Ghost-Free High Dynamic Range Imaging via Rank Minimization

Lee, Chul; Li, Yuelong; Monga, Vishal

doi:10.1109/lsp.2014.2323404

Cited by 94 publications

(20 citation statements)

References 24 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…When considering the background as a low-dimensional subspace, LR models achieve superior performance in background estimation and foreground subtraction tasks. Based on the assumption that the CRF is linear, Oh et al [14] and Lee [15] introduced the rank-1 constraint to the LR model for motion detection in the irradiance domain and used the estimated sparse error as the weight to fuse the irradiance images. By using the truncated nuclear norm to better estimate the nuclear norm, Oh et al [16] and Lee and Lam [17] proposed the LRMC model to directly estimate all of the background irradiances, which are directly used in fusion tasks.…”

Section: Related Workmentioning

confidence: 99%

“…The LR model was first applied in HDR imaging for global registration and inconsistent region detection [14], [15] and was then extended to recover the latent background irradiance [16], [17]. However, these methods rely on the assumption of a linear camera response function (CRF) and fail in very highly saturated regions, e.g., extremely dark or bright regions, where linear CRF breaks down.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Exposure Fusion for Dynamic Scenes Combining Retinex Theory and Low-Rank Matrix Completion

Zhang

Wan

2019

IEEE Access

View full text Add to dashboard Cite

Exposure fusion (EF) directly generates a low dynamic range (LDR) image with similar image quality to the high dynamic range (HDR) image by combining images of different exposures to overcome the limited dynamic range of common digital cameras. EF for dynamic scenes is challenging not only due to the difficulty in removing various ghost artifacts but also because of the requirement of preserving details, especially in the saturated regions. Recently, low-rank (LR) matrix completion (LRMC) has been shown to be efficient in separating the latent background from the sparse motion in the irradiance domain. However, the performance of LRMC models strongly relies on the estimated irradiance images. To address this problem, this paper proposes a novel EF method combining Retinex theory and the LRMC model. The proposed method consists of four steps. First, according to Retinex theory, an image is decomposed into illumination and reflection components. Second, the LRMC model is applied to the reflection component to generate the background reflection component and the sparse error. Third, the motion map is modeled as a Markov random field (MRF), integrating the sparse error with the ordering constraint across all the illumination components. Finally, all the illumination components are fused via a pyramid-based method, where the weight maps are defined based on the obtained motion map and illumination. The experimental results show that the proposed method outperforms the state-of-the-art methods particularly in preserving details in saturated regions.

show abstract

Section: Related Workmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Exposure Fusion for Dynamic Scenes Combining Retinex Theory and Low-Rank Matrix Completion

Zhang

Wan

2019

IEEE Access

View full text Add to dashboard Cite

show abstract

“…Subsequently, there are broad literatures on noise reduction, 8,9 exposure setting, 10-12 image alignment, 4-7 and ghosting removal [13][14][15][16] for HDR imaging. It sequentially takes multiple LDR images of the same scene with different exposures and then merges them into a single HDR image.…”

Section: Multishot High-dynamic-range Imagingmentioning

confidence: 99%

“…48 Without proper treatments in dynamic objects, ghosting artifacts will appear in the produced HDR images. 49 Based on the assumption that the underlying background is static, HDR reconstruction can be formulated into a rank minimization problem, 16,19 which represents all dynamic objects as a sparse matrix. One can handle ghosting artifacts by user corrections, 13 iteratively assigning smaller weights to pixels that are likely to correspond to dynamic objects, 14 or producing ghost-free HDR images with a joint bilateral filter approach.…”

Section: Multishot High-dynamic-range Imagingmentioning

confidence: 99%

Penrose high-dynamic-range imaging

Bai

Lin

et al. 2016

J. Electron. Imaging

View full text Add to dashboard Cite

High-dynamic-range (HDR) imaging is becoming increasingly popular and widespread. The most common multishot HDR approach, based on multiple low-dynamic-range images captured with different exposures, has difficulties in handling camera and object movements. The spatially varying exposures (SVE) technology provides a solution to overcome this limitation by obtaining multiple exposures of the scene in only one shot but suffers from a loss in spatial resolution of the captured image. While aperiodic assignment of exposures has been shown to be advantageous during reconstruction in alleviating resolution loss, almost all the existing imaging sensors use the square pixel layout, which is a periodic tiling of square pixels. We propose the Penrose pixel layout, using pixels in aperiodic rhombus Penrose tiling, for HDR imaging. With the SVE technology, Penrose pixel layout has both exposure and pixel aperiodicities. To investigate its performance, we have to reconstruct HDR images in square pixel layout from Penrose raw images with SVE. Since the two pixel layouts are different, the traditional HDR reconstruction methods are not applicable. We develop a reconstruction method for Penrose pixel layout using a Gaussian mixture model for regularization. Both quantitative and qualitative results show the superiority of Penrose pixel layout over square pixel layout.

show abstract

“…As an alternative to the use of graduated filters, digital image postprocessing has become popular and convenient as it generates High Dynamic Range (HDR) images by blending several images with different exposures 14,15 . However, this technique may cause errors when moving objects come into play, is time consuming, and demands excessive computing power.…”

Section: Introductionmentioning

confidence: 99%

Tunable graduated filters based on electrochromic materials for spatial image control

Hein

Kortz

Oesterschulze

2019

Sci Rep

View full text Add to dashboard Cite

Passive graduated filters with fixed absorption profile are currently used in image recording to avoid overexposure. However, a whole set of filters with prescribed gradients is required to cope with changing illumination conditions. Furthermore, they demand mechanical adjustment during operation. To overcome these deficiencies we present a microfabricated active electrochromic graduated filter which combines multiple functionalities: The overall absorbance, the position of medium transmission as well as the magnitude of its gradient can be tuned continuously by electrical means. Live image control is possible using low operation voltages in the range of ±2 V to reach a high change in optical density ΔOD of 1.01 (400 nm to 780 nm) with a coloration and bleaching time 1.3 s and 0.2 s, respectively. Owing to their low volume and power consumption they are suitable for widespread applications like in smartphones, surveillance cameras or microscopes.

show abstract

Ghost-Free High Dynamic Range Imaging via Rank Minimization

Cited by 94 publications

References 24 publications

Exposure Fusion for Dynamic Scenes Combining Retinex Theory and Low-Rank Matrix Completion

Exposure Fusion for Dynamic Scenes Combining Retinex Theory and Low-Rank Matrix Completion

Penrose high-dynamic-range imaging

Tunable graduated filters based on electrochromic materials for spatial image control

Contact Info

Product

Resources

About