Visual Object Categorization based on Gabor Filter Generalization Via K-Means Clustering

  • Abstract
  • Keywords
  • References
  • PDF
  • Abstract

    Content based object recognition systems need informative image properties to obtain good performance results. Filter bank such as Gabor filters is believed to be one of the most popular methods for complete characterization of images by having some important properties such as selectivity to orientation, scale, frequency and smooth parameters. Furthermore, such properties are very effective for compact image description and analysis. However, these functions show a strong dependence on a certain number of different parameter values. Hence, the different filter parameters values used to construct the functions may give different filter responds or properties. Besides, the large number of these filters leads to expensive computation to create maps for feature extraction, thus it is necessary to reduce the number of candidates and identify subset of effective and discriminative filters to avoid overfitting and hinder generalization performance. In this paper, we first compute Gabor filters using a set of different values for filter parameters. After that, the k-means clustering algorithm is used to group these filter responds into k different clusters. Next the k different clusters are used to convolve images and the edge histogram then apply to the filter outputs for image description. After that, we combine all outputs of image descriptors using SVMs. Experiment results on 20 and 101 classes of the Caltect-101 object database show that the method significantly outperforms using the standard Gabor filter approach.


  • Keywords

    Gabor filter, naïve combination approach, SVM classifier, VOC technique.

  • References

      [1] Abdullah SNHS, Khalid M, Yusof R & Omar K (2007), Comparison of feature extractors in license plate recognition. First Asia International Conference on Modelling & Simulation, 502-506.

      [2] Mizher MAA, Ang MC, Abdullah Siti Norul Huda Sheikh & Ng Kok Weng (2017), Action Key Frames Extraction Using L1-Norm and Accumulative Optical Flow for Compact Video Shot Summarisation. International Visual Informatics Conference, 364-375

      [3] Abdullah, A Supervised Learning Algorithms for Visual Object Categorization, Institute of Information and Computing Sciences, Utrecht University, (2010), 1-151.

      [4] Abdullah A & Wiering MA (2007), CIREC : Cluster Correlogram Image Retrieval and Categorization using MPEG-7 Descriptors. IEEE Symposium on Computational Intelligence in Image and Signal Processing, 431-437.

      [5] Liua Y, Zhanga D, Lua G & Mab WY (2007), A survey of content-based image retrieval with high-level semantics. Journal of Pattern Recognition, 40: 262– 282.

      [6] Wang JZ, Li J & Wiederhold G (2001), SIMPLIcity: Semantics-Sensitive Integrated Matching for Picture LIbraries. IEEE Transactions on Pattern Analysis and Machine Intelligence, 23(9): 947-963.

      [7] Choi WP, Tse SH, Wang KW & Lam KM (2008), Simplified Gabor wavelets for human face recognition. Journal of Pattern Recognition. 41(3): 1186-1199.

      [8] Long F, Zhang H & Feng DD, Multimedia Information Retrieval and Management: Technological Fundamentals and Applications, New York, Springer, (2003), 1-26.

      [9] Nadernejad E, Sharifzadeh S & Hassanpour H (2008), Edge Detection Techniques: Evaluations and Comparisons. Applied Mathematical Sciences, 2: 1507 - 1520.

      [10] Zhu J, Vai MI, & Mak PU (2004), A New Enhanced Nearest Feature Space (ENFS) Classifier for Gabor Wavelets Features-Based Face Recognition. Lecture Notes in Computer Science 3072, Springer, 124 - 131.

      [11] Shen LL & Ji Z (2009), Gabor Wavelet Selection and SVM Classification for Object Recognition. Acta Automatica Sinica, 35(4): 350-355.

      [12] Xing W & Bir B (1997), Gabor wavelet representation for 3-D object recognition. IEEE Transactions on Image Processing, 6(1): 47-64.

      [13] Li W, Mao K, Zhang H & Chai T (2010). Designing compact Gabor filter banks for efficient texture feature extraction. 11th International Conference on Control Automation Robotics & Vision, 1193 – 1197.

      [14] Jones JP & Palmer A (1987), An evaluation of the two-dimensional Gabor filter model of simple receptive fields in cat striate cortex. Journal of Neurophysiology, 58(6): 1233-1258.

      [15] Varma M & Zisserman A (2003), Texture classification: are filter banks necessary?. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, 2:691-698.

      [16] Mignotte M (2012), An Energy-Based Model for the Image Edge-Histogram Specification Problem. IEEE Transactions on Image Processing, 21(1): 379-386.

      [17] Manjunath BS, Salembier P & Sikora T, Introduction to MPEG-7: Multimedia Content Description Interface, John Wiley & Sons, (2002).

      [18] Alemu Y, Koh JB, Ikram M & Kim DK (2009). Image Retrieval in Multimedia Databases: A Survey. Fifth International Conference on Intelligent Information Hiding and Multimedia Signal Processing, 681 – 689.

      [19] Mohammed Salameh, Abdullah A, Shahnorbanun Sahran, (2018), Multiple Descriptors for Visual Odometry Trajectory Estimation, International Journal on Advanced Science, Engineering and Information Technology. vol. (8), issue 4-2, pp 1423-1430.

      [20] Cortes C & Vapnik V (1995), Support-vector networks. Machine Learning, 20(3): 273-297.

      [21] Yi Y, Dong X, Feiping N, Shuicheng Y & Yueting Z (2010). Image Clustering Using Local Discriminant Models and Global Integration. IEEE Transactions on Image Processing, 19(10): 2761-2773.

      [22] Jain AK, Murty MN & Flynn PJ (1999), Data clustering: a review. ACM Computing Surveys, 31(3): 264–323.




Article ID: 24900
DOI: 10.14419/ijet.v8i1.2.24900

Copyright © 2012-2015 Science Publishing Corporation Inc. All rights reserved.