Outdoor Scene Classification Using Multiple SVM

  • Matheel E. Abdulmunem Department of Computer Science, University of Technology-Iraq
  • Eman Hato Department of Computer Science, Mustansiriyah University, Baghdad, Iraq
Keywords: Gist descriptor, OvA, OvO, RBF kernel, Polynomial kernel, SVM


This paper presents a hierarchical two-stage outdoor scene classification method using multi-classes of Support Vector Machine (SVM). In this proposed method, the gist feature of all the images in the database is extracted first to obtain the feature vectors. The image of database is classified into eight outdoor scenes classes, four manmade scenes and four natural scenes. Second, a hierarchical classification is applied, where the first stage classifies all manmade scene classes against all natural scene classes, while the second stage of a hierarchical classification classifies the outputs of first stage into either one of the four manmade scene classes or natural scene classes. Binary SVM and multi-classes SVMs are employed in the first and second stage of a hierarchical classification respectively. The proposed method is designed also to compare and find the most suitable multi-classes SVMs approach and the kernel function for classification task, where their performances are analyzed based on experimental results. The multi-classes SVMs used in this paper are One-versus-All (OvA) and One-versus-One (OvO), while the kernel functions used are linear kernel, Radius Basis Function (RBF) kernel and Polynomial kernel. Experimental results indicate that OvO classifier provides better performance than OvA classifier. The results, also show that the Polynomial kernel function is superior to others kernel function.

How to Cite
Abdulmunem, M. E., & Hato, E. (2018). Outdoor Scene Classification Using Multiple SVM. Iraqi Journal of Science, 59(4C), 2323-2335. Retrieved from https://ijs.uobaghdad.edu.iq/index.php/eijs/article/view/535
Computer Science