Object recognition and pose estimation using appearance manifolds

Abstract

Conventionally, image object recognition and pose estimation are two independent components in machine vision. This paper presented a simple but effective method KNN-SNG, which tightly couples these two components within a single algorithm framework. The basic idea of this method came from the bionic pattern recognition and the manifold ways of perception. Firstly, the shortest neighborhood graphs (SNG) are established for each registered object. SNG can be regarded as a covering and triangulation for a hypersurface on which the training data are distributed. Then for recognition task, the determined test image lies on which SNG by employing the parameter “k”, which could be calculated adaptively. Finally, the local linear approximation method was adopted to build a local map between high-dimensional image space and low-dimensional manifold for pose estimation. The projective coordinates on manifold can depict the pose of object. Experiment results manifested the effectiveness of the method.

Keywords

Object recognition / Pose estimation / Manifold

References

[1] Moranduzzo T, Melgani F (2012) A SIFT-SVM method for detecting cars in UAV images. In: 2012 IEEE international geoscience and remote sensing symposium (IGARSS), pp 6868–6871

[2] Viola P, Jones MJ. Robust real-time face detection. Comput Vision, 2004, 57(2): 137-154.

[3] Seung HS, Lee DD. The manifold ways of perception. Science, 2002, 290: 2268-2269.

[4] Tenenbaum JB, de Silva V, Langford JC. A global geometric framework for nonlinear dimensionality reduction. Science, 2000, 290: 2268-2323.

[5] Roweis LS. Nonlinear dimensionality reduction by locally linear embedding. Science, 2000, 290: 2323-2326.

[6] Belkin M, Niyogi P. Laplacian Eigenmaps for dimensionality reduction and data representation. Neural Comput, 2003, 15: 1373-1396.

[7] Zhang Z, Zha H. Principle manifolds and nonlinear dimension reduction via local tangent space alignment. Sci Comput, 2004, 26(1): 313-338.

[8] Ptucha R, Savakis A (2010) Pose estimation using facial feature points and manifold learning. In: Proceedings of 2010 IEEE 17th international conference on image processing, pp 3261–3264

[9] Schwarz LA, Mateus D, Navab N. Recognizing multiple human activities and tracking full-body pose in unconstrained environments. Pattern Recogn, 2012, 45: 11-23.

[10] Vlachos M, Domeniconi C,Gunopulos D et al (2002) Non-linear dimensionality reduction techniques for classification and visualization. In: Proceedings of 8th ACM SIGKDD international conference on knowledge discovery and data mining, pp 645–651

[11] Kadoury S, Levine MD (2006) Finding faces in gray scale images using locally linear embeddings. In: The 5th international conference on image and video retrieval, pp 221–230

[12] Wang SJ, Chen X. Biomimetic (topological) pattern recognition: a new model of pattern recognition theory and its application. Neural Netw, 2003, 3: 2258-2262.

[13] Zeng JY, Zhai YK, Gan JY et al (2009) An effective iris recognition system based on biomimetic pattern recognition. In: International conference on information science and engineering, pp 3583–3586

[14] Zhang JB, Cong S, Xu ZM et al (2009) A text classifier based on biomimetic pattern recognition. In: 2009 International conference on machine learning and cybernetics, pp 3463–3468