Computational Ghost Imaging Base on Bidirectional Recurrent Neural Network

In recent decades, Ghost Imaging (GI) has emerged as a focal point of research attention among scholars. The imaging quality of GI remains a challenge despite extensive research. In this study, we propose a new model aimed at improving imaging quality. Being inspired by the possible nonlinear connections in GI, we have developed a model based on bidirectional recurrent neural networks (BI-RNN), named GI-BRNN(ghost imaging base on bidirectional recurrent neural network). This model leverages the bidirectional recurrent mechanism and exhibits higher peak signal-to-noise ratio (PSNR) and structural similarity index (SSIM) compared to traditional algorithms. The results of this study demonstrate that the GI-BRNN model significantly improves the quality of ghost imaging. This research provides new insights and approaches for the application of deep learning in the field of GI.