[1]王颖,王爱民.一种鲁棒的二进制图像特征点描述子[J].东南大学学报(自然科学版),2012,42(2):265-269.[doi:10.3969/j.issn.1001-0505.2012.02.014]
 Wang Ying,Wang Aimin.Robust binary feature point descriptor[J].Journal of Southeast University (Natural Science Edition),2012,42(2):265-269.[doi:10.3969/j.issn.1001-0505.2012.02.014]
点击复制

一种鲁棒的二进制图像特征点描述子()
分享到:

《东南大学学报(自然科学版)》[ISSN:1001-0505/CN:32-1178/N]

卷:
42
期数:
2012年第2期
页码:
265-269
栏目:
计算机科学与工程
出版日期:
2012-03-20

文章信息/Info

Title:
Robust binary feature point descriptor
作者:
王颖 王爱民
东南大学仪器科学与工程学院,南京 210096
Author(s):
Wang Ying Wang Aimin
School of Instrument Science and Engineering, Southeast University, Nanjing 210096, China
关键词:
特征点 特征匹配 SURF算法
Keywords:
feature point feature matching speeded up robust feature(SURF)algorithm
分类号:
TP391.4
DOI:
10.3969/j.issn.1001-0505.2012.02.014
摘要:
为了提高特征点匹配的速度,采用二进制方法生成特征点描述,并对描述子进行了尺度和旋转适应性改进.使用特征点邻域小块中随机点的强度对比生成描述,描述子的相似度以Hamming距离度量,以二进制运算提高算法的时间性能.为了检验算法在视角、旋转及尺度变化时的性能,采用Wall和Graffiti图像集及相应的旋转和尺度变换图像集对算法进行测试,得到该算法在各图像集上的匹配准确率,并与SURF算法得到的结果进行比较.结果表明,在2幅图像间进行特征点匹配时,该算法的特征点描述生成时间和匹配时间分别为1 043.67和4 313.36 ms,而使用SURF算法时的相应时间分别为3 950.34和9 951.03 ms,说明该算法的时间特性明显优于SURF算法.此外,在绝大多数测试集上,该算法的匹配准确率明显高于SURF算法.
Abstract:
In order to improve the speed of feature point matching, a binary method is used to generate feature point description, and the descriptor’s adaptability to different scales and rotations is improved. The descriptor is computed using intensity difference tests. The descriptor similarity is evaluated by using Hamming distance, and the time performance of the algorithm is improved by binary operation. The Wall and Graffiti image sets as well as their transformed image sets are used to test the performance of the proposed algorithm for the different perspectives, rotations and scales. The matching accuracies on each image set are obtained. The comparison results of the proposed algorithm and the speeded up robust feature(SURF)algorithm show that during the feature point matching between the two images, the construction time and the matching time of the descriptors of the proposed algorithm are 1 043.67 and 4 313.36 ms, respectively, while the corresponding data of the SURF algorithm are 3 950.34 and 9 951.03 ms, indicating that the time characteristics of the proposed algorithm are better than those of the SURF algorithm. In addition, on most image sets, the matching accuracy of the proposed algorithm is higher than that of the SURF algorithm.

参考文献/References:

[1] Lowe D G.Object recognition from local scale-invariant features[C] //Proceedings of the 7th IEEE International Conference on Computer Vision.Los Alamitos,USA,1999:1150-1157.
[2] Lowe D G.Distinctive image features from scale-invariant keypoints[J].International Journal of Computer Vision,2004,60(2):91-110.
[3] Bay Herbert,Tuytelaars Tinne,Van Gool Lue.SURF:speeded up robust features[C] //Proceedings of the 9th European Conference on Computer Vision.Graz,Austria,2006:404-417.
[4] Mikolajczyk K,Schmid C.A performance evaluation of local descriptors[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2005,27(10):1615-1630.
[5] Ke Y,Sukthankar R.PCA-SIFT:a more distinctive representation for local image descriptors[C] //Proceedings of the 2004 IEEE Computer Society Conference on Computer Vision and Pattern Recognition.Washington DC,USA,2004:506-513.
[6] Lepetit V,Pilet J,Fua P.Point matching as a classification problem for fast and robust object pose estimation[C] //Proceedings of the 2004 IEEE Computer Society Conference on Computer Vision and Pattern Recognition.Washington DC,USA,2004:244-250.
[7] Lepetit V,Fua P.Keypoint recognition using randomized trees[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2006,28(9):1465-1479.
[8] Ozuysal M,Calonder M,Lepetit V,et al.Fast keypoint recognition using random ferns[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2010,32(3):448-461.
[9] Calonder M,Lepetit V,Strecha C,et al.BRIEF:binary robust independent elementary features[C] //Proceedings of the 11th European Conference on Computer Vision.Berlin,Germany,2010:778-792.
[10] Lindeberg T.Feature detection with automatic scale selection[J].International Journal of Computer Vision,1998,30(2):79-116.
[11] Willow Garage.OpenCV[EB/OL].(2010-04-10)[2010-06-03].http://opencv.willowgarage.com.

相似文献/References:

[1]罗翔,席文明,颜景平.一种双目主动立体视觉系统的目标定位算法[J].东南大学学报(自然科学版),2002,32(1):59.[doi:10.3969/j.issn.1001-0505.2002.01.014]
 Luo Xiang,Xi Wenming,Yan Jingping.Scheme of target 3D point positioning for binocular active stereo visual system[J].Journal of Southeast University (Natural Science Edition),2002,32(2):59.[doi:10.3969/j.issn.1001-0505.2002.01.014]
[2]刘天亮,戴修斌,陈昌红,等.对光照变化鲁棒的快速关键点提取与匹配[J].东南大学学报(自然科学版),2012,42(3):413.[doi:10.3969/j.issn.1001-0505.2012.03.004]
 Liu Tianliang,Dai Xiubin,Chen Changhong,et al.Fast keypoint extraction and matching robust to illumination changes[J].Journal of Southeast University (Natural Science Edition),2012,42(2):413.[doi:10.3969/j.issn.1001-0505.2012.03.004]

备注/Memo

备注/Memo:
作者简介: 王颖(1985—),女,博士生; 王爱民(联系人),男,博士,教授,博士生导师,wangam@seu.edu.cn.
基金项目: 国家高技术研究发展计划(863计划)资助项目(2009AA01Z311).
引文格式: 王颖,王爱民.一种鲁棒的二进制图像特征点描述子[J].东南大学学报:自然科学版,2012,42(2):265-269. [doi:10.3969/j.issn.1001-0505.2012.02.014]
更新日期/Last Update: 2012-03-20