Paper Details: Downloads: 873
Serial Number: P1150819004
Title: A New Method of Texture Classification using various Wavelet Transforms based on Primitive Patterns
Authors: V Vijaya Kumar, U S N Raju, K.Chandra Sekaran ,V V Krishna
Abstract: Study of patterns on a local neighborhood of a texture plays an important role in the characterization and classification of the textures. The present paper assumes that texture is characterized not only by gray value at a pixel but also the gray value pattern in a neighborhood, surrounding the pixel. To achieve this goal the present paper studies the accuracy rate of primitive pattern trends in various wavelet transforms. A 4×4 neighborhood of original image will be converted into a 2×2 neighborhood by any wavelet transform in any sub band. To achieve integrity, the patterns of 4×4 mask of original image are compared with same 2×2 mask of wavelet image. A new method is introduced in the present paper which is different from distance function for classification of textures based on primitive patterns. The classification rates are compared with original image and with other wavelet transformed images and also between each of the wavelet images. The experimental result on 24 Brodatz text
Keywords: Texture, Wavelet Transforms, Gray value pattern, Accuracy rate of primitive patterns.
Journal/Conference: International Journal of Graphics, Vision and Image Processing
Volume: 8
Issue: 2
Submission Date: 5/5/2008 12:00:00 AM
Review Date: 5/12/2008 12:00:00 AM
Publishing Date: 6/2/2008 12:00:00 AM
Article Downloads: 873
Download:

Facebook