A Super-Resolution Image Reconstruction Method for a Lensless Microscope

The lensless microscopy technique, with the advantages of low cost and small size, can be widely applied in biotechnology and medicine. In this paper, an innovative image reconstruction scheme is proposed, which is based on the super resolution convolutional neural network (SRCNN). By optimizing the network structure and training strategy, high-quality image reconstruction is achieved for lensless microscopy images. To verify the method's effectiveness, we built a self-built dataset and trained and tested the neural network. The experimental results show that the proposed improved scheme achieves significantly better image reconstruction quality than the original SRCNN network, effectively reduces computational load, and improves reconstruction efficiency.