. doi: 10.1007/s11548-016-1493-1
    MARTIN D, FOWLKES C, TAL D, et al. A database of human segmented natural images and its application to evaluating segmentation algorithms and measuring ecological statistics[C]. IEEE International Conference on Computer Vision, Vancouver, Canada, 2001: 416–423.
    EVERINGHAM M, ESLAMI S M A, VAN GOOL L, et al. The PASCAL visual object classes challenge: A retrospective[J]. International Journal of Computer Vision, 2015, 111(1): 98–136. doi: 10.1007/s11263-014-0733-5
    SOLER L, HOSTETTLER A, AGNUS V, et al. 3D image reconstruction for comparison of algorithm database: A patient-specific anatomical and medical image database[EB/OL]. http://www-sop.inria.fr/geometrica/events/wam/abstract-ircad.pdf.
  • 加载中
图(5) / 表(3)
计量
  • 文章访问数:  4692
  • HTML全文浏览量:  2100
  • PDF下载量:  205
  • 被引次数: 0
出版历程
  • 收稿日期:  2019-02-26
  • 修回日期:  2019-09-03
  • 网络出版日期:  2019-09-20
  • 刊出日期:  2020-02-19

目录

    /

    返回文章
    返回