基于最小生成树的图像结构描述
Image Structure Description Based on MST

作者: 曲智国 * , 谭贤四 , 张伟 , 王红 , 林强 :空军预警学院二系,湖北 武汉;

关键词: 图像结构描述最小生成树评价指标图像匹配场景分类Image Structure Description MST Evaluation Indexes Image Matching Scene Classification

摘要:
最小生成树是表示图像的常用结构,在图像匹配和图像检索等任务中应用广泛。但是,现有方法只是利用最小生成树来表示图像,缺乏分析和描述图像结构的量化指标。本文提出了基于最小生成树的图像结构描述指标,首先采用最小生成树来表示图像中的结构基元,然后基于最小生成树定义了分布均匀度、支撑度和描述符独特性三个量化指标,来衡量图像中结构基元对图像结构的描述能力。采用仿真点集和实际图像进行实验,结果表明,基于最小生成树定义的量化指标可以较好地描述图像的结构,可用来对图像场景粗略分类和评价图像匹配的难易程度。

Abstract: The MST (Minimum Spanning Tree) is an effective structure for representing images in applica-tions such as image matching and image retrieval. However, previous methods only use MST to connect the structural elements in images and are unable to quantitatively analyze the capability of the elements in describing the image. A method for structural description of images based on MST is proposed in this paper. The MST constructed from the structural elements extract from the image supports the image like the skeleton of the image, which reflects the structure of image. Three indexes for evaluating the distribution uniformity, the distinctiveness and the spanning degree of a set of image structural elements are defined based on the MST. The evaluating indexes can be used to analyze the structural information of an image and can also be used to evaluate the capability of a set of structural elements to describe the structure of an image. Experiments on simulated points and real images validate the proposed indexes. The indexes can be used to classify the structures of image scene and to evaluate the difficulty of image matching.

文章引用: 曲智国 , 谭贤四 , 张伟 , 王红 , 林强 (2017) 基于最小生成树的图像结构描述。 图像与信号处理, 6, 106-113. doi: 10.12677/JISP.2017.62013

参考文献

[1] Zhao, M., An, B., Wu, Y., et al. (2015) A Robust Delaunay Triangulation Matching for Multispectral/Multidate Remote Sensing Image Registration. IEEE Geoscience and Remote Sensing Letters, 12, 711-715.
https://doi.org/10.1109/LGRS.2014.2359518

[2] Wang, S., Guo, X., Mu, X., et al. (2015) Advanced Weight Graph Transformation Matching Algorithm. IET Computer Vision, 9, 960-966.
https://doi.org/10.1049/iet-cvi.2014.0339

[3] 支力佳, 张少敏, 赵大哲, 等. 基于最小生成树的DoG关键点医学图像配准[J]. 中国图象图形学报, 2011, 16(4): 647-653.

[4] Bostanci, E., Kanwal, N. and Clark, A.F. (2014) Spatial Statistics of Image Features for Performance Comparison. IEEE Transactions on Image Processing, 23, 153-162.
https://doi.org/10.1109/TIP.2013.2286907

[5] Zhu, Q., Wu, B. and Xu, Z. (2006) Seed Point Selection Method for Triangle Constrained Image Matching Propagation. IEEE Geoscience and Remote Sensing Letters, 3, 207-211.
https://doi.org/10.1109/LGRS.2005.861735

[6] 朱英宏, 李俊山, 杨威, 等. 异源图像特征点边缘描述与匹配[J]. 计算机科学, 2013, 40(7): 277-280.

[7] Redmond, C. and Yukich, J. (1996) Asymptotics for Euclidean Functionals with Power-Weighted Edges. Stochastic Processes and Their Applications, 61, 289-304.

[8] Hero, A., Ma, B. and Michel, O. (2002) Applications of Entropic Spanning Graphs. IEEE Signal Processing Magazine, 19, 85-95.
https://doi.org/10.1109/MSP.2002.1028355

[9] Rényi, A. (1961) On Measures of Entropy and Information. Proceedings of the 4th Berkeley Symposium on Mathematical Statistics and Probability, 1, 547-561.

[10] Hoffman, R. and Jain, A.K. (1983) A Test of Randomness Based on the Minimal Spanning Tree. Pattern Recognition Letters, 1, 175-180.
https://doi.org/10.1016/0167-8655(83)90059-4

[11] Mikolajczyk, K., Tuytelaars, T., Schmid, C., et al. (2005) A Comparison of Affine Region Detectors. International Journal of Computer Vision, 65, 43-72.
https://doi.org/10.1007/s11263-005-3848-x

[12] Mikolajczyk, K. and Schmid, C. (2005) A Performance Evaluation of Local Descriptors. IEEE Transactions on Pattern Analysis and Machine Intelligence, 27, 1615-1630.
https://doi.org/10.1109/TPAMI.2005.188

[13] Umeyama, S. (1988) An Eigen-Decomposition Approach to Weighted Graph Matching Problems. IEEE Transactions on Pattern Analysis and Machine Intelligence, 10, 695-703.
https://doi.org/10.1109/34.6778

分享
Top