An Effective Feature Selection and Classification Technique to Palmprint Biometric Identification Systems

In this paper, we proposed a novel deep-learning technique. In this research, convert the input image to greyscale, crop and resize the image, using CLAHE to enhance the contrast of the palmprint image. After that, extract the features from the image and select important features from extracted features to reduce the time to perform. Finally, employ the Darknet Convolutional Neural Network (DNetCNN) approach to classify the palmprint image, whether matched or not-matched with the Sooty Tern Optimization Algorithm, to improve classification accuracy.