Image classification using principal feature analysis
Omid Khayat, Hamid Reza Shahdoosti, Mohammad Hosein Khosravi · International Conference on Artificial Intelligence · 2008
Classification technology is essential for fast retrieval in large database. This paper proposes a combining Principal Feature Analysis (PFA) and SVM model to content-based image retrieval. The proposed method is also used to classification similar images from database. Joint HSV histogram and average entropy computed from gray-level co-occurrence matrices in the localized image region is employed as input vectors. PFA is employed to select feature subsets (choosing principal features) eliminated irrelevant factors as used inputs and to determine the optimal parameters of Support Vector Machine. Experimental results show that the proposed model outperforms existing method.