Classification of Wild Animals based on SVM and Local Descriptors

Abstract

In this paper, a novel method for object recognition based on hybrid local descriptors is presented. This method utilizes a combination of a few approaches (SIFT - Scale-invariant feature transform, SURF - Speeded Up Robust Features) and consists of second parts. The applicability of the presented hybrid methods are demonstrated on a few images from dataset. Dataset classes represent big animals situated in Slovak country, namely wolf, fox, brown bear, deer and wild boar. The presented method may be also used in other areas of image classification and feature extraction. The experimental results show, that the combination of local descriptors has a positive effect for object recognition.