Pith. sign in

REVIEW 4 major objections 5 minor 49 references

Predicting Asphalt Pavement Friction Using Texture-Based Image Indicator

T0 review · 4 major / 5 minor · reviewed 2026-08-06 · deepseek-v4-flash

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

Pith's one-line read A single photo predicts pavement friction with adjusted R² above 0.90

desk verdict Modest applied extension of threshold-based image indicators for pavement friction, with in-sample fits overstated as predictions and the core brightness-to-protrusion mapping left unvalidated. read the letter →

arxiv 2507.03559 v1 pith:ATJPUNRF submitted 2025-07-04 cs.CV eess.IV

Bingjie Lu , Zhengyang Lu , Yijiashun Qi , Hanzhe Guo , Tianyao Sun , Zunduo Zhao This is my paper
classification cs.CVeess.IV
keywords pavementfrictionskidresistancesurfacetextureimageprocessingaggregateprotrusionareathresholdsegmentationasphaltDynamicTester
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 tries to establish that a single two-dimensional image of an asphalt surface can estimate the friction a tire will feel, using a simple interpretable measure: the total area of aggregate particles that protrude above the surrounding asphalt. The authors call this the aggregate protrusion area and show that it shrinks as tires polish the surface, tracking Dynamic Friction Tester readings closely enough that per-surface linear models reach adjusted $R^2$ values above 0.90 for all three pavement types tested. If the claim holds, skid-resistance checks could become cheap enough for mix design and routine maintenance without laser scanners or skilled testing crews. The paper does not claim a universal model; each pavement type gets its own calibration, and the thresholds need recalibration under field conditions.

What carries the argument

The machinery is a binary segmentation pipeline followed by a pixel-count area measurement. Images are cropped to a fixed region, resized to identical pixel dimensions, converted to grayscale, normalized with contrast-limited adaptive histogram equalization (CLAHE), smoothed with a Gaussian filter, and binarized with the IsoData iterative threshold. The black pixels in the resulting binary image are converted to physical area and regressed against DFT friction separately for each pavement type; the converged thresholds for DGAC, chip seal, and OGFC are 127, 124, and 115 grayscale levels. The load-bearing idea is that after normalization, brightness differences between protruding aggregate and surrounding asphalt encode protrusion height, so a single global threshold per mixture isolates the aggregate that touches the tire.

What would settle it

Scan the same slabs with a 3D laser profilometer after each polishing cycle, compute the true contact area at the DFT measurement plane, and compare it with the paper's binary protrusion area; if the two diverge in trend or magnitude, or if changing only aggregate color while holding protrusion fixed changes the segmented area, the central mechanism fails.

Watch

Extended reading notes

Core claim

The central claim is that the aggregate protrusion area, extracted by a global thresholding algorithm after brightness normalization, is a valid proxy for the tire-contact area that generates friction. For dense-graded asphalt concrete, chip seal, and open-graded friction course, the paper reports adjusted $R^2$ values of 0.9130, 0.9655, and 0.9498 for linear models relating this area to friction measured at 40 km/h with the Dynamic Friction Tester. The paper also claims that this indicator tracks friction changes across polishing cycles better than three literature indicators: aggregate ratio, surface macrotexture index from wavelet energy, and fractal dimension of concave areas, because the proposed measure isolates the protruding fraction of aggregate that actually touches the tire.

Load-bearing premise

The method assumes that, after contrast normalization and smoothing, a pixel's brightness tells whether the aggregate beneath it protrudes above the contact plane, so the thresholded area is a real proxy for tire-contact area.

Editorial extensions

If this is right

  • A digital camera plus this pipeline can rank the skid resistance of these three mixture types in the laboratory without 3D scanning equipment.
  • Because protrusion area falls with polishing, the same indicator can be used to monitor friction loss over traffic life and flag when resurfacing is needed.
  • Mix design can compare candidate aggregates by imaging specimens before and after accelerated polishing, using the per-type models to estimate worn friction.
  • Extending the method to images taken from a moving vehicle would require re-estimating thresholds and regression coefficients, but the paper's pipeline is fast enough to make such network-level surveys feasible.

Reading between the lines

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

  • The brightness-to-protrusion assumption could be checked directly by scanning the same slabs with a 3D laser profilometer; if the binary area diverges from the true contact area at the DFT plane, the high $R^2$ values may reflect wear progression rather than contact-area mechanics.
  • Since friction and area both decline monotonically with polishing cycles, the strongest test is to apply the model to surfaces whose friction rises early in wear, which the paper itself notes for DGAC; the indicator should reproduce that non-monotonic bump if it is truly measuring protrusion.
  • A smartphone-based version with recalibrated thresholds could be evaluated against field British Pendulum or DFT readings, turning the three lab mixture models into a practical pavement-management screening tool.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

4 major / 5 minor

Summary. The paper proposes an image-based indicator, the aggregate protrusion area ('Area'), extracted from 2D pavement images via CLAHE, Gaussian filtering, and IsoData thresholding, and fits linear regression models relating Area to Dynamic Friction Tester (DFT) friction coefficients for three pavement types (DGAC, Chip Seal, OGFC). The authors report adjusted R² values above 0.90 for all three models and compare the proposed indicator against three literature indicators (AR, SMI, FD), concluding that Area more accurately reflects friction changes with polishing cycles and can be used for cost-effective pavement friction evaluation at the mix design stage.

Significance. If the central claim were fully substantiated, a single 2D image would provide a cheap, interpretable estimate of pavement friction for the tested mixture types, which would be practically valuable for mix design and maintenance screening. The paper has clear strengths: it presents a complete, reproducible image-processing pipeline; it collects data under controlled laboratory conditions with multiple lighting angles; and it honestly enumerates limitations including small sample size, per-mixture calibration needs, and the lack of 3D validation. However, the current evidence does not yet support the physical mechanism claimed, because the Area indicator is never validated against true protrusion or contact-area measurements, and the reported R² values are in-sample fits rather than predictive accuracies. The paper is thus a promising methodology case study whose central claim requires additional validation before the conclusions can be accepted.

major comments (4)
  1. [Section 4.1, Table 2, Figure 8] The adjusted R² values in Table 2 are computed on the same data used to fit the regression models, so they are in-sample measures of correlation, not predictive accuracies. Figure 8 shows 'predicted' DFT values that are by construction the fitted values from the same linear regressions, and Eq. (10)–(11) confirm that R² measures fit to the calibration data. Without a validation split, cross-validation, or a genuinely held-out set of polishing cycles or mixtures, the claim that the models 'predict' friction is not supported. I request a temporal hold-out (e.g., fitting on early polishing cycles and testing on later ones) or at least leave-one-cycle-out cross-validation.
  2. [Section 3.2] The load-bearing physical assumption—that grayscale brightness after CLAHE and Gaussian filtering encodes the height of protruding aggregates relative to a tire-contact surface—is asserted but never validated against 3D elevation data or any contact-area measurement. The paper itself notes that 2D images cannot provide depth information, yet it relies on the contrast between 'protruding' and 'non-protruding' parts without independent evidence that the binarized regions correspond to actual protrusions. Because CLAHE deliberately redistributes histograms and the thresholds are fixed per mixture (127, 124, 115), the high correlations could arise from preprocessing artifacts or from both Area and friction trending downward with polishing cycles. I recommend validating the thresholded protrusion area against a 3D scanner-derived contact area or surface elevation for at least a subset of samples, and reporting a sensitivity analysis of the threshold values.
  3. [Table 1, Section 3.2] The claim that 'the Area on the same type of pavement shows a regular linear change with friction' is contradicted by the OGFC row in Table 1: Area increases from 2518.15 mm² at 50k cycles (friction 0.38) to 2634.07 mm² at 90k cycles (friction 0.36), and then decreases to 2456.65 mm² at 150k cycles (friction 0.36). This non-monotonicity means that Area does not strictly track friction for OGFC, and the report should address this explicitly rather than stating a regular linear change. The fitted line may still show a positive overall trend, but the relationship is not as consistent as presented.
  4. [Section 4.2, Figure 10] The paper's conclusion that the proposed indicator 'outperforms other image-based indices' is not uniformly supported by its own results: for DGAC, the SMI-based model has a higher adjusted R² (0.9481) than the Area-based model (0.9130), as acknowledged in the text. The superiority claim should be qualified to state that Area performs best for OGFC and is competitive for Chip Seal, while SMI performs similarly or better for DGAC. This does not invalidate the approach but prevents an overgeneralized conclusion.
minor comments (5)
  1. [Section 3.1] The word 'Frist' at the start of Section 3.1 is a typo for 'First'.
  2. [Equations (2) and (6)] The typesetting of Equation (2) is garbled (the fraction and exponential are not properly rendered), and Equation (6) uses both I and I_gray with an unclear conditional definition; please clarify the notation and ensure all equations are legible.
  3. [Table 1] Table 1 uses '……' to hide intermediate rows, but the reader cannot verify the monotonicity claims across all polishing cycles; consider including the full dataset in supplementary material, given the small number of observations.
  4. [Section 4.1, Figure 8] The phrase 'Most of the data points are found close to the line of equality' is misleading because the figure shows a regression line, not a 1:1 equality line; please rephrase to 'close to the fitted regression line.'
  5. [Section 3.2] The paper states that thresholds for the three mixtures are 127, 124, and 115, but no information is given on the variance or stability of these thresholds across the 10+ images per condition; reporting the threshold distribution would strengthen the reproducibility claim.

Circularity Check

1 steps flagged · score 6.0 of 10

The reported friction 'prediction' accuracy is an in-sample R² of regressions fitted to the same data, so the validation reduces to the fitted values by construction.

  1. fitted input called prediction [Section 4.1, Figure 8, Table 2, Eqs. (9)–(11)]
    "As shown in Figure 8, the Area determined from image analysis is compared to DFT data to develop separate models for different pavement types. Most of the data points are found close to the line of equality, indicating that the proposed indicator is valid."

    The 'line of equality' in Figure 8 is the least-squares regression line fitted to those same DFT/Area points (Table 2, e.g., DFT = 0.2396 + 1.0632E-4 Area for DGAC). The reported R² and adjusted R² are computed from the fitted values via Eqs. (10)-(11) on the same dataset, so the data lying close to the fitted line is an in-sample goodness-of-fit property, not independent prediction. No held-out images, polishing cycles, or cross-validation are reported. The abstract/conclusion calls these 'friction prediction models' with adjusted R² above 0.90, but the reported accuracy reduces to the fitted regression values by construction. The Area extraction itself is independent of DFT; the circularity is confined to the validation step.

full rationale

The central indicator Area is derived from grayscale images via CLAHE, Gaussian filtering, and IsoData thresholding, with no friction data entering the feature extraction, so the correlation between Area and DFT is not definitionally circular. The circularity is in the validation rhetoric: the paper fits a separate linear model for each pavement type and then validates the indicator by showing that the data lie near the same fitted regression line, with R² computed from those fitted values. This is an in-sample fit being labeled as 'prediction,' matching the fitted-input-called-prediction pattern. No train/test split, cross-validation, or external validation is reported. The comparison with AR, SMI, and FD is also in-sample and therefore does not add out-of-sample support. The additional threat that Area may not track true protrusion height is a construct-validity and calibration concern rather than a circularity concern; it is not counted here. Self-citations to the authors' earlier laser/deep-learning work are motivational only and not load-bearing for the main result. Overall score 6 reflects partial circularity in the prediction claim, while the feature-extraction-to-correlation derivation itself retains independent content.

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

The predictive claim rests on a chain of domain assumptions linking 2D grayscale brightness to physical protrusion and tire contact, plus per-mixture thresholds and fitted regression coefficients. No new physical entities are introduced.

free parameters (3)
  • Per-mixture binarization thresholds = DGAC 127, Chip Seal 124, OGFC 115
    Chosen per surface type after preprocessing; the authors state thresholds must be recalibrated for other conditions, so the reported R2 depends on this calibration.
  • Linear regression coefficients = DGAC: 0.2396 + 1.0632e-4 Area; Chip Seal: 0.3151 + 1.4331e-4 Area; OGFC: -0.2504 + 2.4260e-4 Area
    Fit to the same data used to report R2, so the predicted friction is an in-sample fitted value rather than an independent prediction.
  • CLAHE clip limit and Gaussian filter sigma = not reported
    These preprocessing parameters control the binarization output and therefore affect the Area values, but their numerical values are not given.
assumptions (4)
  • domain assumption Grayscale brightness after CLAHE and Gaussian filtering corresponds to physical protrusion height of aggregates above the asphalt surface.
    Invoked in Section 3.2; no 3D elevation or contact-area data are used to verify this mapping.
  • domain assumption A single global threshold per mixture separates protruding aggregate from background across all lighting angles and polishing cycles.
    Used in Section 3.2 and Figure 7; the authors acknowledge recalibration is needed for field use.
  • domain assumption Laboratory three-wheel polishing and DFT measurements at 40 km/h capture the skid-resistance trend relevant to field traffic.
    Adopted in Section 2.1; no field validation is provided.
  • standard math Standard image processing algorithms and linear regression statistics are valid background material.
    Equations 1-11 use textbook definitions of grayscale conversion, Gaussian filtering, IsoData thresholding, and R2 statistics.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Predicting Asphalt Pavement Friction Using Texture-Based Image Indicator." pith.science (2026). https://pith.science/paper/ATJPUNRF

@misc{pith2026250703559,
  author       = {Pith},
  title        = {Pith review of: Predicting Asphalt Pavement Friction Using Texture-Based Image Indicator},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/ATJPUNRF}},
  note         = {Machine review of arXiv:2507.03559}
}
read the original abstract

Pavement skid resistance is of vital importance for road safety. The objective of this study is to propose and validate a texture-based image indicator to predict pavement friction. This index enables pavement friction to be measured easily and inexpensively using digital images. Three different types of asphalt surfaces (dense-graded asphalt mix, open-grade friction course, and chip seal) were evaluated subject to various tire polishing cycles. Images were taken with corresponding friction measured using Dynamic Friction Tester (DFT) in the laboratory. The aggregate protrusion area is proposed as the indicator. Statistical models are established for each asphalt surface type to correlate the proposed indicator with friction coefficients. The results show that the adjusted R-square values of all relationships are above 0.90. Compared to other image-based indicators in the literature, the proposed image indicator more accurately reflects the changes in pavement friction with the number of polishing cycles, proving its cost-effective use for considering pavement friction in mix design stage.

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

49 extracted references · 47 canonical work pages

  1. [1]

    Grande, E

    Z. Grande, E. Castillo, E. Mora, H.K. Lo, Highway and road probabilistic safety assessment based on Bayesian network models, Computer-Aided Civil and Infrastructure Engineering 32(5) (2017) 379- 396

    work page 2017

  2. [2]

    Čelko, M

    J. Čelko, M. Kováč, P. Kotek, Analysis of the pavement surface texture by 3D scanner, Transportation research procedia 14 (2016) 2994-3003

    work page 2016

  3. [3]

    Antle, J

    C.E. Antle, J. Wambold, J. Henry, Z. Rado, International PIARC Experiment to Compare and Harmonize Texture and Skid Resistance Measurement, PIARC Technical Commitee on Surface Characteristics C 1 (1995) 1995

    work page 1995

  4. [4]

    Rezaei, E

    A. Rezaei, E. Masad, A. Chowdhury, P. Harris, Predicting asphalt mixture skid resistance by aggregate characteristics and gradation, Transportation Research Record 2104(1) (2009) 24-33

    work page 2009

  5. [5]

    Z. Rado, T. Yager, J. Wambold, J. Hall, Guide for pavement friction, NCHRP, Project (1-43) (2006)

    work page 2006

  6. [6]

    Asi, Evaluating skid resistance of different asphalt concrete mixes, Building and environment 42(1) (2007) 325-329

    I.M. Asi, Evaluating skid resistance of different asphalt concrete mixes, Building and environment 42(1) (2007) 325-329

    work page 2007

  7. [7]

    Pérez-Acebo, H

    H. Pérez-Acebo, H. Gonzalo-Orden, D.J. Findley, E. Rojí, A skid resistance prediction model for an entire road network, Construction and building materials 262 (2020) 120041

    work page 2020

  8. [8]

    Roy, P.G

    N. Roy, P.G. Mondal, K.K. Kuna, Ima ge-based indices of aggregates for predicting the initial skid resistance of bituminous pavements, Construction and Building Materials 400 (2023) 132776

    work page 2023

Show all 49 references
  1. [9]

    Y. Du, Z. Weng, F. Li, G. Ablat, D. Wu, C. Liu, A novel approach for pavement texture characterisation using 2D-wavelet decomposition, International Journal of Pavement Engineering 23(6) (2022) 1851-1866

    2022

  2. [10]

    Yang, A.A

    G. Yang, A.A. Zhang, K.C. Wang, J.Q. Li, W. Liu, Y. Liu, Deep-learning based non-contact method for assessing pavement skid resistance using 3D laser imaging technology, International Journal of Pavement Engineering 24(2) (2023) 2147520

    2023

  3. [11]

    H. -C. Dan, B. Lu, M. Li, Evaluation of asphalt pavement texture using multiview stereo reconstruction based on deep learning, Construction and Building Materials 412 (2024) 134837

    2024

  4. [12]

    C. Feng, B. Bačić, W. Li, SCA -LSTM: A Deep Learning Approach to Golf Swing Analysis and Performance Enhancement, International Conference on Neural Information Processing, Springer, 2025, pp. 72-86

    2025

  5. [13]

    H.-C. Dan, P. Yan, J. Tan, Y. Zhou, B. Lu, Multiple distresses detection for Asphalt Pavement using improved you Only Look Once Algorithm based on convolutional neural network, International Journal of Pavement Engineering 25(1) (2024) 2308169

    2024

  6. [14]

    Y. Wang, B. Yu, X. Zhang, J. Liang, Automatic extraction and evaluation of pavement three - dimensional surface texture using laser scanning technology, Automation in Construction 141 (2022) 104410

    2022

  7. [15]

    Bazlamit, F

    S.M. Bazlamit, F. Reza, Changes in asphalt pavement friction components and adjustment of skid number for temperature, Journal of Transportation Engineering 131(6) (2005) 470-476

    2005

  8. [16]

    Z. Li, Q. Ji, X. Ling, Q. Liu, A Comprehensive Review of Multi- Agent Reinforcement Learning in Video Games, Authorea Preprints (2025)

    2025

  9. [17]

    M. Yu, Z. Yo u, G. Wu, L. Kong, C. Liu, J. Gao, Measurement and modeling of skid resistance of asphalt pavement: A review, Construction and building materials 260 (2020) 119878

    2020

  10. [18]

    Gonzalez, Digital image processing, Pearson education india2009

    R.C. Gonzalez, Digital image processing, Pearson education india2009. 16

  11. [19]

    P. Li, Q. Yang, X. Geng, W. Zhou, Z. Ding, Y. Nian, Exploring diverse methods in visual question answering, 2024 5th International Conference on Electronic Communication and Artificial Intelligence (ICECAI), IEEE, 2024, pp. 681-685

    2024

  12. [20]

    Y. Liu, X. Qin, Y. Gao, X. Li, C. Feng, SETransformer: A hybrid attention -based architecture for robust human activity recognition, arXiv preprint arXiv:2505.19369 (2025)

    2025 arXiv

  13. [21]

    Pizer, E.P

    S.M. Pizer, E.P. Amburn, J.D. Austin, R. Cromartie, A. Geselowitz, T. Greer, B. ter Haar Romeny, J.B. Zimmerman, K. Zuiderveld, Adaptive histogram equalization and its variations, Computer vision, graphics, and image processing 39(3) (1987) 355-368

    1987

  14. [22]

    C. Wu, H. Huang, L. Zhang, J. Chen, Y. Tong, M. Zhou, Towards automated 3D evaluation of water leakage on a tunnel face via improved GAN and self-attention DL model, Tunnelling and Underground Space Technology 142 (2023) 105432

    2023

  15. [23]

    Zuiderveld, Contrast limited adaptive histogram equalization, Graphics gems IV1994, pp

    K. Zuiderveld, Contrast limited adaptive histogram equalization, Graphics gems IV1994, pp. 474 - 485

  16. [24]

    S. Liu, Y. Zhang, X. Li, Y. Liu, C. Feng, H. Yang, Gated multimodal graph learning for personalized recommendation, arXiv preprint arXiv:2506.00107 (2025)

    2025 arXiv

  17. [25]

    Jähne, Digital image processing, Springer Science & Business Media2005

    B. Jähne, Digital image processing, Springer Science & Business Media2005

  18. [26]

    C. Wu, H. Huang, J. Chen, M. Zhou, S. H an, A novel Tree -augmented Bayesian network for predicting rock weathering degree using incomplete dataset, International Journal of Rock Mechanics and Mining Sciences 183 (2024) 105933

    2024

  19. [27]

    J. -S. Lee, Digital image smoothing and the sigma filter, Compute r vision, graphics, and image processing 24(2) (1983) 255-269

    1983

  20. [28]

    Gonzales, A.L

    A.L. Gonzales, A.L. Ramos, J.M. Lacson, K.S. Go, R.B. Furigay, Optimizing Image and Signal Processing Through the Application of Various Filtering Techniques: A Comparative Study, Novel & Intelligent Digital Systems Conferences, Springer, 2023, pp. 151-170

    2023

  21. [29]

    Zheng, J

    B. Zheng, J. Chen, R. Zhao, J. Tang, R. Tian, S. Zhu, X. Huang, Analysis of contact behaviour on patterned tire- asphalt pavement with 3- D FEM contact model, International Journal of Pavement Engineering 23(2) (2022) 171-186

    2022

  22. [30]

    D. Yun, L. Hu, C. Tang, Tire –road contact area on asphalt concrete pavement and its relationship with the skid resistance, Materials 13(3) (2020) 615

    2020

  23. [31]

    Y. Du, B. Qin, Z. Weng, D. Wu, C. Liu, Promoting th e pavement skid resistance estimation by extracting tire-contacted texture based on 3D surface data, Construction and Building Materials 307 (2021) 124729

    2021

  24. [32]

    Jiang, F

    J. Jiang, F. Ni, L. Gao, L. Yao, Effect of the contact structure characteristics on rutting performance in asphalt mixtures using 2D imaging analysis, Construction and Building Materials 136 (2017) 426-435

    2017

  25. [33]

    Shah, M.E

    S.M.R. Shah, M.E. Abdullah, Effect of aggregate shape on skid resistance of compacted hot mix asphalt (HMA), 2010 Second International Conf erence on Computer and Network Technology, IEEE, 2010, pp. 421-425

    2010

  26. [34]

    P. Li, M. Abouelenien, R. Mihalcea, Z. Ding, Q. Yang, Y. Zhou, Deception detection from linguistic and physiological data streams using bimodal convolutional neural networks, 2024 5th International Conference on Information Science, Parallel and Distributed Systems (ISPDS), IE...

    2024

  27. [35]

    Otsu, A threshold selection method from gray-level histograms, Automatica 11(285-296) (1975) 23-27

    N. Otsu, A threshold selection method from gray-level histograms, Automatica 11(285-296) (1975) 23-27

    1975

  28. [36]

    Ridler, S

    T.W. Ridler, S. Calvard, Picture thresholding using an iterative selection method, IEEE Trans. Syst. Man Cybern 8(8) (1978) 630-632. 17

    1978

  29. [37]

    Sezgin, B.l

    M. Sezgin, B.l. Sankur, Survey over image thresholding techniques and quantitative performance evaluation, Journal of Electronic imaging 13(1) (2004) 146-168

    2004

  30. [38]

    H. Zuo, Z. Li, J. Gong, Z. Tian, Intelligent road crack detection and analysis based on improved YOLOv8, 2025 8th International Conference on Advanced Algorithms and Control Engineering (ICAACE), IEEE, 2025, pp. 1192-1195

    2025

  31. [39]

    Hawkins, The problem of overfitting, Journal of chemical information and computer sciences 44(1) (2004) 1-12

    D.M. Hawkins, The problem of overfitting, Journal of chemical information and computer sciences 44(1) (2004) 1-12

    2004

  32. [40]

    Kaiser, J.N

    E. Kaiser, J.N. Kutz, S.L. Brunton, Sparse identification of nonlinear dynamics for model predictive control in the low-data limit, Proceedings of the Royal Society A 474(2219) (2018) 20180335

    2018

  33. [41]

    Valikhani, A

    A. Valikhani, A. Jaberi Jahromi, S. Pouyanfar, I.M. Mantawy, A. Azizinamini, Machine learning and image processing approaches for estimating concrete surface roughness using basic cameras, Computer-Aided Civil and Infrastructure Engineering 36(2) (2021) 213-226

    2021

  34. [42]

    T. Wan, H. Wang, P. Feng, A. Diab, Concave distribution characterization of asphalt pavement surface segregation using smartphone and image processing based techniques, Construction and Building Materials 301 (2021) 124111

    2021

  35. [43]

    Vieira, U

    T. Vieira, U. Sandberg, S. Erlingsson, Negative texture, positive for the environment: effects of horizontal grinding of asphalt pavements, Road Materials and Pavement Design 22(1) (2021) 1-22

    2021

  36. [44]

    Prowell, J

    B.D. Prowell, J. Zhang, E.R. Brown, Aggregate properties and the performance of superpave- designed hot mix asphalt, Transportation Research Board2005

  37. [45]

    Praticò, R

    F. Praticò, R. Vaiana, A study on the relationship between mean texture depth and mean profile depth of asphalt pavements, Construction and Building Materials 101 (2015) 72-79

    2015

  38. [46]

    Z. Kang, Y. Zhang, X. Deng, X. Li, Y. Zhang, LP-DETR: Layer-wise Progressive Relations for Object Detection, arXiv preprint arXiv:2502.05147 (2025)

    2025 arXiv

  39. [47]

    Zhong, Y

    J. Zhong, Y. Wa ng, Enhancing Thyroid Disease Prediction Using Machine Learning: A Comparative Study of Ensemble Models and Class Balancing Techniques, (2025)

    2025

  40. [48]

    Zhong, Y

    J. Zhong, Y. Wang, D. Zhu, Z. Wang, A Narrative Review on Large AI Models in Lung Cancer Screening, Diagnosis, and Treatment Planning, arXiv preprint arXiv:2506.07236 (2025)

    2025 arXiv

  41. [49]

    Zhang, M

    Y. Zhang, M. Zhu, K. Gui, J. Yu, Y. Hao, H. Sun, Development and application of a monte carlo tree search algorithm for simulating da vinci code game strategies, 2024 5th Internatio nal Conference on Intelligent Computing and Human-Computer Interaction (ICHCI), IEEE, 2024, pp. 5-11

    2024

Pith tools

Reviewed August 6, 2026 · model on record in the stance chip above.