计算机应用 ›› 2014, Vol. 34 ›› Issue (9): 2678-2682.DOI: 10.11772/j.issn.1001-9081.2014.09.2678

• 虚拟现实与数字媒体 • 上一篇    下一篇

基于旋转不变特征的SIFT描述子在图像配准中的应用

王帅1,2,孙伟2,姜树明1,刘晓辉1,彭蓬3   

  1. 1. 山东省科学院 情报研究所,济南250014;
    2. 中国矿业大学 信息与电气工程学院,江苏 徐州 221008;
    3. 兖矿集团有限公司,山东 邹城 273500
  • 收稿日期:2014-02-13 修回日期:2014-04-28 出版日期:2014-09-01 发布日期:2014-09-30
  • 通讯作者: 孙伟
  • 作者简介: 
    王帅(1983-),男,山东潍坊人,博士〖BP(〗研究生〖BP)〗,主要研究方向:图像处理、模式识别;
    孙伟(1963-),男,山东徐州人,教授,博士生导师,主要研究方向:人工智能、复杂过程控制;
    姜树明(1979-),男,山东烟台人,副研究员,主要研究方向:图像处理、计算机视觉。
  • 基金资助:

    山东省信息产业发展专项;山东省博士基金资助项目;济南市科技明星计划项目

Application of scale invariant feature transform descriptor based on rotation invariant feature in image registration

WANG Shuai1,2,SUN Wei2,JIANG Shuming1,LIU Xiaohui1,PENG Peng3   

  1. 1. Information Research Institute, Shandong Academy of Sciences, Jinan Shandong 250014, China
    2. School of Information and Electrical Engineering, China University of Mining and Technology, Xuzhou Jiangsu 221008, China
    3. Yankuang Group Company Limited, Zoucheng Shandong 273500, China
  • Received:2014-02-13 Revised:2014-04-28 Online:2014-09-01 Published:2014-09-30
  • Contact: SUN Wei

摘要:

针对尺度不变特征变换(SIFT)算法中描述子维度高造成配准过程中计算量过大的问题,提出了一种改进的SIFT算法。该算法利用圆形的旋转不变性,以特征点为中心,在近似大小的圆形特征点邻域内构造特征描述子,以每个圆环作为一个子环,每个子环内只有像素位置发生了改变,像素之间其他相对信息是保持不变的。当图像发生旋转时,统计每个圆环内元素的梯度累加值进行排序,生成特征向量描述子,降低了算法的维度及复杂度,把特征描述子的维数从128维降低到48维。实验结果表明,改进算法旋转配准重复率在85%以上;在图像旋转、缩放和光照变化情况下,与SIFT算法相比,平均配准准确率提高5%,平均配准耗时降低30%左右,有效实现了对SIFT的改进。

Abstract:

To solve the problem that high dimension of descriptor decreases the matching speed of Scale Invariant Feature Transform (SIFT) algorithm, an improved SIFT algorithm was proposed. The feature point was acted as the center, the circular rotation invariance structure was used to construct feature descriptor in the approximate size circular feature points' neighborhood, which was divided into several sub-rings. In each sub-ring, the pixel information was to maintain a relatively constant and positions changed only. The accumulated value of the gradient within each ring element was sorted to generate the feature vector descriptor when the image was rotated. The dimensions and complexity of the algorithm was reduced and the dimensions of feature descriptor were reduced from 128 to 48. The experimental results show that, the improved algorithm can improve rotating registration repetition rate to more than 85%. Compared with the SIFT algorithm, the average matching registration rate increases by 5%, the average time of image registration reduces by about 30% in the image rotation, zoom and illumination change cases. The improved SIFT algorithm is effective.

中图分类号: