Pith. sign in

REVIEW 4 major objections 4 minor 36 references

Temperature calibration of surface emissivities with an improved thermal image enhancement network

T0 review · 4 major / 4 minor · reviewed 2026-08-15 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read A single neural enhancement network can correct emissivity-induced temperature errors in infrared images, bringing enhanced temperatures to within about 0.6°C of contact-thermometry ground truth while also improving contrast.

desk verdict A sensible two-material thermography calibration pipeline whose reported accuracy is inflated by GT-anchored rescaling; worth a revised round, not acceptance as is. read the letter →

arxiv 2506.16803 v1 pith:VT7HLNR4 submitted 2025-06-20 eess.IV cs.CV

Ning Chu , Siya Zheng , Shanqing Zhang , Li Li , Caifang Cai , Ali Mohammad-Djafari , Feng Zhao , Yuanbo Song This is my paper · ORCID
classification eess.IVcs.CV
keywords infraredthermographysurfaceemissivitytemperaturecalibrationimageenhancementZero-DCEcross-attentionKLdivergenceindustrialconditionmonitoring
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This paper aims to prove that one neural enhancement network can fix both visual artifacts and temperature errors caused by differing surface emissivities in infrared images. It targets a two-material industrial blower (aluminum and iron), forces the enhanced target region's gray-level statistics to match a reference region, and converts the enhanced gray levels back to temperature through a sensor-specific linear radiation calibration. Reported errors against contact-thermometry ground truth are small (0.57±0.25°C and −0.02±0.51°C across two experiments), and the Euclidean distance of the temperature profile from ground truth drops from 102.5 to 11.2. If right, the method provides a low-cost way to read true surface temperatures from thermal cameras without per-pixel emissivity maps, surface coatings, or contact sensors.

What carries the argument

The load-bearing object is the dual-constrained loss function of Eqs. (7)–(8): $\mathcal{L}_{\mathrm{stat}}$ aligns first and second moments of the target and reference regions, and $\mathcal{L}_{\mathrm{hist}}$ minimizes the symmetric Kullback-Leibler divergence between their gray-level histograms. Around this, the architecture combines emissivity-aware normalization (Eq. 3), cross-attention fusing aluminum (Query) and iron (Key, Value) features (Eq. 4), a seven-layer skip CNN, and an eight-step nonlinear iteration that adapts contrast. The final temperature conversion uses a least-squares fitted linear calibration $G(t) = 0.915t + 0.05$ derived from the sensor's gray-temperature scatter plot.

What would settle it

Run the pipeline on a scene where a contact probe confirms the target region is genuinely warmer than the reference region (e.g., a local hot spot on the aluminum while the iron stays cool). If the enhanced target temperature is pulled toward the iron reference instead of tracking the probe, the statistical-matching assumption fails exactly in the predicted way.

Watch

Extended reading notes

Core claim

The central claim is that statistical alignment—matching the mean, variance, and histogram of the target region to a reference region of the same object—is a sufficient proxy for emissivity correction, once the regions are emissivity-normalized and the result is re-calibrated through a linear gray-temperature relation. The network, an improved Zero-DCE with cross-attention and eight nonlinear iterations, is trained with a dual loss of mean-variance alignment and symmetric KL-divergence histogram matching. On blower sequences taken at two distances and speeds, the enhanced temperatures closely track contact-thermometry ground truth, leading the authors to state that the algorithm achieves both visual and physical corrections.

Load-bearing premise

The target aluminum region and the reference iron region must be at the same true temperature; otherwise forcing their image statistics to match will shift the enhanced temperature toward the reference rather than correcting it.

Editorial extensions

If this is right

  • Temperature maps of multi-material equipment can be obtained from a single thermal camera without contact sensors or emissivity coatings, as long as a same-temperature reference region is visible in the same frame.
  • The same enhancement output serves both visual inspection (higher contrast, higher entropy) and quantitative condition monitoring, so one preprocessing step replaces two separate pipelines.
  • Training requires only about 300 image-temperature pairs, making per-installation recalibration practical for industrial deployments.
  • The method is currently limited to two materials with different emissivities; the authors explicitly state that extending to three or more emissivity values is the intended next step.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • Beyond the paper: the statistical-matching strategy implicitly assumes emissivity is the only between-region difference. A direct test would be to image a single material with two different surface finishes (e.g., painted vs bare) at the same true temperature and confirm the network leaves both regions unaltered.
  • Beyond the paper: extending to ≥3 materials will require a common reference anchor or a pairwise alignment scheme; a natural formulation is aligning all material regions to one calibrated reference with a known temperature.
  • Beyond the paper: since the gray-temperature linear fit is sensor-specific and fitted on the training sequence, a cheap on-site deployment check would be to measure the residual of that linear fit on a fresh scene; the reported training-time MSE of 0.00234 gives a baseline for acceptable deviation.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

4 major / 4 minor

Summary. The paper proposes a deep-learning framework for correcting emissivity-induced temperature distortions in thermal infrared sequences of an industrial blower. The pipeline segments aluminum and iron regions, normalizes them by measured emissivities, fuses features through cross-attention, enhances the target region via a nonlinear iterative CNN, and converts enhanced images to temperatures through a radiation calibration. A dual loss (statistical mean/variance alignment plus KL histogram matching) is used to train the network. The authors claim both visual enhancement and physical temperature correction, supported by two experiments (321 and 263 image-temperature pairs) reporting ERR of 0.57±0.25 °C and -0.02±0.51 °C, and a reduction in Euclidean profile distance from about 102.5 to 11.2 and 84.1 to 8.3.

Significance. If the stated accuracy were established independently, the work would be practically useful: it addresses a real industrial need, combines image enhancement and radiometric calibration in a single trainable framework, and the authors demonstrate that a relatively small dataset (about 300 images) suffices. The paper also includes explicit calibration equations and reports quantitative metrics, which is commendable. However, the current evaluation protocol is confounded by a GT-anchored rescaling, so the central physical-correction claim is not yet supported by the reported numbers.

major comments (4)
  1. [Section 3.3 and Section 4.3] The evaluation rescales the enhanced temperature using GT-derived bounds: Section 3.3 states that the normalized enhanced temperature is 'restored to the original temperature range through the GT temperature,' and Section 4.3 sets the 95% maximum GT temperature as the upper limit and the minimum GT temperature as the lower limit. This injects two per-sequence scalars from the ground truth into every predicted temperature profile, so Tables 4 and 5 measure the network combined with a GT-anchored affine transform, not the network's absolute calibration accuracy. The claim in Section 5 that the algorithm 'achieves both visual and physical corrections' is therefore not supported by these comparisons. Please recompute the errors using the fixed calibration curve in Eq. (10) without GT-bounds rescaling, or report results both with and without the rescaling.
  2. [Section 3.2.1, Eqs. (7)-(8)] The loss function directly enforces equality of the enhanced target region's mean and variance to those of the reference region, and additionally aligns histograms via KL divergence. Since the reference region is treated as the calibration anchor, the small error between enhanced and GT temperatures is partly an artifact of optimizing exactly the statistics that the evaluation then checks. This does not invalidate the method, but it means the reported ERR values are not evidence of independent physical accuracy. Provide a separate evaluation using contact thermometry measurements (e.g., the calibration points in Fig. 9) taken during the dynamic blower runs, rather than relying solely on statistical alignment with the reference region.
  3. [Section 3.3, Eq. (10) and Section 4.3] The calibration curve G(t)=0.915t+0.05 is fitted by least squares to the same sensor data (temperature and gray values) that is used to define GT. Consequently, the radiometric anchor is not independent of the evaluation: any systematic offset in the sensor's grayscale-temperature relationship is absorbed by this fit. Please elaborate how the sensor's absolute calibration was established (e.g., against a blackbody or a calibrated contact thermometer over the full temperature range), or demonstrate that the evaluation conclusions are robust to the fitted parameters.
  4. [Section 3.2.1 and Section 4.2] The homogeneity assumption (aluminum and iron surfaces have the same true temperature) is load-bearing for the loss function, but it is validated at only three surface points over a limited RPM range (Fig. 9-10), while the blower's average temperature rises from 26 °C to 37 °C during the run (Fig. 7). Transient temperature gradients between aluminum and iron are plausible, and if the two surfaces are ever at different true temperatures, the statistical matching will bias the reading rather than correct it. Please provide additional validation of the homogeneity assumption during the dynamic phase, or characterize how such gradients affect the reported error.
minor comments (4)
  1. [Abstract] The abstract contains a typo: 'segments the ROIs of the image and and initially corrected the firing rate' should likely read 'segments the ROIs of the image and initially corrects the gray values' or similar; 'firing rate' seems to be a placeholder or translation error.
  2. [Section 4.3] The text states that a '16×16 neighborhood is randomly selected to calculate the average temperature distribution,' yet Tables 4-5 give deterministic per-sequence distances and errors. Please clarify how the random selection is used and why the reported values are reproducible.
  3. [Table 3] Table 3 lists SSIM, CEI, and Entropy values, but it is unclear what the reference image is for SSIM (original vs. enhanced?) and whether these values are averages over sequence frames. Please state the evaluation protocol and the standard deviations.
  4. [Eq. (5)] The nonlinear iteration in Eq. (5) closely resembles Zero-DCE, but the differences are not discussed. Please clarify what modifications the 'improved Zero-DCE' introduces beyond the cross-attention and the new loss.

Circularity Check

1 steps flagged · score 6.0 of 10

GT-anchored rescaling confounds reported accuracy: Tables 4-5 evaluate an affine mapping fixed by GT bounds, not the network's absolute temperature calibration.

  1. fitted input called prediction [Section 3.3 (Radiation calibration and temperature conversion module), Section 4.3 (Sequence image enhancement results), Tables 4-5, Eq. (11)]
    "Then the normalized enhanced temperature is restored to the original temperature range through the GT temperature. ... To compare the enhanced image temperature with the actual temperature, the normalized enhanced temperature is calibrated using the GT temperature and reduced to the original value. The 95% maximum GT temperature is set as the upper limit and the minimum value is set as the lower limit."

    The enhanced temperature profile that is compared with GT in Tables 4-5 is produced by an affine rescaling whose lower and upper bounds are the minimum and 95th-percentile of the GT temperature sequence itself. Each reported ERR and Euclidean distance therefore measures a composite of the enhancement network and a GT-anchored normalization, not the network's standalone radiometric accuracy. Any monotone output with the correct relative ordering will be forced into the GT range by this rescaling, and the absolute error is artificially bounded by the construction. This is load-bearing because the paper's concluding claim that the algorithm 'achieves both visual and physical corrections' rests on these very comparisons.

full rationale

The central derivation chain is: segment Al and Fe regions; normalize by measured emissivities; train a network with L_stat and L_hist to force the enhanced Al region's mean, variance, and histogram to match the Fe reference; map gray values through a radiometric calibration curve G(t) fitted to the sensor; then 'restore' the normalized enhanced temperature to the original range using the GT minimum and 95% maximum. The final step is the circular element: the prediction is affinely rescaled with parameters taken from the very ground truth it is later compared against. Consequently, the small errors reported (ERR 0.57±0.25 °C and −0.02±0.51 °C; Dis reduction from 102.5 to 11.2) partly reflect this GT-anchored normalization rather than the network's absolute accuracy. The paper's novelty claims and the physical-correction conclusion are therefore supported in part by construction. However, this is not full circularity: the emissivity normalization uses independent ASTM-style measured emissivities, the segmentation and enhancement architecture provide nontrivial learned transformations, the radiometric model Eq. (1) is a standard physical model, and the GT is anchored by contact thermometry at three surface points. The homogeneity assumption (Section 3.2.1) is a modeling assumption that could bias results if Al and Fe true temperatures differ, but it is not itself a circular definition. Thus the appropriate score is 6: one or more reported predictions reduce by construction, yielding partial circularity in the accuracy claim.

Assumptions & free parameters 3 free parameters · 4 assumptions · 0 invented entities

The physical model (Eqs. 1-2) and the emissivity values come from prior literature or advance measurement, which is legitimate input. The load-bearing premise the paper does not independently establish is the homogeneity of the surface temperature between aluminum and iron regions, on which the loss and the calibration transfer rest. The sensor grayscale-temperature fit (Eq. 10) and the GT-derived rescaling bounds are numbers fitted to data that appear in the central accuracy claim. No new physical entities are postulated; the emissivity-aware attention module is purely architectural.

free parameters (3)
  • Sensor grayscale-temperature calibration coefficients a and b = a = 0.915, b = 0.05 (Eq. 10)
    Fitted by least squares to the observed grayscale-temperature scatter (Fig. 8); all temperature readouts, including the claimed accuracy, pass through this linear map.
  • GT-derived temperature bounds for rescaling the enhanced prediction = GT minimum as lower bound, 95th percentile of GT maximum as upper bound
    Section 4.3 maps the normalized enhanced temperature back into the GT temperature range before computing errors against GT, so the comparison scale comes from the evaluation target itself.
  • Iteration count and loss weighting = 8 iterations; L_stat and L_hist combined with equal weight (Eq. 6)
    Hand-chosen hyperparameters that determine enhancement strength and hence the final temperature values; no sensitivity analysis is provided.
assumptions (4)
  • domain assumption Homogeneous internal temperature across aluminum and iron regions with negligible material thickness
    Stated in Section 3.2.1 as the basis for the loss in Eqs. (7)-(8); the statistical matching only corrects emissivity artifacts if both surfaces are truly at the same temperature. Verified at only three calibration points on one blower.
  • domain assumption Infrared radiation model of Eq. (1) with sensor exponent n = 4.09 for HgCdTe
    Adopted from the thermography literature [5] and used to compute GT temperatures from measured signals; standard in the field but not independently validated in this setup.
  • domain assumption Atmospheric transmittance model of Eq. (2) with tabulated water condensation values (Table 1)
    Adopted from reference [35]; the tabulated values shift the transmittance and therefore the GT temperatures used for evaluation.
  • domain assumption The DNLNet segmentation network [14] transfers to blower material masks
    The paper applies a blood vessel segmentation architecture to extract aluminum and iron regions without describing training or fine-tuning, so the mask accuracy is assumed rather than demonstrated.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Temperature calibration of surface emissivities with an improved thermal image enhancement network." pith.science (2026). https://pith.science/paper/VT7HLNR4

@misc{pith2026250616803,
  author       = {Pith},
  title        = {Pith review of: Temperature calibration of surface emissivities with an improved thermal image enhancement network},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/VT7HLNR4}},
  note         = {Machine review of arXiv:2506.16803}
}
read the original abstract

Infrared thermography faces persistent challenges in temperature accuracy due to material emissivity variations, where existing methods often neglect the joint optimization of radiometric calibration and image degradation. This study introduces a physically guided neural framework that unifies temperature correction and image enhancement through a symmetric skip-CNN architecture and an emissivity-aware attention module. The pre-processing stage segments the ROIs of the image and and initially corrected the firing rate. A novel dual-constrained loss function strengthens the statistical consistency between the target and reference regions through mean-variance alignment and histogram matching based on Kullback-Leibler dispersion. The method works by dynamically fusing thermal radiation features and spatial context, and the model suppresses emissivity artifacts while recovering structural details. After validating the industrial blower system under different conditions, the improved network realizes the dynamic fusion of thermal radiation characteristics and spatial background, with accurate calibration results in various industrial conditions.

Figures

Figures reproduced from arXiv: 2506.16803 by the authors.

Figure 1
Figure 1. These figures illustrate the effect of the emissivity of an object’s surface on temperature measurements. In the figure, (a)(b) are two bottles of the same shape and different materials, which contain water of the same temperature but present temperature differences in the infrared image; the blower in (c)(d) are composed of different materials. While the internal temperature is homogeneous, the temperature distribu… view at source ↗
Figure 2
Figure 2. A schematic presentation of the whole processing steps of the proposed method. A sequence of IR images are used as input. They are segmented to obtain RoIs, which are used to focus on two area of the images (composed of Al and Fer). As it is assumed that the inside temperature is the same for the two area, the gray levels of the Al area are modified to have the same statistical properties (means, variance) than the … view at source ↗
Figure 3
Figure 3. Schematic diagram of the image enhancement network structure. This network is an improved Zero-DCE. The input of the network is two images containing only the reference area (𝐼𝐹 𝑒) and only the target area (𝐼𝐴𝑙). A cross attention module is added before the skip convolutional neural network, and the nonlinear iteration module performs adaptive enhancement based on the grayscale of the original target area. (a) iter … view at source ↗
Figures from the paper (5 more)
Figure 8
Figure 8. Figure 8: Image of radiation calibration results.The red line is the regression line of the fitted grayscale-temperature correspondence [PITH_FULL_IMAGE:figures/full_fig_p010_8.png]
Figure 9
Figure 9. Figure 9: Photography of the blower and calibration points [PITH_FULL_IMAGE:figures/full_fig_p010_9.png]
Figure 7
Figure 7. Figure 7: Example plot of sequence temperature. Each data point represents the average temperature of the target area in the sequence image. In this experiment the temperature is increasing from 26 at the startin point of recording to about 37 at the end of recording [PITH_FULL…
Figure 12
Figure 12. Figure 12: Images after running the blower for different times with corresponding enhanced images. (a) Experiment 1 (b) Experiment 2 [PITH_FULL_IMAGE:figures/full_fig_p011_12.png]
Figure 13
Figure 13. Figure 13: Temperature profile of enhanced temperature, GT temperature and sensor reading temperature. The results of the temperature profiles for the two sets of experiments correspond to the two sets of experiments in Fig.12. : Preprint submitted to Elsevier Page 11 of 8 [PIT…

Review history (2 revisions) →

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

36 extracted references · 36 canonical work pages

  1. [1]

    Cordoni, G

    F. Cordoni, G. Bacchiega, G. Bondani, R. Radu, R. Muradore, A multi–modalunsupervisedfaultdetectionsystembasedonpowersig- nals and thermal imaging via deep autoencoder neural network, En- gineering Applications of Artificial Intelligence 110 (2022) 104729

    work page 2022

  2. [2]

    Berg, Detection and tracking in thermal infrared imagery, Ph.D

    A. Berg, Detection and tracking in thermal infrared imagery, Ph.D. thesis, Linköping University Electronic Press (2016)

    work page 2016

  3. [3]

    D.Lopez-Perez,J.Antonino-Daviu,Applicationofinfraredthermog- raphy to failure detection in industrial induction motors: case stories, IEEE Transactions on Industry Applications 53 (3) (2017) 1901– 1908

    work page 2017

  4. [4]

    Resendiz-Ochoa, D

    E. Resendiz-Ochoa, D. A. Elvira-Ortiz, A. Y. Jaen-Cuellar, J. J. Saucedo-Dorantes,Non-invasivediagnosismethodologybasedonin- frared thermography and deep feature learning for detecting multiple faults in electromechanical systems, Measurement (2025) 117076

    work page 2025

  5. [5]

    J. Shen, Y. Zhang, T. Xing, The study on the measurement accuracy of non-steady state temperature field under different emissivity using infrared thermal image, Infrared Physics & Technology 94 (2018) 207–213

    work page 2018

  6. [6]

    Z. Li, W. Chen, Progress in dynamic emissivity regulation: control methods, material systems, and applications, Materials Chemistry Frontiers 5 (17) (2021) 6315–6332

    work page 2021

  7. [7]

    N. Chu, X. Yan, Y. Zhong, L. Wang, L. Yu, C. Cai, A. Mohammad- Djafari,Anefficienttemperaturecalibrationmethodbasedontheim- provedinfraredforwardmodelandbayesianinference,IEEESensors Journal (2024)

    work page 2024

  8. [8]

    G.Ulutas,B.Ustubioglu,Underwaterimageenhancementusingcon- trast limited adaptive histogram equalization and layered difference representation,MultimediaToolsandApplications80(2021)15067– 15091

    work page 2021

Show all 36 references
  1. [9]

    W. Liu, P. Zhao, Y. Zhao, Y. Fu, J. Dai, L. Zhou, A reflectance- correctionretinexframeworkforthermalimageenhancementinnon- destructivedefectdetectionofcfrp,Measurement237(2024)115070

    2024

  2. [10]

    J. Y. Chiang, Y.-C. Chen, Underwater image enhancement by wave- length compensation and dehazing, IEEE transactions on image pro- cessing 21 (4) (2011) 1756–1769

    2011

  3. [11]

    X. Yi, H. Xu, H. Zhang, L. Tang, J. Ma, Diff-retinex: Rethinking low-light image enhancement with a generative diffusion model, in: ProceedingsoftheIEEE/CVFInternationalConferenceonComputer Vision, 2023, pp. 12302–12311

    2023

  4. [12]

    Hambarde, S

    P. Hambarde, S. Murala, A. Dhall, Uw-gan: Single-image depth estimation and image enhancement for underwater images, IEEE Transactions on Instrumentation and Measurement 70 (2021) 1–12

    2021

  5. [13]

    R.Cong,W.Yang,W.Zhang,C.Li,C.-L.Guo,Q.Huang,S.Kwong, Pugan: Physical model-guided underwater image enhancement using gan with dual-discriminators, IEEE Transactions on Image Process- ing 32 (2023) 4472–4485

    2023

  6. [14]

    J.Ni,J.Wu,A.Elazab,J.Tong,Z.Chen,Dnl-net:deformednon-local neuralnetworkforbloodvesselsegmentation,BMCMedicalImaging 22 (1) (2022) 109

    2022

  7. [15]

    ASTM International, ASTM E1933-14: Standard Test Methods for MeasuringandCompensatingforEmissivityUsingInfraredImaging Radiometers (2018)

    2018

  8. [16]

    S. Kim, W. Kang, E. Lee, J. Paik, Wavelet-domain color image enhancement using filtered directional bases and frequency-adaptive shrinkage,IEEETransactionsonConsumerElectronics56(2)(2010) 1063–1070

    2010

  9. [17]

    3076–3085

    X.Deng,R.Yang,M.Xu,P.L.Dragotti,Waveletdomainstyletransfer for an effective perception-distortion tradeoff in single image super- resolution,in:ProceedingsoftheIEEE/CVFinternationalconference on computer vision, 2019, pp. 3076–3085

    2019

  10. [18]

    K. Xu, X. Yang, B. Yin, R. W. Lau, Learning to restore low-light images via decomposition-and-enhancement, in: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, 2020, pp. 2281–2290

    2020

  11. [19]

    Zhang, L

    W. Zhang, L. Dong, T. Zhang, W. Xu, Enhancing underwater image via color correction and bi-interval contrast enhancement, Signal Processing: Image Communication 90 (2021) 116030

    2021

  12. [20]

    Jiang, Y

    Q. Jiang, Y. Zhang, F. Bao, X. Zhao, C. Zhang, P. Liu, Two- step domain adaptation for underwater image enhancement, Pattern Recognition 122 (2022) 108324

    2022

  13. [21]

    C. Guo, C. Li, J. Guo, C. C. Loy, J. Hou, S. Kwong, R. Cong, Zero- referencedeepcurveestimationforlow-lightimageenhancement,in: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, 2020, pp. 1780–1789. :Preprint submitted to Elsevier Page 7 of 8

    2020

  14. [22]

    L. Ji, Q. Zhu, Y. Zhang, J. Yin, R. Wei, J. Xiao, D. Xiao, G. Zhao, Cross-domainheterogeneousresidualnetworkforsingleimagesuper- resolution, Neural Networks 149 (2022) 84–94

    2022

  15. [23]

    Voronin, O

    V. Voronin, O. Tokareva, N. Gapon, I. Naumov, A. Zelensky, E. Se- menishchev, Image haze removal using block-based multiscale en- hancementmethod,in:MultimodalImageExploitationandLearning 2023, Vol. 12526, SPIE, 2023, pp. 216–221

    2023

  16. [24]

    J. Hao, Y. Du, S. Wang, W. Ren, Infrared image enhancement al- gorithm based on wavelet transform and improved bilateral filtering, Infrared Technology 46 (9) (2024) 1051

    2024

  17. [25]

    Xiong, Z

    Z. Xiong, Z. Sheng, Y. Mao, Feature multi-scale enhancement and adaptive dynamic fusion network for infrared small target detection, Remote Sensing (2025)

    2025

  18. [26]

    Rehman, T

    A. Rehman, T. Mahmood, F. S. Alamri, T. Saba, S. Naseem, Ad- vanced feature learning and classification of microscopic breast ab- normalities using a robust deep transfer learning technique, Mi- croscopy Research and Technique (2024)

    2024

  19. [27]

    Zhuang, J

    P. Zhuang, J. Wu, F. Porikli, C. Li, Underwater image enhancement with hyper-laplacian reflectance priors, IEEE Transactions on Image Processing 31 (2022) 5442–5455

    2022

  20. [28]

    L. Hu, M. Qin, F. Zhang, Z. Du, R. Liu, Rscnn: A cnn-based method to enhance low-light remote-sensing images, Remote Sensing 13 (1) (2020) 62

    2020

  21. [29]

    Huang, H

    H. Huang, H. Tao, H. Wang, A convolutional neural network based method for low-illumination image enhancement, in: Proceedings of the 2nd international conference on artificial intelligence and pattern recognition, 2019, pp. 72–77

    2019

  22. [30]

    G. Nie, Z. Zhang, Z. Jiao, Y. Li, M. Shao, X. Dai, A novel intelligent bearing fault diagnosis method based on image enhancement and improved convolutional neural network, Measurement 242 (2025) 116148

    2025

  23. [31]

    W. Zhou, X. Lin, J. Lei, L. Yu, J.-N. Hwang, Mffenet: Multiscale feature fusion and enhancement network for rgb–thermal urban road scene parsing, IEEE Transactions on Multimedia 24 (2022) 2526– 2538

    2022

  24. [32]

    Liang, Y

    T. Liang, Y. Jin, W. Liu, Y. Li, Cross-modality transformer with modality mining for visible-infrared person re-identification, IEEE Transactions on Multimedia 25 (2023) 8432–8444

    2023

  25. [33]

    Zhang, W

    H. Zhang, W. Qian, M. Wan, K. Zhang, Infrared image enhancement algorithm using local entropy mapping histogram adaptive segmen- tation, Infrared Physics & Technology 120 (2022) 104000

    2022

  26. [34]

    Z. Pang, G. Liu, G. Li, J. Gong, C. Chen, C. Yao, An infrared image enhancement method via content and detail two-stream deep convolutional neural network, Infrared Physics & Technology 133 (2023) 104761

    2023

  27. [35]

    Jha, Transmission Characteristics of IR Signals in Atmosphere, John Wiley & Sons, Hoboken, NJ, USA, 2000

    A. Jha, Transmission Characteristics of IR Signals in Atmosphere, John Wiley & Sons, Hoboken, NJ, USA, 2000

    2000

  28. [36]

    (a) (b) (c) (d) Figure 1:These figures illustrate the effect of the emissivity of an object’s surface on temperature measurements

    S.Kullback,R.A.Leibler,Oninformationandsufficiency,Theannals of mathematical statistics 22 (1) (1951) 79–86. (a) (b) (c) (d) Figure 1:These figures illustrate the effect of the emissivity of an object’s surface on temperature measurements. In the figure, (a)(b) are two bottles...

    1951

Pith tools

Reviewed August 15, 2026 · model on record in the stance chip above.