An Efficient Hierarchical Clustering Algorithm and its Application for Image Retrieval

Wang Ming Xu; Kang Ling Fang; Hai Ru Zhang

doi:10.4028/www.scientific.net/AMM.44-47.3757

Paper Titles

Microscopic Phase-Field Simulation of the Order-Disorder Transition of Antiphase Domian Boundary Formed between DO₂₂ Phases
p.3736

A Dynamic Sensory Quality Evaluation Model Based on Panel Data and Linguistic Information Processing
p.3741

A Novel Smooth Support Vector Regression Based on CHKS Function
p.3746

The Application of Agent Technique in E-Commerce
p.3752

An Efficient Hierarchical Clustering Algorithm and its Application for Image Retrieval
p.3757

Condenser Fault Diagnosis Based on FNN and Data Fusion
p.3762

Application of Time Series Analysis in the Analysis and Forecasting of Track Geometry Status
p.3767

Study on Data Management of High-Speed Railway Transport Equipment Based on Metadata
p.3771

Calculating Urban Storm Intensity Formula Using Genetic Algorithm
p.3776

HomeApplied Mechanics and MaterialsApplied Mechanics and Materials Vols. 44-47An Efficient Hierarchical Clustering Algorithm and...

An Efficient Hierarchical Clustering Algorithm and its Application for Image Retrieval

Abstract:

Clustering is an efficient and fundamental unsupervised learning algorithm for many vision-based applications. This paper aim at the problems of fast indexing high-dimensional local invariant features of images (e.g. SIFT features) and quick similarity searching of images in a scalable image database by using a hierarchical clustering algorithm. We adopt the hierarchical K-means (HKM) clustering method to build a visual vocabulary tree efficiently on given training data and represent image as a “bag of visual words” which are the leaf nodes of the visual vocabulary tree. For the application of image retrieval, we adopt an usually-used indexing structure called “inverted file” to record the mapping of each visual word to the database images containing that visual word along with the number of times it appears in each image. We propose a weighted voting strategy for the application of content-based image retrieval and achieve desirable performance through experiments.