Метод поиска изображений с использованием вейвлет-технологии

Статья в Elpub

Радиостроение. 2015; : 87-100

Метод поиска изображений с использованием вейвлет-технологии

Аннотация

В статье рассмотрен метод поиска заданных изображений в базах данных больших размеров. Предложена модифицированная версия данного метода, основанная на применении вейвлет базиса Добеши. Для определения степени близости двух изображений рассмотрена метрика, основанная на сравнении коэффициентов вейвлет преобразования. Предложен алгоритм определения коэффициентов на основе метода логистической регрессии. Проведено экспериментальное исследование представленного модифицированного метода. Продемонстрировано его преимущество по сравнению с использованными ранее методами с точки зрения точности реализации запроса и быстродействия. DOI: 10.7463/rdopt.0515.0813256

Список литературы

1. Э.Столниц, Т.Дероуз, Д.Салезин. Вейвлеты в компьютерной графике. - Ижевск, НИЦ «Регулярная и хаотическая динамика», 2003, 272 с.

2. SmithJ.R., ChangS.F. ToolsandTechniquesforColorImageRetrieval. // SPIE (ColumbiaUniv., USA, 1996): Proceedings of the SPIE, vol. 2670, 1996. pp. 426-437.

3. Duda R., Hart P. Pattern Classification and Scene Analysis. John Wiley and Sons publishing, 1973. 512 p.

4. D. G. Lowe. Distinctive Image Features from Scale-Invariant keypoints. Режим доступа: https://www.cs.ubc.ca/~lowe/papers/ijcv04.pdf (дата обращения 02.06.2015).

5. Niblack W., Barber R., Equitz W., Flickner M., Glasman E., Petkovic D., Yanker P., Faloutsos C., Taubin G. The QBICproject: Querying images by content using color, texture, and shape volume. // SPIE (Bellingham, WA, 1993): Storage and Retrieval, 1993. pp. 173-187.

6. E.Stollnitz, E.J.,DeRose T.D., Salesin D.H. Wavelets for Computer Graphics. Theory and Applications. Morgan Kaufmann PublishersInc., 1996. 245p.

7. BeylkinО.,Coifman R., RokhlinУ. Fast wavelet transforms and numerical algorithms. // Communications on pure and Applied Mathematics, 1991, vol. 44. pp. 141-183.

8. Kankanhalli A., Zhang H.J., Low C.Y. Using texture for image retrieval. // International Conference оn Automation, Robotics and Computer Vision. (Nanyang Technological University, Singapore, 1994), IEEE publ., 1994. pp. 935-939.

9. Gibson A.S. Exposure and Understanding the Histogram. PeachpitPresspubl., 2011. 75 p.

10. С.E.Jacobs, A.Finkelstein, D.H.Salesin. Fast multiresolution image quering. Proceedings of SIGGRAPH, ACM, New York, 1995, pp. 277-286.

11. Najmi A.H. Wavelets: A Concise Guide. Johns Hopkins University Press publ., 2012. 304 p.

12. Fugal D.L. Conceptual Wavelets in Digital Signal Processing. Space & Signals Technical Publishing, 2010. 374 p.

13. Vetterli M., Kovacevic E., Goyal V.K. Fourier and Wavelet Signal Processing. Cambridge University Press, 2014. 294p.

14. Davis J., Goadrich M. The Relationship Between Precision-Recall and ROC Curves // Proc. Of 23 International Conference on Machine Learning, Pittsburgh, PA, 2006, pp.233-240.

Radio Engineering. 2015; : 87-100

The Method of Image Retrieval Using Wavelet Technology

Filippov M. V., Gaiduk S. E.

Abstract

Recently, the problem of quick search of specified images has taken developers’ attention. The main reason for this interest lies in the substantially increasing capacity of graphic information, which stipulates the need to create instant search algorithms. An important problem is also to develop the metric for determining an affinity degree of the two images.
A number of papers describe a wavelet- based technology method. Herewith the metric is based on a comparison of the wavelet transform coefficients. These papers demonstrate the benefits of such an approach. This article describes a modification of the method described above in order to increase the effectiveness of the image retrieval on the image-request, as well as to reduce the retrieval time. The presented method has the following features. Firstly, Daubechies wavelets have been proposed as the decomposition basis, in contrast to previously used Haar wavelets. This reduced the number of expansion coefficients and thus, reduced the search time. Secondly, to determine the coefficients has been used so-called logistic regression algorithm based on the statistical model. The paper gives a detailed description of the algorithm to implement said procedure of image search.
To assess the effectiveness of the presented method based both on the criteria of accuracy of selecting a given image (the percentage of successfully completed requests) and on the speed, there were conducted numerical experiments to search for images in databases of different capacities.
The paper has shown that the proposed metric provides a substantially greater speed and accuracy than the standard metric L1 and L2 . It has also demonstrated the advantage of using Daubechies wavelet- basis.

References

1. E.Stolnits, T.Derouz, D.Salezin. Veivlety v komp'yuternoi grafike. - Izhevsk, NITs «Regulyarnaya i khaoticheskaya dinamika», 2003, 272 s.

2. SmithJ.R., ChangS.F. ToolsandTechniquesforColorImageRetrieval. // SPIE (ColumbiaUniv., USA, 1996): Proceedings of the SPIE, vol. 2670, 1996. pp. 426-437.

3. Duda R., Hart P. Pattern Classification and Scene Analysis. John Wiley and Sons publishing, 1973. 512 p.

4. D. G. Lowe. Distinctive Image Features from Scale-Invariant keypoints. Rezhim dostupa: https://www.cs.ubc.ca/~lowe/papers/ijcv04.pdf (data obrashcheniya 02.06.2015).

6. E.Stollnitz, E.J.,DeRose T.D., Salesin D.H. Wavelets for Computer Graphics. Theory and Applications. Morgan Kaufmann PublishersInc., 1996. 245p.

7. BeylkinO.,Coifman R., RokhlinU. Fast wavelet transforms and numerical algorithms. // Communications on pure and Applied Mathematics, 1991, vol. 44. pp. 141-183.

8. Kankanhalli A., Zhang H.J., Low C.Y. Using texture for image retrieval. // International Conference on Automation, Robotics and Computer Vision. (Nanyang Technological University, Singapore, 1994), IEEE publ., 1994. pp. 935-939.

9. Gibson A.S. Exposure and Understanding the Histogram. PeachpitPresspubl., 2011. 75 p.

10. S.E.Jacobs, A.Finkelstein, D.H.Salesin. Fast multiresolution image quering. Proceedings of SIGGRAPH, ACM, New York, 1995, pp. 277-286.

11. Najmi A.H. Wavelets: A Concise Guide. Johns Hopkins University Press publ., 2012. 304 p.

12. Fugal D.L. Conceptual Wavelets in Digital Signal Processing. Space & Signals Technical Publishing, 2010. 374 p.

13. Vetterli M., Kovacevic E., Goyal V.K. Fourier and Wavelet Signal Processing. Cambridge University Press, 2014. 294p.

14. Davis J., Goadrich M. The Relationship Between Precision-Recall and ROC Curves // Proc. Of 23 International Conference on Machine Learning, Pittsburgh, PA, 2006, pp.233-240.