NUCLEAR TECHNIQUES, Volume. 48, Issue 4, 040402(2025)
Effect of datasets on imaging quality of coded aperture γ camera based on convolutional neural network algorithm
Fig. 2. (a) 11-dimensional extended coded aperture mask (the central part is the base coded aperture mask), (b) Schematic diagram of the point-source model imaging
Fig. 3. Schematic diagram of dataset construction process (color online)
Fig. 4. Schematic diagram of the CNN training model architecture (color online)
Fig. 5. CNR of different energy sources forming projections on the detector (color online)
Fig. 6. Relationship between CNR and capacity of datasets for orphan point-source
Fig. 7. Variations of reconstructed CNR of the model with different datasets varies with the imaging time
Fig. 8. Reconstruction results of diverse datasets models at different times when the source is at the center of the FOV (color online)
Fig. 9. Relationship between CNR and capacity of datasets for multi-point source
Fig. 10. Reconstruct results of different number of sources (circle is the actual location of sources, color online)
Fig. 11. Reconstruction results of different energy sources exist simultaneously (color online)
|
Get Citation
Copy Citation Text
Wenrui XU, Yushou SONG, Chunzhi ZHOU, Yingwei HOU, Huilan LIU. Effect of datasets on imaging quality of coded aperture γ camera based on convolutional neural network algorithm[J]. NUCLEAR TECHNIQUES, 2025, 48(4): 040402
Category: NUCLEAR ELECTRONICS AND INSTRUMENTATION
Received: Sep. 10, 2024
Accepted: --
Published Online: Jun. 3, 2025
The Author Email: Yushou SONG (宋玉收), Chunzhi ZHOU (周春芝)