文章摘要
曾步衢.一种改进的图像低维表示方法[J].包装工程,2017,38(9):230-235.
ZENG Bu-qu.An Improved Method of Low-Dimensional Representation of Images[J].Packaging Engineering,2017,38(9):230-235.
一种改进的图像低维表示方法
An Improved Method of Low-Dimensional Representation of Images
投稿时间:2016-03-31  修订日期:2017-05-10
DOI:
中文关键词: 图像低维表示  L2范数  稳健型光流算法  SIFT流算法
英文关键词: low-dimensional representation of images  L2 norm  robust optical flow  SIFT flow
基金项目:河南省教育厅重点科技攻关项目(13A520786)
作者单位
曾步衢 黄淮学院驻马店 463000 
摘要点击次数:
全文下载次数:
中文摘要:
      目的 解决当前方法需要对图像中的相应点手动标记界标,且局限于特定对象或形状变形的问题。方法 提出一种可以同时实现图像颜色、外观和形态的图像低维表示算法。结果 该算法通过将形态和外观的流形约束到低维子空间上,进一步降低了流形学习的采样复杂性。结论 文中方法的性能远优于目前典型的稳健型光流算法和SIFT流算法。在图像编辑和关节学习关任务中取得了令人满意的定性结果。
英文摘要:
      The work aims to solve the problem that the existing solutions either require manually specified landmarks for corresponding points in the images, or are restricted to specific objects or shape deformations. A low-dimensional representation of images for simultaneously recovering color, appearance and shape was proposed. The proposed algorithm further reduced sample complexity of manifold learning as the manifolds of shape and appearance were restricted to low-dimensional subspaces. The proposed method significantly outperformed the current typical methods of robust optical flow and SIFT flow. Our qualitative results in some related tasks such as image deformation and joint learning are encouraging.
查看全文   查看/发表评论  下载PDF阅读器
关闭

关于我们 | 联系我们 | 投诉建议 | 隐私保护 | 用户协议

您是第20814097位访问者    渝ICP备15012534号-2

版权所有:《包装工程》编辑部 2014 All Rights Reserved

邮编:400039 电话:023-68795652 Email: designartj@126.com

    

渝公网安备 50010702501716号