Palm print recognition technology is one of the most important biometric technologies and has received great attention in terms of scientific research and practical applications. In this article, the technology is based on the ResNet50 algorithm, also known as the Residual Neural Network, and the Convolutional Neural Network (CNN) framework for palm print recognition. First, based on the geometric features of the palm print, the region of interest (ROI) area for the palm print was sculpted. The post-processing ROI is then considered an input to the convolutional neural network. The network can adapt to different palmprints by developing training data. Intermediate Layer Activations for the palm print have also been visualizing and this is a distinct work of the article. The CNN-based palmprint recognition system has an overall accuracy of 98.55 percent, and the model parameters are substantially lower than the standard network model, according to the test findings. It outperforms the typical palmprint recognition method in terms of accuracy.