International Journal of Computer Applications |
Foundation of Computer Science (FCS), NY, USA |
Volume 94 - Number 16 |
Year of Publication: 2014 |
Authors: Seema Anand Chaurasia, Vaishali Suryawanshi |
10.5120/16446-6119 |
Seema Anand Chaurasia, Vaishali Suryawanshi . Hybrid Algorithm for Image Retrieval using LBG and K-means. International Journal of Computer Applications. 94, 16 ( May 2014), 40-43. DOI=10.5120/16446-6119
In this paper a hybrid algorithm for image retrieval based on texture feature extraction is proposed. Proposed algorithm can be implemented for texture feature retrieval using Vector Quantization (VQ). For texture feature retrieval Linde-Buzo-Gray (LBG) algorithms is used by dividing each image into pixel blocks of size 2X2 where each pixel consists of green, red and blue component. A training vector of dimension 12 can be obtained by putting these in a row. A training set is collection of such training vectors. Size of codebook will be 16X12. In the proposed method K-means algorithm is applied on existing LBG codebook and results are compared with LBG algorithm. From experiments it is found that proposed algorithm gives better relevance percentage as compared to the LBG algorithm.