AUTOREGRESSIVE MODEL BASED ON BAYESIAN APPROACH FOR TEXTURE REPRESENTATION
Journal: ICTACT Journal on Image and Video Processing (IJIVP) (Vol.3, No. 1)Publication Date: 2012-08-01
Authors : T. Karthikeyan; R. Krishnamoorthy;
Page : 485-491
Keywords : Texnum; Texspectnum; Microtexture; K-Means Algorithm; Supervised Classification; Unsupervised Classification;
Abstract
In this study autoregressive model based on Bayesian approach is proposed for texture classification. Based on auto correlation coefficients, micro textures are identified and represented locally and then globally. The identified micro texture is represented as a local description, called texnum. The global descripter, texspectnum, is obtained by simply observing the numbers of occurrences of the texnums that cover the entire image. The proposed representation scheme has been employed in both supervised and unsupervised classifications of textured images. The supervised classification is based on simple tests of hypotheses and the unsupervised classification is based on the modified K-means algorithm with minimum distance classifiers. The proposed method is demonstrated for classification of different types natural textured images. The average correct classification is better than the existing methods.
Other Latest Articles
- AUTOMATIC FAST VIDEO OBJECT DETECTION AND TRACKING ON VIDEO SURVEILLANCE SYSTEM
- MULTI MODAL ONTOLOGY SEARCH FOR SEMANTIC IMAGE RETRIEVAL
- ANNOTATION SUPPORTED OCCLUDED OBJECT TRACKING
- DESIGN AND ANALYSIS OF LOW POWER MULTIPLY AND ACCUMULATE UNIT USING PIXEL PROPERTIES REUSABILITY TECHNIQUE FOR IMAGE PROCESSING SYSTEMS
- A NEW RECOGNITION TECHNIQUE NAMED SOMP BASED ON PALMPRINT USING NEURAL NETWORK BASED SELF ORGANIZING MAPS
Last modified: 2013-12-05 17:51:28