TY  - JOUR
T1  - ANN and SVM Based Image Classification Using Wavelet Decomposition
AU - , V. Devendran AU - , Amitabh Wahi AU - , Hemalatha Thiagarajan 
JO  - Asian Journal of Information Technology
VL  - 6
IS  - 11
SP  - 1174
EP  - 1180
PY  - 2007
DA  - 2001/08/19
SN  - 1682-3915
DO  - ajit.2007.1174.1180
UR  - https://makhillpublications.co/view-article.php?doi=ajit.2007.1174.1180
KW  - Artificial neural networks
KW  -image classification
KW  -multi-level wavelet decomposition
KW  -support vector machines
AB  - This study discusses the comparative study of ANN and SVM in the classification problem domain. We are classifying the static images into two categories i.e., images with cars and images with no cars. This is tested without any segmentation processes. Feed forward neural networks with backpropagation algorithm are used for paradigm of ANN. On the other hand, support vector machine with radial basis kernel function with p = 5 is used for the same image classification problem. Comparative results show that the classification rate of SVM is far better than the classification rate of ANN. Features are extracted from the static images using Multi-level wavelet decomposition method. The performance of ANN and SVM is tested in 5 different levels of decomposition. This research is carried out using real world data set.
ER  -