The International Arab Journal of Information Technology (IAJIT)

..............................
..............................
..............................


The Statistical Quantized Histogram Texture Features Analysis for Image Retrieval Based on Median and Laplacian Filters in the DCT Domain  

An effective Content!Based Image Retrieval (CBIR) system is based on efficient feature extraction and accurate retrieval of similar images. Enhanced images by usi ng proper filter methods can also, play an importa nt role in image retrieval in a compressed frequency domain since cu rrently most of the images are represented in the compressed format by using the Discrete Cosine Transformation ( DCT) blo cks transformation. In compression, some crucial information is lost and perceptual information is left, which has significa nt energy requirement for retrieval in a compressed domain. In this paper, the statistical texture features are extracted from the enhanced images in the DCT domain using only t he DC and first three AC coefficients of the DCT blocks of image having m ore significant information. We study the effect of filters in image retrieval using texture features. We perform an exp erimental comparison of the results in terms of accuracy on the basis of median, median with edge extraction and Laplacian f ilters using quantized histogram texture features in a DCT domain. Experiments using the Corel database, give the impr oved results on the basis of filters; more specifically, the Laplacian filter with sharpened images gives good performance in ret rieval of JPEG format images as compared to the med ian filter in the DCT frequency domain.    

 

[1]  Alnihoud  J.,  “Content-Based  Image  Retrieval  System  Based  on  Self  Organizing  Map,  Fuzzy  Color  Histogram  and  Subtractive  Fuzzy  Clustering,”  International Arab Journal of Information Technology ,  vol.  9,  no.  5,  pp.  452- 458, 2012. 

[2]  Anjum  M.  and  Javed  M.,  “Multiresolution  and  Varying Expressions Analysis of Face Images for  Recognition,”  Information Technology Journal ,  vol. 6, no. 1, pp. 57-65, 2007. 

[3]  Bae  H.  and  Jung  S.,  “Image  Retrieval  using  Texture  Based  on  DCT,”  in Proceedings of International Conference on Information, Communications and Signal Processing ,  South  Korea, vol. 2, pp. 1065-1068, 1997. 

[4]  Chakravarti  R.  and  Meng  X.,  “A  Study  of  Color  Histogram  Based  Image  Retrieval,”  in Proceedings of Information Technology: New Generations , Las Vegas, pp. 1323-1328, 2009. 

[5]  Feng  G.  and  Jiang  J.,  “JPEG  Compressed  Image  Retrieval  Via  Statistical  Features,”  Pattern Recognition , vol. 36, no. 4, pp. 977-985, 2003. 

[6]  Gonzalez  R.,  Woods  R.,  and  Eddins  S.,  Digital Image Processing using MATLAB ,  Pearson  Prentice Hall,  2004. 

[7]  Lay  J.  and  Guan  L.,  “Image  Retrieval  Based  on  Energy  Histograms  of  the  Low  Frequency  DCT  Coefficients,”  in Proceedings of IEEE International Conference on Acoustics, Speech, and Signal Processing , Phoenix, vol. 6, pp. 3009- 3012, 1999. 

[8]  Liu  Y.,  Zhang  D.,  Lu  G., and  Ma  W.,  “A  Survey  of  Content-Based  Image  Retrieval  with  High- Level  Semantics,”  Pattern Recognition , vol.  40,  no. 1, pp. 262-282, 2007. 

[9]  Mandal  M.,  Idris  F.,  and  Panchanathan  S.,  “A  Critical  Evaluation  of  Image  and  Video  Indexing  Techniques  in  the  Compressed  Domain,”  Image and Vision Computing , vol.  17,  no.  7,  pp.  513- 529, 1999. 

[10]  Mohamed  A.,  Khellfi  F.,  Weng  Y.,  Jiang  J.,  and  Ipson  S.,  “An  Efficient  Image  Retrieval  through  DCT Histogram Quantization,”  in Proceedings of  International Conference on CyberWorlds ,  Bradford, pp. 237-240, 2009. 

[11]  Nezamabadi-Pour  H.  and  Saryazdi  S.,  “Object  Based  Image  Indexing  and  Retrieval  in  DCT  Domain  using  Clustering  Techniques,”  World Academy of Science Engineering and Technology ,  vol.  3,  no.  1307-6884,  pp.  207-210,  2005. 

[12]  Selvarajah S. and Kodituwakku S., “Analysis and  Comparison  of  Texture  Features  for  Content  Based Image Retrieval,”  International Journal of Latest Trends in Computing , vol.  2,  no.  1,  pp.  108-113, 2011.  624 The International Arab Journal of Information Tech nology, Vol. 10, No. 6, November 2013    

[13]  Sergy´an  S.,  “Color  Histogram  Features  Based  Image  Classification  in  Content-Based  Image  Retrieval  Systems,”  in Proceedings of the  6 th International Symposium on Applied Machine Intelligence and Informatics ,  Herlany,  pp.  221- 224, 2008. 

[14]  Shan  Z.  and  Liu  Y.,  “Research  on  JPEG  Image  Retrieval,”  in Proceedings of the 5 th International Conference on Wireless Communications, Networking and Mobile Computing ,  Beijing,  pp.  1-4, 2009. 

[15]  Thawari P. and Janwe N., “CBIR Based on Color  and  Texture,” International Journal of Information Technology and Knowledge Management , vol. 4, no. 1, pp. 129-132, 2011. 

[16]  Tsai  T.,  Huang  Y.,  and  Chiang  T.,  “Dominant  Feature  Extraction  in  Block-DCT  Domain,”  in Proceedings of  IEEE International Conference on Systems, Man and Cybernetics , Taipei,  vol.  5,  pp. 3623-3628, 2006. 

[17]  Vailaya  A.,  Jain  A.,  and  Zhang  H.,  “On  Image  Classification:  City  Images  vs.  Landscapes,”  Pattern Recognition , vol.  31,  no.  12,  pp.  1921- 1935, 1998. 

[18]  Veltkamp  R.  and  Tanase  M.,  “Content  Based  Image  Retrieval  Systems:  A  Survey,”  Technical Report , Utrecht  University,  Utrecht  the  Netherlands, 2002. 

[19]  Wang J., Li J., and Wiederhold G., “SIMPLIcity:  Semantics-Sensitive  Integrated  Matching  for  Picture   LIbraries,”  Presented at the Advances in Visual Information Systems ,  vol.  23,  no.  9,  pp.  171-193, 2000. 

[20]  Zhao  H.,  Kim  P.,  and  Park  J.,  “Feature  Analysis  Based  on  Edge  Extraction  and  Median  Filtering  for  CBIR,”  in Proceedings of the 11 th International Conference on Computer Modelling and Simulation ,  Cambridge,  pp.  245- 249, 2009. 

[21]  Zhong  D.  and  Defee  I.,  “DCT  Histogram  Optimization  for  Image  Database  Retrieval,”  Pattern Recognition Letters , vol.  26,  no.  14,  pp.  2272-2281, 2005.    Fazal  Malik received  his  BSc  and  MSc  in  computer  science  degrees  from  University  of  Peshawar,  Pakistan.  He  is  currently  a  PhD  student  at  the  Department  of  Computer  and  Information Sciences,  Universiti  Teknologi  PETRONAS,  Malaysia.  His  current  research  interests  include  im age  retrieval and indexing.          Baharum  Baharudin  received  his  BSc  and  MSc  of  science  degrees  from  Central  Michigan  University,  USA  and  the  PhD  degree  from  the  University  of  Bradford,  UK.  Currently,  he  is  an  associate  professor  at  the  Department  of  Computer  and  Information  Sciences,  Universiti  Teknologi  PETRONAS  Malaysia.  His  research  interests  include  image  processing,  data  mining  and   knowledge management.