A SIFT Feature Descriptor Based on Sector Area Partitioning
-
摘要: 提出了一种在圆形区域内基于扇形区域分割的特征描述符构建方法. 首先, 针对SIFT描述符维数过高, 导致匹配速度慢的弱点, 提出在半径为9像素的圆形特征区域内划分为8个扇区, 在这些扇形特征邻域内统计8个方向的灰度梯度直方图, 形成64维描述符的方法,降低了描述符的维数. 同时, 针对SIFT构建描述符的运算复杂性较高的事实, 提出在圆形区域内计算像素灰度梯度主方向, 以主方向为基准点把该区域划分为8个等面积扇区的方法, 取消了对特征区域的旋转变换, 降低了构建描述符的运算复杂性. 通过与OpenCV SIFT和Lowe SIFT进行多方面对比实验, 结果表明该方法的综合匹配速度具有显著提升, 在两幅图像存在一定程度的视点、模糊、旋转、比例、光照变化等情形下, 匹配性能有所增强.Abstract: This paper presents a new method of constructing feature descriptor based on sector area partitioning in a circular region. The large dimension of descriptor will decrease the matching speed of SIFT algorithm. In order to solve the problem, we partition a circular region whose radius is nine pixels into eight bins in angular direction. By computing the gradient orientation histogram in the eight directions for each bin, a descriptor with 64 dimensions is constituted. Therefore, the dimension of the descriptor is reduced. Meanwhile, based on the fact that computation of constructing SIFT descriptor is complex, we introduce a strategy that computes the dominate orientation of pixel gray gradients in the circular region, and then partitions the circular region into eight identical sector areas starting from the dominate orientation. Consequently, the computational complexity is reduced due to cancellation of rotation operation. By comprehensive comparison with the OpenCV SIFT and Lowe SIFT, the results indicate that the proposed method can increase the average matching speed significantly. Even if there exist affine distortion, defocusing, rotation, scaling or illumination variation, the matching performance can also be improved.
-
Key words:
- Machine vision /
- feature descriptor /
- image matching /
- SIFT
-
[1] Zhuang Yan, Chen Dong, Wang Wei, Han Jian-Da, Wang Yue-Chao. Status and development of natural scene understanding for vision-based outdoor mobile robot. Acta Automatica Sinica, 2010, 36(1): 1-11(庄严, 陈东, 王伟, 韩建达, 王越超. 移动机器人基于视觉室外自然场景理解的研究与进展. 自动化学报, 2010, 36(1): 1-11)[2] Lin Hai-Feng, Ma Yu-Feng, Song Tao. Research on object tracking algorithm based on SIFT. Acta Automatica Sinica, 2010, 36(8): 1204-1208(蔺海峰, 马宇峰, 宋涛. 基于SIFT特征目标跟踪算法研究. 自动化学报, 2010, 36(8): 1204-1208)[3] Lowe D G. Object recognition from local scale-invariant features. In: Proceedings of the 7th IEEE International Conference on Computer Vision. Kerkyra, Greece: IEEE, 1999. 1150-1157[4] Lowe D G. Distinctive image features from scale-invariant keypoints. International Journal of Computer Vision, 2004, 60(2): 91-110[5] Mikolajczyk K, Matas J. Improving descriptors for fast tree matching by optimal linear projection. In: Proceedings of the 11th IEEE International Conference on Computer Vision. Rio de Janeiro, Brazil: IEEE, 2007. 1-8[6] Klare B, Li Z F, Jain A K. Matching forensic sketches to mug shot photos. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2011, 33(3): 639-646[7] Liu Li, Peng Fu-Yuan, Zhao Kun, Wan Ya-Ping. Simplified SIFT algorithm for fast image matching. Infrared and Laser Engineering, 2008, 37(1): 181-184(刘立, 彭复员, 赵坤, 万亚平. 采用简化SIFT算法实现快速图像匹配. 红外与激光工程, 2008, 37(1): 181-184)[8] Mikolajczyk K, Schmid C. A performance evaluation of local descriptors. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2005, 27(10): 1615-1630[9] Huang C R, Chen C S, Chung P C. Contrast context histogram —— an efficient discriminating local descriptor for object recognition and image matching. Pattern Recognition, 2008, 41(10): 3071-3077[10] Liu Ping-Ping, Zhao Hong-Wei, Zang Xue-Bai, Dai Jin-Bo. A fast local feature description algorithm. Acta Automatica Sinica, 2010, 36(1): 40-45(刘萍萍, 赵宏伟, 臧雪柏, 戴金波. 一种快速局部特征描述算法. 自动化学报, 2010, 36(1): 40-45)[11] Tola E, Lepetit V, Fua P. DAISY: an efficient dense descriptor applied to wide-baseline stereo. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2010, 32(5): 815-830[12] Lv Ji, Gao Hong-Min, Wang Bo, Zhou Zhi-Qiang. Fast SIFT algorithm for autonomous image guidance. Systems Engineering and Electronics, 2009, 31(5): 1147-1151(吕冀, 高洪民, 汪渤, 周志强. 图像制导中的SIFT快速算法. 系统工程与电子技术, 2009, 31(5): 1147-1151)[13] Ke Y, Sukthankar R. PCA-SIFT: a more distinctive representation for local image descriptors. In: Proceedings of the 2004 IEEE Computer Society Conference on Computer Vision and Pattern Recognition. Washington D. C., USA: IEEE, 2004. 506-513[14] Wong Y M. Invariant Local Feature for Image Matching [Master dissertation], Chinese University of Hong Kong, China, 2006[15] Cui Y, Hasler N, Thormahlen T, Seidel H P. Scale invariant feature transform with irregular orientation histogram binning. In: Proceedings of the 6th International Conference on Image Analysis and Recognition. Halifax, Canada: Springer, 2009. 258-267[16] Chen J, Shan S G, He C, Zhao G Y, Pietikainen M, Chen X L, Gao W. WLD: a robust local image descriptor. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2010, 32(9): 1705-1720[17] Heikkil M, Pietikainen M, Schmid C. Description of interest regions with local binary patterns. Pattern Recognition, 2009, 42(3): 425-436[18] Zeng Hui, Mu Zhi-Chun, Wang Xiu-Qing. A robust method for local image feature region description. Acta Automatica Sinica, 2011, 37(6): 658-664(曾慧, 穆志纯, 王秀青. 一种鲁棒的图像局部特征区域的描述方法. 自动化学报, 2011, 37(6): 658-664)[19] Bay H, Ess A, Tuytelaars T, Gool L V. SURF: speeded up robust features. Computer Vision and Image Understanding, 2008, 110(3): 346-359[20] Meng Fan-Jie, Guo Bao-Long. Image retrieval by using local distribution features of interest points and multiple-instance learning. Journal of Xidian University (Natural Science), 2011, 38(2): 47-53(孟繁杰, 郭宝龙. 使用兴趣点局部分布特征及多示例学习的图像检索方法. 西安电子科技大学学报(自然科学版), 2011, 38(2): 47- 53)[21] Li C L, Ma L Z. A new framework for feature descriptor based on SIFT. Pattern Recognition Letters, 2009, 30(5): 544-557[22] Zeng L, Tan J B. Robust matching algorithm for image mosaic. In: Proceedings of the 6th International Symposium on Precision Engineering Measurements and Instrumentation. Hangzhou, China: SPIE, 2010. 166-175
点击查看大图
计量
- 文章访问数: 1845
- HTML全文浏览量: 51
- PDF下载量: 1721
- 被引次数: 0