In many areas where personal identification is important, security is of great importance. Biometric or multi-biometric systems, which include the physiological and behavioral features of individuals, are more preferred because traditional methods are insufficient and cannot provide security. In the study, a new approach of multimodal biometric identification is proposed consisting of the fingerprint and finger knuckle print (FKP). A deep convolutional neural network (CNN) based hybrid feature extraction technique is utilized along with the GLCM (Gray Level Co-occurrence Matrix) and wavelet moments. Classification of extracted features is achieved by using random forest classifier with simulation results in terms of F-Score, accuracy, sensitivity,precision, and specificity

CNN, FKP, GLCM, Random Forest Classifier, Wavelet Moments