High dynamic range (HDR) imaging offers an enhanced visual experience by capturing a wider range of real-world luminance levels in digital images. Driven by the increasing demand for high-quality visuals, HDR monitor technology has seen significant advancements. As such monitors become commonplace in both consumer and professional settings, efficient methods are needed for both converting standard dynamic range (SDR) content to HDR-known as reverse tone mapping-and optimizing natural HDR lighting content for display on HDR monitors. A reverse tone mapping procedure aims to produce natural lighting levels, but even on high-end HDR monitors, such images still require adjustment to avoid hard clipping. This paper presents a solution that jointly does both steps: (1) reverse tone mapping to a display-aware HDR representation, and (2) direct generation of an image tailored for a chosen monitor brightness value. We propose a novel neural network architecture conditioned on the target peak brightness via a lightweight multi-layer perceptron (MLP) module injected at the bottleneck, which predicts a bracketed stack of LDR exposures serving as the method's HDR representation. In this manner, the ill-posed tone mapping problem is guided by auxiliary information about display characteristics, improving visual quality. Experiments throughout the full consumer HDR range (100-4000 nits) show consistent improvements over the display-agnostic baseline in peak luminance utilization, local contrast, color and perceptual quality.

Enhancing images to make them visually appealing is a persistent challenge in computer vision. Many deep-learning methods train models on paired datasets to replicate expert editing styles. However, these approaches struggle with two key issues: (1) interpretability and (2) a parametrization suitable for user adjustments. To address these challenges, we present NamedCurves+, an approach inspired by the concept of Color Naming, a universal set of familiar colors widely used in software tools for intuitive editing. Our method integrates color names into a learning-based framework, enabling global adjustments for each named color through tone curves. To address local image variations, we incorporate a transformer block that captures spatial dependencies, enabling context-aware edits across the image. NamedCurves+ enhances the retouching process's interpretability and supports user interaction, allowing flexible modifications of individual tone curves to refine the retouched image according to personal preferences. Extensive experiments on tasks such as image retouching, tone mapping, and exposure correction demonstrate that NamedCurves+ outperforms state-of-the-art methods. Notably, our approach is both explainable, as the tone curves explicitly represent how each color name contributes to the enhancement, and interactive, allowing users to customize the retouching process and achieve results tailored to their liking. Source code and models will be publicly available at: https://namedcurves.github.io.

The xDR dataset: A cinematic natively graded HDR & SDR dataset for evaluation of inverse tone mapping methods

ABSTRACT High dynamic range (HDR) images often need to be rendered on standard dynamic range (SDR) displays, where naïve dynamic‐range compression can cause highlight clipping, contrast distortion, and hue or saturation shifts. This paper proposes a perception‐guided HDR‐to‐SDR framework that combines tone mapping and color correction using perceptual supervision from a cross‐media appearance‐matching experiment. In the experiment, color‐chart patches were presented under three illuminance levels (50, 1000, and 35 000 lx), and observers adjusted their perceived appearance on a calibrated wide‐gamut SDR display. The collected data showed good observer consistency, with mean values of 1.4 and 3.4 for intra‐ and inter‐observer variation, respectively. Based on these data, a compact tone‐compression model was developed using a shared power‐law mapping parameterized by adapting‐white luminance and applied locally through pixel‐wise white estimation. In addition, an appearance‐guided colorfulness correction was introduced to reduce colorfulness drift while preserving hue stability. Quantitative evaluation on two benchmark HDR datasets, comprising 105 images from Fairchild's HDR Photographic Survey and 457 images from LVZ‐HDR, showed that the proposed method performs competitively in terms of TMQI, FSITM, and HDR‐VDP‐3, indicating strong performance in both fidelity and perceptual quality.

Ultrasound diagnostics is a key tool in obstetrics for detecting fetal anomalies and monitoring pregnancy, but image quality often declines in obese patients due to reduced contrast resolution. This pilot study develops and preliminarily validates a novel tone-mapping algorithm for enhancing contrast resolution in high dynamic range (HDR) ultrasound images. The method employs multi-resolution fusion with depth-adaptive weighting and depth compensation to improve contrast, enhance tissue differentiation, reduce noise, and preserve fine details. The algorithm was tested on 20 fetal ultrasound images focused on fetal kidney visualization. Quantitative evaluation showed a 5.4% mean increase in entropy and mean generalized contrast-to-noise ratio (gCNR) improvements of 15.79%, 8.93%, and 17.39% between fetal kidneys and amniotic fluid, far-field objects and fluid, and fetal kidneys and adjacent tissues, respectively, compared with an existing method. These results demonstrate improved anatomical visualization, particularly of fetal kidneys, with potential clinical relevance.

The deep learning revolution has strongly impacted low-level image processing tasks such as style/domain transfer, enhancement/restoration, and visual quality assessments. Despite often being treated separately, the aforementioned tasks share a common theme of understanding, editing, or enhancing the appearance of input images without modifying the underlying content. We leverage this observation to develop a novel disentangled representation learning method that decomposes inputs into content and appearance features. The model is trained in a self-supervised manner and we use the learned features to develop a new quality prediction model named DisQUE. We demonstrate through extensive evaluations that DisQUE achieves state-of-the-art accuracy across quality prediction tasks and distortion types. Moreover, we demonstrate that the same features may also be used for image processing tasks such as HDR tone mapping, where the desired output characteristics may be tuned using example input-output pairs.

A Multi-Stage Adaptive Brightness Adjustment Framework for HDR Image Synthesis

Tone-mapping is a complex process for displaying HDR pictures on standard displays. There are several ways to tone-map pictures, thus it's important to come up with a fair quality measure to choose the best tone-mapping operator (TMO) and adjust its parameters to get the greatest reproduction quality. This is a novel way to objectively assess tone-mapped photos of real-life settings. It combines perceptually meaningful features picked using an acceptable technique. Additionally, the selection highlights the significance of perceptual factors in evaluating tone-mapped HDR video. A number of state-of-the-art criteria and three publicly available datasets are used to assess the feature combination. A different goal is suggested to optimize the picture quality. The foundation of this strategy is the DL-CNN fusion of many perceptually significant characteristics that have been meticulously chosen utilizing adequate high-level features.

All-sky imagers (ASIs) used in solar energy meteorology face an extreme intra-image dynamic range, with the circumsolar neighborhood orders of magnitude brighter than the diffuse dome. Many operational ASI pipelines address this gap with high-dynamic-range (HDR) bracketing inside the camera's image signal processor (ISP), i.e., after demosaicing and color processing in a nonlinear 8-bit RGB domain. Near the Sun, such ISP-domain HDR can down-weight the shortest exposure, retain clipped or near-clipped samples from longer frames, and compress highlight contrast, thereby increasing circumsolar saturation and flattening aureole gradients. A radiance-linear HDR fusion in the sensor/RAW domain (RAW-HDR) is therefore contrasted with the vendor ISP-based HDR mode (ISP-HDR). Solar-based geometric calibration enables Sun-centered analysis. Paired, interleaved acquisitions under clear-sky and broken-cloud conditions are evaluated using two circumsolar performance criteria per RGB channel: (i) saturated-area fraction in concentric rings and (ii) a median-based radial gradient in defined arcs. All quantitative analyses operate on the radiance-linear HDR result; post-merge tone mapping is only used for visualization. Across conditions, ISP-HDR exhibits roughly double the near-saturation within 0-4° of the Sun and about a three- to fourfold weaker circumsolar radial gradient within 0-6° relative to RAW-HDR. These findings indicate that radiance-linear fusion in the RAW domain better preserves circumsolar structure than the examined ISP-domain HDR mode and thus provides more suitable input for downstream tasks such as cloud-edge detection, aerosol retrieval, and irradiance estimation.

The application of digital image processing to medicinal and aromatic leaves has become increasingly important in industries such as pharmaceuticals, cosmetics, and food, as leaf quality directly influences their usability and market value. This study investigates various image enhancement techniques applied to the DIMPSAR and MAP 177 Medicinal Leaf Datasets to improve visual quality, facilitating better feature extraction and analysis. Methods such as contrast enhancement, edge enhancement, sharpening, noise reduction, morphological operations, High Dynamic Range (HDR) enhancement, denoising and restoration, and color enhancement are systematically evaluated based on performance metrics such as Peak Signal-to-Noise Ratio (PSNR), Mean Squared Error (MSE), and Root MSE (RMSE). The results indicate that Linear Contrast Stretching (LCS) and median filtering are the most effective techniques, offering significant contrast and noise reduction improvements while preserving essential structural leaf details. Additionally, techniques such as white balance and unsharp masking enhance image consistency and sharpness, whereas Histogram Equalization (HE) and tone mapping introduce distortions that degrade image quality, limiting their applicability. Lastly, edge detection and morphological operations, primarily used for structure extraction, were found to amplify noise and distortions rather than improve image clarity.

Fast tone mapping operator for high dynamic range image using prior information

A controlled experimental setup used multichannel LED lighting to create HDR scenes. Ten observers performed visual matching tasks between real illuminated scenes and HDR display content across eight lighting conditions. Jzazbz, CIECAM16, and CAM16-UCS were evaluated, analysis using STRESS metrics showed CAM16-UCS achieved the best performance for both lightness and colorfulness predictions. Based on these findings, a tone mapping operator was developed utilizing CAM16-UCS color space with local adaptation and gamma adjustments derived from experimental data. The results demonstrate that CAM16-UCS provides superior color appearance prediction for HDR content and serves as an effective foundation for tone mapping applications.

Tone mapping algorithms are used to compress dynamic range, make image details more conspicuous and generally enhance the image for preference. Global tone mapping manipulates the brightnesses of pixels by applying a single function - or tone curve - to every pixel in the image. Tone curve generation algorithms often constrain the shape of their tone curves and it has been argued that tone curves should be simple, meaning they have one or zero inflexion points. In this work, we investigate whether tone curves should be simplified even further. We present our method which finds the zero inflexion tone curve - which we call a Very Simple (VS) curve - that best approximates a potentially complex tone curve. For the MIT-Adobe FiveK dataset, comprising 25,000 expert tone adjustments, we calculate the best VS approximations and find these curves produce visually similar images compared with more complex counterparts.

This paper proposes a novel image tone-mapping framework that incorporates meta-learning, a psychophysical model, Bayesian optimization, and light-field virtual diffraction. First, we formalize the virtual diffraction process as a mathematical operator defined in the frequency domain to reconstruct high-dynamic-range (HDR) images through phase modulation, enabling the precise control of image details and contrast. In parallel, we apply the Stevens power law to simulate the nonlinear luminance perception of the human visual system, thereby adjusting the overall brightness distribution of the HDR image and improving the visual experience. Unlike existing methods that primarily emphasize structural fidelity, the proposed method strikes a balance between perceptual fidelity and visual naturalness. Secondly, an adaptive parameter tuning system based on Bayesian optimization is developed to conduct optimization of the Tone Mapping Quality Index (TMQI), quantifying uncertainty using probabilistic models to approximate the global optimum with fewer evaluations. Furthermore, we propose a task-distribution-oriented meta-learning framework: a meta-feature space based on image statistics is constructed, and task clustering is combined with a gated meta-learner to rapidly predict initial parameters. This approach significantly enhances the robustness of the algorithm in generalizing to diverse HDR content and effectively mitigates the cold-start problem in the early stage of Bayesian optimization, thereby accelerating the convergence of the overall optimization process. Experimental results demonstrate that the proposed method substantially outperforms state-of-the-art tone-mapping algorithms across multiple benchmark datasets, with an average improvement of up to 27% in naturalness. Furthermore, the meta-learning-guided Bayesian optimization achieves two- to five-fold faster convergence. In the trade-off between computational time and performance, the proposed method consistently dominates the Pareto frontier, achieving high-quality results and efficient convergence with a low computational cost.

Critical details in both bright and dark regions are frequently lost in high dynamic range (HDR) images when they are displayed on low dynamic range (LDR) devices. To mitigate this issue, tone mapping operators (TMOs) have been developed to convert HDR images into LDR representations while maintaining perceptual quality. However, it is challenging to effectively balance various key visual attributes, such as naturalness and structural fidelity. To overcome this limitation, a two-stage Bayesian optimization approach was proposed in this work to enhance the perceptual quality of tone-mapped images across multiple evaluation metrics. The first stage adaptively optimizes TMQI parameters to capture image-specific perceptual characteristics, while the second stage refines the tone mapping function to further improve detail preservation and visual realism. Extensive experiments using three distinct HDR benchmark datasets were conducted, indicating that the proposed method generally performs better than the existing tone mapping techniques across most evaluated metrics, including TMQI, Naturalness, and Structural Fidelity. Our adaptive approach offers a robust and effective solution for optimizing HDR image conversion, resulting in a significantly improved perceptual quality compared to traditional methods.

Abstract Tone mapping is an essential step in an acquisition or a rendering pipeline to map high dynamic range (HDR) content to a reference display range. The simplest tone mapping approach is to apply a function to the luminance channel of an HDR image and then to propagate the change to the red, green, and blue channels. However, this often causes color distortions since luminance and chrominance channels are interdependent, and modifying one affects the other. We propose a novel tone mapping approach that preliminarily decomposes the image into intrinsic components and leverages them to perform the actual operation. This strategy effectively mitigates color distortions, eliminating the need for post-processing color correction required by many state-of-the-art methods, and it also assists tone mapping operators (TMOs), improving the overall image quality. Our method was validated through quantitative metrics and a psychophysical experiment, both demonstrating its effectiveness.

Abstract Capturing the full luminance range of real-world scenes exceeds the capabilities of most digital cameras, often resulting in detail loss, particularly in bright regions. Inverse tone mapping aims to reconstruct High Dynamic Range (HDR) images from Standard Dynamic Range (SDR) inputs, but typically fails to recover clipped details. This paper presents a novel semantic-aware diffusion-based inpainting approach for inverse tone mapping1. Our method introduces two key contributions: (1) a semantic graph-guided diffusion process to inpaint saturated SDR regions, and (2) a principled HDR lifting formulation inspired by traditional HDR bracketing, designed to complement generative inpainting techniques. Experiments demonstrate that our approach outperforms existing methods both quantitatively and qualitatively across multiple datasets.

ABSTRACT In recent years, immersive head-mounted displays (HMDs) have seen increasing use in wide swathes of research and industry domains. In architectural education, lighting design shows particular interest in immersive technologies. While HMDs offer the potential to visualize various daylit environments, the fidelity of the luminous atmosphere of photographic scenes projected in virtual reality (VR) remains less explored. This study aims to evaluate the fidelity of immersive representations in VR HMDs by comparing subjective impressions of daylit iconic buildings with those experienced in VR through 360° stereoscopic photographs. This comparison supports impression evaluation and lighting decisions in daylight design and education. A challenge addressed is assessing the impact of the choice of tone-mapping operators (TMOs) on the subjective impressions in VR. Experiments were conducted in seven iconic buildings involving 225 participants, rating 70 different impressions divided into five aspects. Sixty-three participants evaluated the same scenes in the VR HMD, using three different TMOs. The results indicate that VR HMDs are a reasonable tool to replicate subjective impressions of luminous atmospheres. However, challenges emerge from limitations, such as the headset’s weight, restricted field of view, and the static nature of the image. Significant differences were observed in bodily sensations, possibly attributed to the headset’s weight, and in spatial impressions, due to its restricted field of view. Other atmospheric impressions were less accurately reproduced, likely impacted by the static nature of the image projection. Finally, the study highlights how the choice of TMO influences impressions of light, such as “alternating light and dark.”

When taking images against strong light sources, the resulting images often contain heterogeneous flare artifacts. These artifacts can significantly affect image visual quality and downstream computer vision tasks. While collecting real data pairs of flare-corrupted/flare-free images for training flare removal models is challenging, current methods utilize the direct-add approach to synthesize training data. However, these methods do not consider automatic exposure and tone mapping in the image signal processing pipeline (ISP), leading to the limited generalization capability of deep model training using such data. Besides, existing light source recovery methods hardly recover multiple light sources due to the different sizes, shapes, and illuminance of various light sources. In this paper, we propose a solution to improve the performance of lens flare removal by revisiting the ISP, remodeling the principle of automatic exposure in the synthesis pipeline, and designing a more reliable light source recovery strategy. The new pipeline approaches realistic imaging by discriminating the local and global illumination through a convex combination, avoiding global illumination shifting and local over-saturation. Moreover, the current deep models are only generalized to specific devices due to the diversity of cameras' ISPs. To achieve better generalization on different devices, we formulate the generalization problem as an adversarial training problem and embed an adversarial curve learning (ACL) paradigm in the synthesis pipeline to gain better performance. For recovering multiple light sources, our strategy convexly averages the input and output of the neural network based on illuminance levels, thereby avoiding the need for a hard threshold in identifying light sources. We also contribute a new flare removal testing dataset containing the flare-corrupted images captured by fifteen types of consumer electronics. The dataset facilitates the verification of the generalization capability of flare removal methods. Extensive experiments show that our solution can effectively improve the performance of lens flare removal and push the frontier toward more general situations.

ABSTRACTA new adaptive tone mapping algorithm, which includes global contrast compression, local detail enhancement, and dark region enhancement, is proposed in this paper for HDR images encoded with perceptual quantization (PQ) or hybrid log‐gamma (HLG) transfer functions with the absolute luminance values. In particular, the global contrast compression adopts a luminance mapping curve considering the relationship between the HDR diffuse white and SDR nominal peak luminance, as well as the mapping of the skin tone. The local detail enhancement adopts an enhancement factor to enhance the details while avoiding halo effects. Moreover, the dark region enhancement is performed on nighttime images to enhance the visibility of details in dark regions. When compared to eight popular tone mapping methods, the proposed method is found to achieve a great global contrast, maintain local details, and also enhance great visibility of the details of nighttime images.