Comparison of ResNet50 Algorithm with AlexNet Algorithm in Precise Biometric Palm Print Recognition

This work's main objective is to replace the AlexNet algorithm with the Novel ResNet50 Algorithm to increase the biometric palm print identification accuracy. A sample size of ten for each group, the novel ResNet50 and Alexnet algorithms were the two study groups in this experiment. Selecting photos from the Kaggle source, 1640 photographs of the left and right hands make up the Mobile Palmprint Database at Birjand University (BMPD). The sample size was calculated using a G Power of 0.8 and a 95% confidence range. The ResNet50 algorithm performs with better recognition by having 95.8% accuracy compared to the AlexNet algorithm having 92.6%. A value of p=0.001 The outcomes demonstrated that the Novel ResNet50 and AlexNet algorithms differed statistically significantly. (Independent sample T-test p<0.05). The results show that ResNet50 recognises biometric palm prints with a greater level of accuracy than AlexNet.