Acta Optica Sinica, Volume. 40, Issue 16, 1611003(2020)
Lens-Free Imaging Method Based on Generative Adversarial Networks
Fig. 1. Lens-free imaging experimental device and light path. (a) Diagram of experimental device; (b) diagram of light path
Fig. 2. Flow chart of lens-free imaging processing
Fig. 3. Calculation principle of defocus distance
Fig. 4. Object image reconstructed by back-propagation
Fig. 5. Diagram of GAN principle
Fig. 6. Main components of GAN. (a) GN; (b) DN
Fig. 7. Image processing results based on GAN. (a) GAN input; (b) partial enlarged image of
Fig. 8. Apical bud slitting images. (a) Lens-free image; (b) reconstructed object plane image through back-propagation; (c) GAN result; (d) microscope image
Fig. 9. PSNR. (a) Reconstructed object plane image through back-propagation; (b) CNN result; (c) GAN result; (d) microscope image
Fig. 10. Performance comparison between GAN and CNN
Get Citation
Copy Citation Text
Chao Zhang, Tao Xing, Zizhen Liu, Haokun He, Hua Shen, Yinxu Bian, Rihong Zhu. Lens-Free Imaging Method Based on Generative Adversarial Networks[J]. Acta Optica Sinica, 2020, 40(16): 1611003
Category: Imaging Systems
Received: Jan. 14, 2020
Accepted: May. 18, 2020
Published Online: Aug. 7, 2020
The Author Email: Shen Hua (edward_bayun@163.com)