Model-based deep learning for fiber bundle infrared image restoration

Wang, Bowen; Li, Le; Yang, Hai‐Bo; Chen, Jiaxin; Li, Yuhai; Chen, Qian; Zuo, Chao

doi:10.1016/j.dt.2022.12.011

Cited by 5 publications

(2 citation statements)

References 27 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…12,13 However, they use a method of blocking the incident light to obtain different coding masks in these studies, which will cause a significant decrease in the system luminous flux, thus affecting the imaging quality. 14 In this work, to exceed the Nyquist sampling frequency and improve the pixel resolution, we propose a new phase wavefront modulation method that can change the point spread function(PSF) of the imaging system.…”

Section: Introductionmentioning

confidence: 99%

Pixel super resolution imaging technology based on phase wavefront modulation

Yang,

Zhong,

et al. 2024

Sixth Conference on Frontiers in Optical Imaging and Technology: Novel Technologies in Optical Systems

View full text Add to dashboard Cite

In the field of aperture coding super-resolution imaging, an important reason limiting the image super-resolution effect is that aperture coding greatly reduces the incident light flux of the system. Phase wavefront encoding modulation technology can effectively increase the system light flux and improve image super-resolution results. However, using phase wavefront coding instead of aperture coding will cause complex frequency domain calculations. In this paper, the feasibility of the single-seam phase encoding pattern is proved through the mathematical calculation of phase wavefront coding modulation. In the subsequent reconstruction process, the Fourier ptychographic (FP) imaging algorithm is used for verification. The resolvable line pairs of the USAF are increased from Group0 Number2 to Group1 Number3, means the image resolution is approximately doubled. The optical system built based on this phase mask prove the luminous flux of the acquired down-sampled image is increased about more than 5 times, and the image signal-to-noise ratio (SNR) is increased about more than 7 times. This makes it possible for higher super-resolution computational imaging.

show abstract

Section: Introductionmentioning

confidence: 99%

Pixel super resolution imaging technology based on phase wavefront modulation

Yang,

Zhong,

et al. 2024

Sixth Conference on Frontiers in Optical Imaging and Technology: Novel Technologies in Optical Systems

View full text Add to dashboard Cite

show abstract

“…The current mainstream approach for single-frame image super-resolution uses deep learning to learn the exact mapping relationships and generate highresolution images containing rich information. [3][4][5][6][7] Multi-frame image super-resolution reconstructs high-resolution images from multiple frames of low-resolution images with sub-pixel level offsets from each other. Compared with single-frame super-resolution imaging, multi-frame super-resolution imaging recover high-resolution images from low-resolution images with different information in physical scenes, playing a crucial role in application scenarios that pursue high image confidence.…”

Section: Introductionmentioning

confidence: 99%

Wavefront coded-aperture pixel super-resolution imaging based on circular patterns

Liang,

Wang,

Zuo

2023

Third International Computing Imaging Conference (CITA 2023)

View full text Add to dashboard Cite

In multi-frame super-resolution imaging, we need a physical moving device or scanning mechanism to acquire multiple samples of the original scene to recover detailed information about the high-resolution image. However, these imaging systems are usually bulky and complex in structure, relying on a high-precision displacement stage for scanning, which is costly, not easily portable, and difficult to miniaturize. We propose a wavefront coded aperture pixel super-resolution imaging method based on circular patterns, where a variable aperture is placed in front of the imaging lens to achieve modulation of the point spread function, and an incoherent Fourier ptychography imaging algorithm with adaptive step size used to reconstruct high-frequency detail information from a series of low-resolution images. Experimental results show that the imaging resolution can be improved by a factor of 1.41 using the coded aperture super-resolution imaging method. The proposed method only requires a variable aperture placed in front of the imaging lens, without the need for scanning devices or increasing the size and volume of the imaging system and is expected to be applied in super-resolution imaging for miniaturized devices.

show abstract