BL-ASM에서 U-Net 기반 위상 홀로그램의 스펙클 노이즈 감소와 이미지 품질 향상
Speckle Noise Reduction and Image Quality Improvement in U-net-based Phase Holograms in BL-ASM
0
조회
0
다운로드
0
인용
0
참고문헌
키워드
초록
Band-limited angular spectrum method (BL-ASM)는 공간주파수 제어의 문제로 aliasing 오류가 발생한다. 본 논문에서는 위상 홀로그 램에 대한 표본화 간격 조정 기법과 딥 러닝 기반의 U-Net 모델을 사용한 스펙클 노이즈 감소 및 이미지 품질 향상 기법을 제안하였다. 제안한 기법에서는 넓은 전파 범위에서 aliasing 오류를 제거할 수 있도록 먼저 샘플링 팩터를 계산하여 표본화 간격 조절에 의한 공간주파수를 제어 함으로써 스펙클 노이즈를 감소시킨다. 그 후 딥 러닝 모델을 적용한 위상 홀로그램을 학습시켜 복원 이미지의 품질을 향상시킨다. 다양한 샘플 이미지에 대한 S/W 시뮬레이션에서 기존의 BL-ASM과의 peak signal-to-noise ratio (PSNR), structural similarity index measure (SSIM)을 비교할 때 각각 평균 5%, 0.14% 정도 비율이 향상됨을 확인하였다.
Abstract
The band-limited angular spectrum method (BL-ASM) causes aliasing errors due to spatial frequency control problems. In this paper, a sampling interval adjustment technique for phase holograms and a technique for reducing speckle noise and improving image quality using a deep-learning-based U-net model are proposed. With the proposed technique, speckle noise is reduced by first calculating the sampling factor and controlling thespatial frequency by adjusting the sampling interval so that aliasing errors can be removed in a wide range of propagation. The next step is to improve the quality of the reconstructed image by learning the phase hologram to which the deep learning model is applied. In the S/W simulation of varioussample images, it was confirmed that the peak signal-to-noise ratio (PSNR) and structural similarity index measure (SSIM) were improved by 5% and 0.14% on average, compared with the existing BL-ASM.
참고문헌 (13)
- P. Chakravarthula, "Wirtinger holography for near-eye displays", ACM Trans. Graph, 38, pp.213~ (2019)
- Y. K. Kim, "Low-noise high-efficiency double-phase hologram by multiplying a weight factor", Opt. Lett, 44, pp.3649~3652 (2019)
- Y. Zhao, "Fast calculation method of computer-generated hologram using a depth camera with point cloud gridding", Opt. Commun, 411, pp.166~169 (2018)
- L. Shi, "Towards real-time photorealistic 3D holography with deep neural networks", Nature, 591, pp.234~239 (2021)
- Y. Peng, "Neural holography with camera-in-the-loop training", ACM Trans. Graph, 39, pp.185~ (2020)
- J. Wu, "High-speed computergenerated holography using an autoencoder-based deep neural network", Opt. Lett, 46, pp.2908~2911 (2021)
- K. Matsushima, "Band limited angular spectrum method for numerical simulation of free-space propagation in far and near fields", Opt. Express, 17, pp.19662~19673 (2009)
- 차준영, "전파 거리에 따른 위상 홀로그램 복원성능 분석 및 BL-ASM 개선 방안 연구", 방송공학회 논문지, 28(1), pp.3~20 (2023), 한국방송∙미디어공학회[DOI]
- J. Cho, "DIML", CVL RGB-D dataset : 2M RGBD images of natural indoor and outdoor scenes, 2021
- Y. Kim, "Structure selective depth super-resolution for RGB-D cameras", IEEE Trans. Image Process, 25, pp.5527~5538 (2016)
- S. Kim, "Deep stereo confidence prediction for depth estimation", Proc. 2017IEEE International Conference on Image Processing-ICIP, pp.992~996 (2017)
- Y. Kim, "Deep monocular depth estimation via integration of global and local predictions", IEEE Trans. Image Process, 27, pp.4131~4143 (2018)
- J. Cho, "Deep monocular depth estimation leveraging a large-scale outdoor stereo dataset", Expert Syst. Appl, 178, pp.114877~ (2021)