It has great importance to provide the highest accuracy from fingerprint identification and verification systems, which have a large number of biometric features. Fingerprint recognition systems are more widely utilized than other biometric feature recognition systems. For this reason, a fingerprint recognition system must be fast and reliable to realize the separation of the fake and live fingerprints and provide high accuracy. In this study fingerprint liveness detection system is presented using LivDet2015 dataset. SVM (Support Vector Machine), CNN (Convolutional Neural Network), CNN+SVM methods are used for classification and their performances are compared. Especially the classifying performance of CNN method is analyzed. Before the classification process is performed with SVM, edge enrichment, transformation and feature extraction steps are applied on images as preprocessing steps. The highest accuracy rate is obtained by using CNN-deep learning classifier.