• Corpus ID: 237372148

Perceptually Optimized Deep High-Dynamic-Range Image Tone Mapping

@article{Le2021PerceptuallyOD,
  title={Perceptually Optimized Deep High-Dynamic-Range Image Tone Mapping},
  author={Chenyang Le and Jiebin Yan and Yuming Fang and Kede Ma},
  journal={ArXiv},
  year={2021},
  volume={abs/2109.00180}
}
We describe a deep high-dynamic-range (HDR) image tone mapping operator that is computationally efficient and perceptually optimized. We first decompose an HDR image into a normalized Laplacian pyramid, and use two deep neural networks (DNNs) to estimate the Laplacian pyramid of the desired tone-mapped image from the normalized representation. We then end-to-end optimize the entire method over a database of HDR images by minimizing the normalized Laplacian pyramid distance (NLPD), a recently… 

Figures and Tables from this paper

Perceptual Optimization of a Biologically-Inspired Tone Mapping Operator

TLDR
This paper develops a two-stage neural network-based HDR image TMO that is biologically-inspired, computationally efficient, and perceptually optimized, and fully automatic to tone map uncalibrated HDR images.

References

SHOWING 1-10 OF 27 REFERENCES

High dynamic range image tone mapping by maximizing a structural fidelity measure

TLDR
This paper proposes a substantially different tone mapping approach, where instead of explicitly designing a new computational structure for TMO, the algorithm searches in the space of images to find better quality images in terms of a recent objective measure that can assess the structural fidelity between two images of different dynamic ranges.

A Hybrid l1-l0 Layer Decomposition Model for Tone Mapping

TLDR
A hybrid l1-l0 decomposition model is proposed that achieves visually compelling results with little halo artifacts, outperforming the state-of-the-art tone mapping algorithms in both subjective and objective evaluations.

Adaptive Logarithmic Mapping For Displaying High Contrast Scenes

TLDR
A fast, high quality tone mapping technique to display high contrast images on devices with limited dynamic range of luminance values and taking into account user preference concerning brightness, contrast compression, and detail reproduction is proposed.

Objective Quality Assessment of Tone-Mapped Images

TLDR
An objective quality assessment algorithm for tone-mapped images is proposed by combining: 1) a multiscale signal fidelity measure on the basis of a modified structural similarity index and 2) a naturalness measure onThe basis of intensity statistics of natural images.

Compressing and companding high dynamic range images with subband architectures

TLDR
This work uses a symmetrical analysis-synthesis filter bank, and applies local gain control to the subbands to demonstrate that multi-scale image processing techniques, which are widely used for many image processing tasks, can work when properly implemented.

Consistent tone reproduction

TLDR
This work proposes an efficient global tone reproduction method that achieves robust results across a large variety of HDR images without the need to adjust parameters, which makes this method highly suitable for automated dynamic range compression, which for instance is necessary when a large number of HDR pictures need to be converted.

Perceptually Optimized Image Rendering

TLDR
A framework for rendering photographic images by directly optimizing their perceptual similarity to the original visual scene is developed, yielding results of comparable visual quality to current state-of-the-art methods, but without manual intervention or parameter adjustment.

High dynamic range image rendering with a retinex-based adaptive filter

TLDR
The novelties of the method is first to use an adaptive filter, whose shape follows the image high-contrast edges, thus reducing halo artifacts common to other methods, and only the luminance channel is processed.

Fast bilateral filtering for the display of high-dynamic-range images

We present a new technique for the display of high-dynamic-range images, which reduces the contrast while preserving detail. It is based on a two-scale decomposition of the image into a base layer,

Fast Image Processing with Fully-Convolutional Networks

TLDR
This work presents an approach to accelerating a wide variety of image processing operators using a fully-convolutional network that is trained on input-output pairs that demonstrate the operator’s action, and demonstrates that the presented approach is significantly more accurate than prior approximation schemes.