采用改进的尺度不变特征变换算法计算物体旋转角度

朱齐丹; 李科; 蔡成涛; 程甘霖

doi:10.3788/OPE.20111907.1669

您当前的位置：

首页 >

文章列表页 >

采用改进的尺度不变特征变换算法计算物体旋转角度

信息科学 | 更新时间：2020-08-12

- 采用改进的尺度不变特征变换算法计算物体旋转角度
- Calculation of object rotation angle by improved SIFT
- 光学精密工程 2011年19卷第7期页码：1669-1676
- 作者机构：
  
  哈尔滨工程大学自动化学院,黑龙江哈尔滨,150001
- 作者简介：
- 基金信息：
  
  国家自然科学基金资项目(No.60875025)()
- DOI：10.3788/OPE.20111907.1669
  中图分类号： TB92;TP391.4
- 收稿日期：2010-09-06，
  
  修回日期：2010-11-18，
  
  网络出版日期：2011-07-25，
  
  纸质出版日期：2011-07-25
- 稿件说明：
移动端阅览
朱齐丹, 李科, 蔡成涛, 程甘霖. 采用改进的尺度不变特征变换算法计算物体旋转角度[J]. 光学精密工程, 2011,19(7): 1669-1676

ZHU Qi-dan, LI Ke, CAI Cheng-tao, CHENG Gan-lin. Calculation of object rotation angle by improved SIFT[J]. Editorial Office of Optics and Precision Engineering, 2011,19(7): 1669-1676
朱齐丹, 李科, 蔡成涛, 程甘霖. 采用改进的尺度不变特征变换算法计算物体旋转角度[J]. 光学精密工程, 2011,19(7): 1669-1676 DOI： 10.3788/OPE.20111907.1669.

ZHU Qi-dan, LI Ke, CAI Cheng-tao, CHENG Gan-lin. Calculation of object rotation angle by improved SIFT[J]. Editorial Office of Optics and Precision Engineering, 2011,19(7): 1669-1676 DOI： 10.3788/OPE.20111907.1669.

摘要

改进了传统的尺度不变特征变换(SIFT)算法

使其在进行图像匹配的同时

可以求取出物体的旋转角度。首先

利用SIFT特征对旋转保持不变的特性

按照原算法提取出旋转前后两幅图像的SIFT特征

分析特征点主方向的计算过程

记录每个特征点主方向的角度值进行特征匹配。然后

计算出每对匹配的SIFT特征点的主方向角度之差

得到特征点的旋转角度;采用迭代自组织聚类的方法分析得到的特征点旋转角度数据

依据类内方差和类内样本数目

选取正确的样本类。最后

选用该样本类的均值作为物体的最终旋转角度。实验结果表明

该方法在图像畸变不大时的误差在3以内

即使在部分遮挡的情况下

也能较好地计算出旋转角度。在时间复杂度增加不大的情况下

使SIFT算法具有了计算旋转角度的功能

拓宽了应用方向。

Abstract

The Scale-invariant Feature Transform(SIFT) algorithm was improved in this paper

which could match two pictures and could also compute the object rotation angles in the pictures. Firstly

the SIFT was used to extract two images according to the SIFT feature invariance. Then

the computing process for the main direction of the feature point was analyzed

and the main angle for every key point was recorded. After matching two pictures

the angle difference of main direction for each pair of matched SIFT feature points was calculated and the rotation angle of feature point was obtained. Afterthat

all of the rotation angles of the feature points were analyzed by iterative self-organizing clustering method. Finally

the correct class of samples was selected by the variance and the number of samples within the classe

and the mean of the correct class was used as the final rotation angle of the object. The experiment results indicate that the rotation angle error is within 3 when the image distortion is not significant and it can also well estimate the rotation angle even if the object is partially occluded. Furthermore

in the case of the time complexity does not increase obvionsly

the SIFT can compute the rotation angle of the object

which expands its applications.

关键词

Keywords

references

高春甫,唐可洪,胡庆玉,等. 机械手姿态识别的立体视觉匹配[J]. 光学精密工程,2010,18(2):464-469. GAO CH F, TANG K H, HU Q Y, et al. Stereo vision matching for posture recognition of manipulator [J]. Opt. Precision Eng., 2010,18(2):464-469. (in Chinese)[2] 刘书桂,赵金才,李建苹,等. 智能坐标测量系统中零件位姿自动识别[J]. 天津大学学报,2006,39(10):1241-1246. LIU SH G, ZHAO J C, LI J P, et al. Automatic recognition for position and orientation of parts in intelligent coordinate measuring system[J]. Journal of Tianjin University,2006,39(10):1241-1246. (in Chinese)[3] 孙凤杰,郭凤顺,范杰清,等. 基于图像处理技术的表盘指针角度识别研究[J]. 中国机电工程学报,2005,25(16):73-78. SUN F J, GUO F SH, FAN J Q, et al. Studies of the recognition of pointer angle of dial based on image processing[J]. Proceedings of the CSEE, 2005,25(16):73-78. (in Chinese)[4] 白福忠,包晓艳,李义军. 角度的图像测量法[J]. 计量学报,2007,28(4):314-316. BAI F ZH, BAO X Y, LI Y J. An image measuring method for angle[J]. Acta Metrologica Sinica, 2007,28(4):314-316. (in Chinese)[5] 徐文福,梁斌,李成,等. 空间机器人捕获非合作目标的测量与规划方法[J]. 机器人,2010,32(1):61-69. XU W F, LIANG B, LI CH, et al. Measurement and planning approach of space robot for capturing non-cooperative target[J]. Robot, 2010,32(1):61-69. (in Chinese)[6] VINCENT L, FRANCESC M, PASCAL F.An accurate O(n) solution to the PnP problem[J]. International Journal of Computer Vision, 2009,81(2):155-166.[7] TANG J L,CHEN W SH, WANG J. A novel linear algorithm for P5P problem[J]. Applied Mathematics and Computation, 2008,205(2):628-634.[8] 许海霞,王耀南,袁小芳,等. 基于矢量差分的未标定摄像机P5P问题的求解[J]. 自动化学报,2009,35(8):1140-1144. XU H X, WANG Y N, YUAN X F, et al. Solution to the P5P problem with un-calibrated camera based on vector difference[J]. Acta Automatica Sinica,2009,35(8):1140-1144. (in Chinese)[9] DAVID G.Lowe distinctive image features from scale-invariant keypoints[J]. International Journal of Computer Vision, 2004,60(2):91-110.[10] PLINIO M, ALEXANDRE B, JOS S. Improving the SIFT descriptor with smooth derivative filters[J]. Pattern Recognition Letters, 2009, 30(1):18-26.[11] 杨晓敏,吴炜,卿粼波,等. 图像特征点提取及匹配技术[J]. 光学精密工程,2009,17(9):2276-2282. YANG X M, WU W, QING L B, et al. Image feature extraction and matching technology [J]. Opt. Precision Eng., 2009,17(9):2276-2282. (in Chinese)[12] 贾鑫. 台架试验中车轮位姿视觉识别算法的研究. 长春:吉林大学,2005. JIA X. Research on vision recognition algorithm of wheel pose and position parameters in bench testing .Changchun: Jilin University,2005.(in Chinese)

浏览量

666

下载量

CSCD

文章被引用时，请邮件提醒。

提交

工具集

关联资源

基于二维混合式位置编码的高精度测角

数字天顶仪中倾角仪参数的标定

微球测速聚类分析的流式液路稳定性评估