Learning-based lensless imaging through optically thick scattering media

Meng Lyu; Hao Wang; Guowei Li; Shanshan Zheng; Guohai Situ

doi:doi:10.1117/1.AP.1.3.036002

Advanced Photonics, 2019, 1 (3): 036002, Published Online: Jun. 19, 2019

Learning-based lensless imaging through optically thick scattering media Download： 1087次

Meng Lyu ^1,2Hao Wang ^1,2Guowei Li ^1,2Shanshan Zheng ^1,2Guohai Situ ^1,2,*

Author Affiliations

¹ Chinese Academy of Sciences, Shanghai Institute of Optics and Fine Mechanics, Shanghai, China

² University of Chinese Academy of Sciences, Center for Materials Science and Optoelectronics Engineering, Beijing, China

Figures & Tables

Fig. 1. (a) Experimental setup for imaging through scattering media, SLM represents an amplitude-only SLM, P1 and P2 are linear polarizers and the slab is a 3-mm-thick white polystyrene. The images captured at the (b) front and (c) back surfaces of the scattering medium. (d) The side view and (e) the top view of the polystyrene.

下载图片查看原文

Fig. 2. The diagram of the proposed HNN model.

下载图片查看原文

Fig. 3. The reconstructed results. (a) The speckle patterns ( $64 \times 64$ pixels) cropped from the raw acquired scattered pattern ( $512 \times 512$ pixels), (b) the reconstructed images by using the proposed HNN, (c) the ground-truth images, and (d) the reconstructed images by using memory effect.

下载图片查看原文

Fig. 4. Comparison of reconstruction performance. (a) The speckle patterns cropped from the raw scattered patterns. The images reconstructed by (b) the HNN, (c) a DNN, and (d) a CNN. (e) The ground-truth images.

下载图片查看原文

Fig. 5. The positions of the three $64 \times 64 -$ pixel blocks randomly selected from the acquired scattered patterns of $512 \times 512$ pixels.

下载图片查看原文

Fig. 6. The images reconstructed by using the subspeckle pattern located at A1, A2, and A3.

下载图片查看原文

Fig. 7. The images of handwritten digits and English letters reconstructed from the randomly selected pixels. (a) The subspeckle patterns formed by the randomly selected 3096 pixels from the $512 \times 512$ raw scattered patterns, (b) the reconstructed images, and (c) the ground-truth images.

下载图片查看原文

Fig. 8. The images of handwritten digits and English letters reconstructed from the subspeckle patterns of six different sizes.

下载图片查看原文

Fig. 9. The result of the digits and English letters with different gray intervals. The images in the first row are the speckle images with different gray-level intervals, the second row shows the predicted objects by the HNN model, images in the third row are the ground-truth images. (G1) Images with gray value interval (25,000 to 30,000), (G2) images with gray value interval (35,000 to 40,000), (G3) images with gray value interval (40,000 to 45,000), and (G4) images with gray value threshold of 35,000.

下载图片查看原文

Fig. 10. The setup of the system to measure the optical depth.

下载图片查看原文

Meng Lyu, Hao Wang, Guowei Li, Shanshan Zheng, Guohai Situ. Learning-based lensless imaging through optically thick scattering media[J]. Advanced Photonics, 2019, 1(3): 036002.

Learning-based lensless imaging through optically thick scattering media Download： 1087次

Fig. 2. The diagram of the proposed HNN model.

Fig. 4. Comparison of reconstruction performance. (a) The speckle patterns cropped from the raw scattered patterns. The images reconstructed by (b) the HNN, (c) a DNN, and (d) a CNN. (e) The ground-truth images.

Fig. 5. The positions of the three $64 \times 64 -$ pixel blocks randomly selected from the acquired scattered patterns of $512 \times 512$ pixels.

Fig. 6. The images reconstructed by using the subspeckle pattern located at A1, A2, and A3.

Fig. 7. The images of handwritten digits and English letters reconstructed from the randomly selected pixels. (a) The subspeckle patterns formed by the randomly selected 3096 pixels from the $512 \times 512$ raw scattered patterns, (b) the reconstructed images, and (c) the ground-truth images.

Fig. 8. The images of handwritten digits and English letters reconstructed from the subspeckle patterns of six different sizes.

Fig. 10. The setup of the system to measure the optical depth.

关于本站 Cookie 的使用提示

全站搜索

Learning-based lensless imaging through optically thick scattering media Download： 1087次

Fig. 2. The diagram of the proposed HNN model.

Fig. 3. The reconstructed results. (a) The speckle patterns (64×64 pixels) cropped from the raw acquired scattered pattern (512×512 pixels), (b) the reconstructed images by using the proposed HNN, (c) the ground-truth images, and (d) the reconstructed images by using memory effect.

Fig. 4. Comparison of reconstruction performance. (a) The speckle patterns cropped from the raw scattered patterns. The images reconstructed by (b) the HNN, (c) a DNN, and (d) a CNN. (e) The ground-truth images.

Fig. 5. The positions of the three 64×64-pixel blocks randomly selected from the acquired scattered patterns of 512×512 pixels.

Fig. 6. The images reconstructed by using the subspeckle pattern located at A1, A2, and A3.

Fig. 7. The images of handwritten digits and English letters reconstructed from the randomly selected pixels. (a) The subspeckle patterns formed by the randomly selected 3096 pixels from the 512×512 raw scattered patterns, (b) the reconstructed images, and (c) the ground-truth images.

Fig. 8. The images of handwritten digits and English letters reconstructed from the subspeckle patterns of six different sizes.

Fig. 10. The setup of the system to measure the optical depth.

相关论文

相关资讯

关于本站 Cookie 的使用提示

全站搜索

Fig. 5. The positions of the three $64 \times 64 -$ pixel blocks randomly selected from the acquired scattered patterns of $512 \times 512$ pixels.

Fig. 7. The images of handwritten digits and English letters reconstructed from the randomly selected pixels. (a) The subspeckle patterns formed by the randomly selected 3096 pixels from the $512 \times 512$ raw scattered patterns, (b) the reconstructed images, and (c) the ground-truth images.