Release Time:2019-03-11  Hits:

Indexed by: Conference Paper

Date of Publication: 2010-08-07

Included Journals: Scopus、CPCI-S、EI

Volume: 7820

Key Words: image classification; bag of words; spatial visual vocabulary; SVM

Abstract: In this paper, we propose an approach to recognize scene categories by means of a novel method named spatial visual vocabulary. Firstly, we hierarchically divide images into sub regions and construct the spatial visual vocabulary by grouping the low-level features collected from every corresponding spatial sub region into a specified number of clusters using k-means algorithm. To recognize the category of a scene, the visual vocabulary distributions of all spatial sub regions are concatenated to form a global feature vector. The classification is obtained using LIBSVM, a support vector machine classifier. Our goal is to find a universal framework which is applicable to various types of features, so two kinds of features are used in the experiments: "V1-like" filters and PACT features. In almost all experimental cases, the proposed model achieves superior results. Source codes are available by email.