Science AI Journal
Computer Science · Research topic

Open research questions in Image Enhancement Techniques

155 gap statements mined from Image Enhancement Techniques papers in our 4.5M-paper local library, which holds 1,791 papers on the topic — drawn mostly from each paper's own stated research gap, future-work, challenge and limitation notes, and its abstract. The ones listed below are a selection still marked open; each names the study that raised it, with a DOI link where the paper has one.

Choosing where to publish on Image Enhancement Techniques? See the ranked Computer Science journals.

What the literature leaves open

  • Conventional methods are limited by their rigid reliance on simplified assumptions, leading to artifacts and inaccurate transmission estimates. Large kernels fail to capture multi-scale haze structures. Conventional attention cannot adapt to non-uniform haze distribution.

    SpectraMixNet: lightweight multi-scale spectral-frequency attention network for single image dehazing · 2026 · DOI

  • Existing methods, such as gamma correction and histogram equalization, have limitations in terms of detail and structure preservation. Retinex-based methods have limitations in estimating the illumination component. There is a need for a more effective approach to address the challenges of low-light image enhancement.

    Low-frequency SNR-guided CNN-transformer network for high-frequency restoration in low-light image enhancement · 2026 · DOI

  • Existing methods lack adaptability to content in image restoration tasks. The need for a model that can effectively accomplish dehazing, deraining, and desnowing for multi-weather image restoration.

    MARNet: multi-scale aggregated and residual network for adverse weather image restoration · 2026 · DOI

  • The traditional methods are limited by fixed mathematical or physical models.

    DualOadamNet: Dual-Branch Lightweight Network for Underwater Image Processing with Optical-Aware Detail Augmentation · 2026 · DOI

  • The lack of targeted datasets for training and evaluating models under non-uniform illumination conditions. The need to balance visual quality and computational consumption in the design of UNIR-Net.

    UNIR-Net: A novel approach for restoring underwater images with non-uniform illumination using synthetic data · 2025 · DOI

  • Existing scanning strategies are not suitable for shadow removal. Traditional shadow removal methods have limitations, such as quadratic computational complexity or inability to model global information. There is a need for a method that can capture local details while enhancing semantic continuity between them.

    ShadowMamba: State-space model with boundary-region selective scan for shadow removal · 2025 · DOI

  • Low light image enhancement (LLIE) using supervised deep learning is limited by the scarcity of matched low/normal light image pairs.

    Back Projection Generative Strategy for Low and Normal Light Image Pairs With Enhanced Statistical Fidelity and Diversity · 2024 · DOI

  • Existing methods lack robustness in color restoration and texture preservation. Deep learning methods have limitations in handling color distortion and uneven lighting. There is a need for a method that can effectively balance brightness enhancement and color recovery.

    DADNet: Dual-Branch Low-Light Image Enhancement Network Based on Attention Mechanism and Dark Channel Prior · 2026 · DOI

  • In conclusion, while the current advancements in underwater image enhancement show significant promise, future work should focus on developing real-time enhancement models for dynamic underwater environments, integrating multi-modal data for more comprehensive improvements, and exploring hybrid models that combine deep learning with physical models to overcome challenges posed by underwater conditions.

    Advances in underwater image and video enhancement systematic review of conventional and deep learning-based method · 2026 · DOI

  • The proposed framework addresses this limitation by integrating virtual flash simulation with transformer-based feature fusion.

    LUMEN: Low-light Unified Multi-stage Enhancement Network using depth-guided flash, clustering, and attention-based Transformers · 2026

  • Existing deep learning approaches require paired training data and incur substantial computational overhead.

    Bio-Inspired Low-Light Image Enhancement with Large Kernel Convolution and Attention · 2026 · DOI

  • Conventional underwater detection methods rely on handcrafted descriptors that extract a limited set of features.

    FSD-Net: underwater object detection based on frequency and spatial domain feature enhancement · 2026 · DOI

  • The limited local receptive fields of traditional convolution operators make it difficult to handle globally degraded edge information.

    PFENet: Physics-Informed Frequency-Enhanced YOLO for Object Detection in Hazy Scenes · 2026 · DOI

  • Collecting well-aligned pairs of low-and normal-light images is expensive and labor-intensive, - The method relies on pre-trained diffusion models, - The experiments are conducted on a limited number…

    Zero-Shot Low-Light Image Enhancement via Diffusion with Joint Frequency and Spatial Guidance · 2026 · DOI

  • Investigating the use of other types of priors for low-light image enhancement, - Exploring the application of the proposed method to other image restoration tasks, - Developing more efficient and scalable methods for low-light image…

    Zero-Shot Low-Light Image Enhancement via Diffusion with Joint Frequency and Spatial Guidance · 2026 · DOI

  • Investigating denoising modules or noise-aware training strategies, - Investigating Swin Transformer and other attention-based backbones for long-range dependency modeling and hierarchical feature representation in low-light enhancement, - Semantic-aware enhancement could preserve important objects while adjusting background…

    Bio-Inspired Low-Light Image Enhancement with Large Kernel Convolution and Attention · 2026 · DOI

  • Further exploration of frequency–space collaborative modeling for image dehazing, - Investigation of SpectraMixNet's applicability to other low-level vision…

    SpectraMixNet: lightweight multi-scale spectral-frequency attention network for single image dehazing · 2026 · DOI

  • Evaluating CMSR with various VLMs, - Exploring the application of CMSR to other computer vision…

    VLM-driven fine-grained semantic regularization for low-light image enhancement · 2026 · DOI

  • Existing methods focus primarily on salient objects while neglecting other critical discrepancies like color and background context - The lack of fine-grained semantic consistency in low-light image enhancement - The need for a method that leverages pre-trained vision-language models to establish multi-level consistency…

    VLM-driven fine-grained semantic regularization for low-light image enhancement · 2026 · DOI

  • Improving the generalization ability of underwater image enhancement methods. Developing more efficient and effective methods for underwater image processing.

    DualOadamNet: Dual-Branch Lightweight Network for Underwater Image Processing with Optical-Aware Detail Augmentation · 2026 · DOI

  • The computational and memory constraints of drones pose significant challenges to deploying complex image restoration models, - Prior-based methods have limited robustness in complex or diverse…

    Simultaneous Learning Knowledge Distillation for Image Restoration: Efficient Model Compression for Drones · 2025 · DOI

  • Further refinement of the SLKD framework to enhance its efficiency and robustness, - Exploration of other applications of the SLKD framework beyond image…

    Simultaneous Learning Knowledge Distillation for Image Restoration: Efficient Model Compression for Drones · 2025 · DOI

  • The absence of real paired ground-truth data hinders a comprehensive evaluation of its effectiveness in real-world scenarios - Some domain gap may exist when deploying models trained on PUNI to real-world underwater…

    UNIR-Net: A novel approach for restoring underwater images with non-uniform illumination using synthetic data · 2025 · DOI

  • Under hazy weather condition, images captured by electronic imaging devices frequently encounter a number of issues, such as blurring of image details and the poorly defined edges.

    Improving Skip Connection in U-Net Through Fusion Perspective With Mamba for Image Dehazing · 2024 · DOI

  • The brightness enhancement branch is relatively complex and consumes substantial computational resources - The color correction branch learns only a color matrix through the attention mechanism, ignoring other color features - DADNet shows limited generalization…

    DADNet: Dual-Branch Low-Light Image Enhancement Network Based on Attention Mechanism and Dark Channel Prior · 2026 · DOI

Most-cited papers in Image Enhancement Techniques

Most recent work

Find a gap in your own Image Enhancement Techniques sub-topic

This page shows what the Image Enhancement Techniques literature already flags as unresolved. To narrow it to your specific question, search the Research Gap Finder: the search is free with a free account and lists the papers closest to your topic first. Unlocking that topic (50 credits, charged once) fills the comparison table from our 4.5M-paper local library and writes the gaps from its rows.

Open the Research Gap Finder →

Related topics in Computer Science

Online Learning and Analytics
1,151 open questions
AI in Service Interactions
1,014 open questions
Virtual Reality Applications and Impacts
928 open questions
Computational Drug Discovery Methods
925 open questions
Advanced Neural Network Applications
913 open questions
Quantum Computing Algorithms and Architecture
809 open questions
Topic Modeling
712 open questions
Teaching and Learning Programming
708 open questions
Open questions raised by papers listed on this page include: (1) There is a need for a more effective approach to address the challenges of low-light image enhancement (“Low-frequency SNR-guided CNN-transformer network for high-frequency restoration…”, 2026); (2) There is a need for a method that can effectively balance brightness enhancement and color recovery (“DADNet: Dual-Branch Low-Light Image Enhancement Network Based on Attention…”, 2026); (3) The lack of targeted datasets for training and evaluating models under non-uniform illumination conditions (“UNIR-Net: A novel approach for restoring underwater images with non-uniform…”, 2025); (4) There is a need for a method that can capture local details while enhancing semantic continuity between them (“ShadowMamba: State-space model with boundary-region selective scan for shadow…”, 2025). 155 gap statements have been mined from Image Enhancement Techniques papers in our 4.5M-paper local library, which holds 1,791 papers on the topic; the gaps come from whichever of those papers state one. They are mostly the research gaps the authors state and the papers' abstracts, plus future-work, limitations and challenges passages. The ones listed below are a selection still marked open; each names the study that raised it, with a DOI link where the paper has one, so you can read the original claim in context.

Tools for your next paper

Gaps & Literature SearchFind open questions in any field — 120,000+ open gaps mined from our own library.Pre-CheckIs your paper ready? Acceptance odds for top, mid and open/emerging journals, in seconds.Journal FinderA ranked shortlist of target journals from a 17,500-venue index.Duplicate Publication CheckerCheck for prior publication across 7 sources and our local library.Citation GeneratorFreeA DOI or title → APA, MLA, Vancouver, Chicago, BibTeX, RIS. No signup.Graphical Abstract MakerFreeTurn your findings into a colorblind-safe graphical abstract.

How Pre-Check worksHow AI Review works

Compare the category — Honest roundups of the AI research tools, ours listed alongside the alternatives.

Command palette

Jump anywhere, run any action.