{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,5,11]],"date-time":"2026-05-11T16:12:16Z","timestamp":1778515936098,"version":"3.51.4"},"reference-count":55,"publisher":"MDPI AG","issue":"10","license":[{"start":{"date-parts":[[2026,5,9]],"date-time":"2026-05-09T00:00:00Z","timestamp":1778284800000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"funder":[{"name":"Postgraduate Research and Practice Innovation Program of Jiangsu Province","award":["SJCX25_2013"],"award-info":[{"award-number":["SJCX25_2013"]}]}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":["Sensors"],"abstract":"<jats:p>Millimeter-wave radar imaging plays an increasingly important role in autonomous driving road perception due to its robustness under adverse weather conditions. However, radar images are inherently contaminated by multiplicative speckle noise, which severely degrades structural continuity, weakens target boundaries, and limits the perception of road scenes and surrounding objects. To address this problem, this paper proposes a structural-prior deep learning network for millimeter-wave radar image enhancement. The proposed framework first introduces an adaptive Otsu-based masking strategy to extract salient scattering structures and generate a coarse image structural prior for subsequent restoration. Guided by this prior, the network performs progressive feature enhancement through a continuous attention mechanism that integrates residual channel attention, context-aware feature extraction, and convolutional block attention, thereby enabling effective multi-scale representation learning while suppressing signal-dependent speckle interference. In addition, a composite loss function is designed by combining logarithmic denoising gain, total variation regularization, and a \u03b2-index edge-preservation term to jointly improve noise suppression, spatial smoothness, and structural fidelity. The proposed method is evaluated on the synthetic UC Merced dataset under different noise intensities and via cross-domain inference on the real-world RADIATE millimeter-wave radar dataset for autonomous driving scenarios. Experimental results demonstrate that the proposed network consistently outperforms conventional filtering methods and representative deep learning baselines in terms of PSNR, SSIM, \u03b2-index, and ENL while providing a superior preservation of road structures, target contours, and scene geometry. Ablation studies further confirm the effectiveness of the structural-prior guidance and continuous attention design. Furthermore, the network achieves a rapid inference latency of 12.35 milliseconds. These results indicate that the proposed method provides an effective and robust solution for millimeter-wave radar image enhancement and offers practical value for downstream road-scene perception in autonomous driving environments.<\/jats:p>","DOI":"10.3390\/s26102976","type":"journal-article","created":{"date-parts":[[2026,5,11]],"date-time":"2026-05-11T15:39:32Z","timestamp":1778513972000},"page":"2976","update-policy":"https:\/\/doi.org\/10.3390\/mdpi_crossmark_policy","source":"Crossref","is-referenced-by-count":0,"title":["Structural-Prior Deep Learning Network for Millimeter-Wave Radar Image Enhancement in Autonomous Driving Road Sensing"],"prefix":"10.3390","volume":"26","author":[{"ORCID":"https:\/\/orcid.org\/0009-0003-1117-3281","authenticated-orcid":false,"given":"Hongyan","family":"Chen","sequence":"first","affiliation":[{"name":"School of Transportation and Civil Engineering, Nantong University, Nantong 226019, China"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Tonghui","family":"Huang","sequence":"additional","affiliation":[{"name":"School of Transportation and Civil Engineering, Nantong University, Nantong 226019, China"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-1183-103X","authenticated-orcid":false,"given":"Yuexia","family":"Wang","sequence":"additional","affiliation":[{"name":"School of Transportation and Civil Engineering, Nantong University, Nantong 226019, China"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-3609-0258","authenticated-orcid":false,"given":"Jiajia","family":"Shi","sequence":"additional","affiliation":[{"name":"School of Transportation and Civil Engineering, Nantong University, Nantong 226019, China"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-5645-3610","authenticated-orcid":false,"given":"Zhihuo","family":"Xu","sequence":"additional","affiliation":[{"name":"School of Transportation and Civil Engineering, Nantong University, Nantong 226019, China"}],"role":[{"role":"author","vocabulary":"crossref"}]}],"member":"1968","published-online":{"date-parts":[[2026,5,9]]},"reference":[{"key":"ref_1","doi-asserted-by":"crossref","unstructured":"Qiu, Z., Shao, J., Guo, D., Yin, X., Zhai, Z., Duan, Z., and Xu, Y. (2025). A Multi-Feature Fusion Approach for Road Surface Recognition Leveraging Millimeter-Wave Radar. Sensors, 25.","DOI":"10.3390\/s25123802"},{"key":"ref_2","doi-asserted-by":"crossref","unstructured":"Yao, L., Chung, K.L., Tang, L., Ye, T., Wang, S., Xu, P., Bi, Y., and Wu, Y. (2026). A Hybrid Millimeter-Wave Radar\u2013Ultrasonic Fusion System for Robust Human Activity Recognition with Attention-Enhanced Deep Learning. Sensors, 26.","DOI":"10.3390\/s26031057"},{"key":"ref_3","doi-asserted-by":"crossref","unstructured":"Steinmetz, N., Magori, E., Balal, Y., Sudai, Y.B., and Balal, N. (2026). A 94 GHz Millimeter-Wave Radar System for Remote Vehicle Height Measurement to Prevent Bridge Collisions. Sensors, 26.","DOI":"10.3390\/s26061921"},{"key":"ref_4","doi-asserted-by":"crossref","unstructured":"Peter, E., Ang, L.M., Seng, K.P., and Srivastava, S. (2026). Recent Advances in Deep Learning for SAR Images: Overview of Methods, Challenges, and Future Directions. Sensors, 26.","DOI":"10.3390\/s26041143"},{"key":"ref_5","doi-asserted-by":"crossref","first-page":"4606","DOI":"10.1109\/TIV.2024.3380244","article-title":"4D mmWave Radar for Autonomous Driving Perception: A Comprehensive Survey","volume":"9","author":"Fan","year":"2024","journal-title":"IEEE Trans. Intell. Veh."},{"key":"ref_6","doi-asserted-by":"crossref","first-page":"3946","DOI":"10.1109\/TFUZZ.2025.3598211","article-title":"Exploring Fuzzy Priors From Multimapping GAN for Robust Image Dehazing","volume":"33","author":"Zhang","year":"2025","journal-title":"IEEE Trans. Fuzzy Syst."},{"key":"ref_7","doi-asserted-by":"crossref","first-page":"1448","DOI":"10.1109\/TIP.2026.3657636","article-title":"IHDCP: Single Image Dehazing Using Inverted Haze Density Correction Prior","volume":"35","author":"Liu","year":"2026","journal-title":"IEEE Trans. Image Process."},{"key":"ref_8","doi-asserted-by":"crossref","first-page":"20408","DOI":"10.1109\/JIOT.2022.3175375","article-title":"Fusing mmWave Radar With Camera for 3-D Detection in Autonomous Driving","volume":"9","author":"Liu","year":"2022","journal-title":"IEEE Internet Things J."},{"key":"ref_9","doi-asserted-by":"crossref","unstructured":"St\u00e4cker, L., Heidenreich, P., Rambach, J., and Stricker, D. (2022, January 3\u20138). Fusion Point Pruning for Optimized 2D Object Detection with Radar-Camera Fusion. Proceedings of the 2022 IEEE\/CVF Winter Conference on Applications of Computer Vision (WACV), Waikoloa, HI, USA.","DOI":"10.1109\/WACV51458.2022.00134"},{"key":"ref_10","doi-asserted-by":"crossref","first-page":"7","DOI":"10.1186\/s13634-022-00944-6","article-title":"CenterTransFuser: Radar point cloud and visual information fusion for 3D object detection","volume":"2023","author":"Li","year":"2023","journal-title":"EURASIP J. Adv. Signal Process."},{"key":"ref_11","doi-asserted-by":"crossref","first-page":"18185","DOI":"10.1109\/TITS.2024.3417826","article-title":"Traffic Object Detection for Autonomous Driving Fusing LiDAR and Pseudo 4D-Radar Under Bird\u2019s-Eye-View","volume":"25","author":"Meng","year":"2024","journal-title":"IEEE Trans. Intell. Transp. Syst."},{"key":"ref_12","doi-asserted-by":"crossref","first-page":"7316","DOI":"10.1109\/TPAMI.2023.3301553","article-title":"CoIR: Compressive Implicit Radar","volume":"47","author":"Farrell","year":"2025","journal-title":"IEEE Trans. Pattern Anal. Mach. Intell."},{"key":"ref_13","doi-asserted-by":"crossref","unstructured":"Borts, D., Liang, E., Broedermann, T., Ramazzina, A., Walz, S., Palladin, E., Sun, J., Brueggemann, D., Sakaridis, C., and Van Gool, L. (August, January 27). Radar Fields: Frequency-Space Neural Scene Representations for FMCW Radar. Proceedings of the SIGGRAPH \u201924: ACM SIGGRAPH 2024 Conference Papers, Denver, CO, USA.","DOI":"10.1145\/3641519.3657510"},{"key":"ref_14","doi-asserted-by":"crossref","first-page":"1611","DOI":"10.1049\/iet-rsn.2015.0571","article-title":"Insights into prior learning for despeckling SAR images","volume":"10","author":"Xu","year":"2016","journal-title":"IET Radar Sonar Navig."},{"key":"ref_15","doi-asserted-by":"crossref","first-page":"5204017","DOI":"10.1109\/TGRS.2021.3065461","article-title":"Speckle2Void: Deep Self-Supervised SAR Despeckling With Blind-Spot Convolutional Neural Networks","volume":"60","author":"Molini","year":"2022","journal-title":"IEEE Trans. Geosci. Remote Sens."},{"key":"ref_16","doi-asserted-by":"crossref","first-page":"4321","DOI":"10.1109\/JSTARS.2021.3071864","article-title":"SAR2SAR: A Semi-Supervised Despeckling Algorithm for SAR Images","volume":"14","author":"Dalsasso","year":"2021","journal-title":"IEEE J. Sel. Top. Appl. Earth Obs. Remote Sens."},{"key":"ref_17","doi-asserted-by":"crossref","first-page":"5201914","DOI":"10.1109\/TGRS.2023.3237466","article-title":"Multitemporal Speckle Reduction With Self-Supervised Deep Neural Networks","volume":"61","author":"Meraoumia","year":"2023","journal-title":"IEEE Trans. Geosci. Remote Sens."},{"key":"ref_18","doi-asserted-by":"crossref","first-page":"2963","DOI":"10.1109\/JSTARS.2023.3345538","article-title":"Robust SAR Image Despeckling by Deep Learning From Near-Real Datasets","volume":"17","author":"Guan","year":"2024","journal-title":"IEEE J. Sel. Top. Appl. Earth Obs. Remote Sens."},{"key":"ref_19","doi-asserted-by":"crossref","first-page":"165","DOI":"10.1109\/TPAMI.1980.4766994","article-title":"Digital image enhancement and noise filtering by use of local statistics","volume":"PAMI-2","author":"Lee","year":"1980","journal-title":"IEEE Trans. Pattern Anal. Mach. Intell."},{"key":"ref_20","doi-asserted-by":"crossref","first-page":"1260","DOI":"10.1109\/TIP.2002.804276","article-title":"Speckle reducing anisotropic diffusion","volume":"11","author":"Yu","year":"2002","journal-title":"IEEE Trans. Image Process."},{"key":"ref_21","doi-asserted-by":"crossref","first-page":"3142","DOI":"10.1109\/TIP.2017.2662206","article-title":"Beyond a gaussian denoiser: Residual learning of deep cnn for image denoising","volume":"26","author":"Zhang","year":"2017","journal-title":"IEEE Trans. Image Process."},{"key":"ref_22","doi-asserted-by":"crossref","unstructured":"Zhang, Q., Yuan, Q., Li, J., Yang, Z., and Ma, X. (2018). Learning a dilated residual network for SAR image despeckling. Remote Sens., 10.","DOI":"10.3390\/rs10020196"},{"key":"ref_23","doi-asserted-by":"crossref","unstructured":"Zhao, X., Ren, F., Sun, H., and Qi, Q. (2024). Synthetic aperture radar image despeckling based on a deep learning network employing frequency domain decomposition. Electronics, 13.","DOI":"10.3390\/electronics13030490"},{"key":"ref_24","doi-asserted-by":"crossref","first-page":"5207717","DOI":"10.1109\/TGRS.2023.3271998","article-title":"Residual in Residual Scaling Networks for Polarimetric SAR Image Despeckling","volume":"61","author":"Lin","year":"2023","journal-title":"IEEE Trans. Geosci. Remote Sens."},{"key":"ref_25","doi-asserted-by":"crossref","first-page":"6372","DOI":"10.1109\/JSTARS.2023.3292325","article-title":"A Deep Neural Network Based on Prior-Driven and Structural Preserving for SAR Image Despeckling","volume":"16","author":"Lin","year":"2023","journal-title":"IEEE J. Sel. Top. Appl. Earth Obs. Remote Sens."},{"key":"ref_26","doi-asserted-by":"crossref","first-page":"10420","DOI":"10.1109\/JSTARS.2023.3327332","article-title":"MFAENet: A Multiscale Feature Adaptive Enhancement Network for SAR Image Despeckling","volume":"16","author":"Liu","year":"2023","journal-title":"IEEE J. Sel. Top. Appl. Earth Obs. Remote Sens."},{"key":"ref_27","doi-asserted-by":"crossref","first-page":"5216812","DOI":"10.1109\/TGRS.2023.3314857","article-title":"SAR Despeckling Using Multiobjective Neural Network Trained With Generic Statistical Samples","volume":"61","author":"Vitale","year":"2023","journal-title":"IEEE Trans. Geosci. Remote Sens."},{"key":"ref_28","doi-asserted-by":"crossref","unstructured":"Pongrac, B., and Gleich, D. (2023). Despeckling of SAR images using residual twin CNN and multi-resolution attention mechanism. Remote Sens., 15.","DOI":"10.3390\/rs15143698"},{"key":"ref_29","doi-asserted-by":"crossref","unstructured":"Bu, L., Zhang, J., Zhang, Z., Yang, Y., and Deng, M. (2023). Deep learning for integrated speckle reduction and super-resolution in multi-temporal SAR. Remote Sens., 16.","DOI":"10.3390\/rs16010018"},{"key":"ref_30","doi-asserted-by":"crossref","first-page":"4036","DOI":"10.1109\/JSTARS.2024.3355220","article-title":"ANED-Net: Adaptive Noise Estimation and Despeckling Network for SAR Image","volume":"17","author":"Wang","year":"2024","journal-title":"IEEE J. Sel. Top. Appl. Earth Obs. Remote Sens."},{"key":"ref_31","doi-asserted-by":"crossref","first-page":"19380","DOI":"10.1109\/JSTARS.2024.3483786","article-title":"HTCNet: Hybrid Transformer-CNN for SAR Image Denoising","volume":"17","author":"Huang","year":"2024","journal-title":"IEEE J. Sel. Top. Appl. Earth Obs. Remote Sens."},{"key":"ref_32","doi-asserted-by":"crossref","first-page":"15525","DOI":"10.1109\/JSTARS.2024.3450429","article-title":"SAR Image Despeckle With CNN Using a Novel Logarithmic Discrete Cosine Transform-Based Loss","volume":"17","author":"Saleem","year":"2024","journal-title":"IEEE J. Sel. Top. Appl. Earth Obs. Remote Sens."},{"key":"ref_33","doi-asserted-by":"crossref","first-page":"28272","DOI":"10.1109\/ACCESS.2021.3059003","article-title":"Feature-Guided CNN for Denoising Images From Portable Ultrasound Devices","volume":"9","author":"Dong","year":"2021","journal-title":"IEEE Access"},{"key":"ref_34","doi-asserted-by":"crossref","first-page":"5215417","DOI":"10.1109\/TGRS.2024.3419083","article-title":"Despeckling SAR Images With Log-Yeo\u2013Johnson Transformation and Conditional Diffusion Models","volume":"62","author":"Ma","year":"2024","journal-title":"IEEE Trans. Geosci. Remote Sens."},{"key":"ref_35","doi-asserted-by":"crossref","unstructured":"Pan, Y., Zhong, L., Chen, J., Li, H., Zhang, X., and Pan, B. (2024). SAR Image Despeckling Based on Denoising Diffusion Probabilistic Model and Swin Transformer. Remote Sens., 16.","DOI":"10.3390\/rs16173222"},{"key":"ref_36","doi-asserted-by":"crossref","unstructured":"Lin, H., Su, X., Zeng, Z., Xing, C., and Yin, J. (2025). Speckle2Self: Learning Self-Supervised Despeckling with Attention Mechanism for SAR Images. Remote Sens., 17.","DOI":"10.3390\/rs17233840"},{"key":"ref_37","doi-asserted-by":"crossref","first-page":"3","DOI":"10.1109\/JSTARS.2021.3132027","article-title":"SAR image despeckling using continuous attention module","volume":"15","author":"Ko","year":"2021","journal-title":"IEEE J. Sel. Top. Appl. Earth Obs. Remote Sens."},{"key":"ref_38","doi-asserted-by":"crossref","unstructured":"Liang, J., Cao, J., Sun, G., Zhang, K., Van Gool, L., and Timofte, R. (2021, January 11\u201317). Swinir: Image restoration using swin transformer. Proceedings of the IEEE\/CVF International Conference on Computer Vision, Montreal, BC, Canada.","DOI":"10.1109\/ICCVW54120.2021.00210"},{"key":"ref_39","doi-asserted-by":"crossref","first-page":"5223311","DOI":"10.1109\/TGRS.2025.3628887","article-title":"Despeckling of Synthetic Aperture Radar Images using Linear\u2013Angular Attention Transformer","volume":"63","author":"Kundu","year":"2025","journal-title":"IEEE Trans. Geosci. Remote Sens."},{"key":"ref_40","doi-asserted-by":"crossref","first-page":"5211912","DOI":"10.1109\/TGRS.2024.3397325","article-title":"Trans-NLM network for SAR image despeckling","volume":"62","author":"Xiao","year":"2024","journal-title":"IEEE Trans. Geosci. Remote Sens."},{"key":"ref_41","doi-asserted-by":"crossref","unstructured":"Zhang, L., Zheng, L., Wen, Y., Zhang, F., Bo, F., and Cen, Y. (2025). Effective SAR image despeckling using noise-guided transformer and multi-scale feature fusion. Remote Sens., 17.","DOI":"10.3390\/rs17233863"},{"key":"ref_42","doi-asserted-by":"crossref","first-page":"376","DOI":"10.1016\/j.isprsjprs.2024.11.003","article-title":"Contrastive learning for real SAR image despeckling","volume":"218","author":"Fang","year":"2024","journal-title":"ISPRS J. Photogramm. Remote Sens."},{"key":"ref_43","doi-asserted-by":"crossref","unstructured":"Perera, M.V., Bandara, W.G.C., Valanarasu, J.M.J., and Patel, V.M. (2022). Transformer-based SAR image despeckling. Proceedings of the IGARSS 2022-2022 IEEE International Geoscience and Remote Sensing Symposium, IEEE.","DOI":"10.1109\/IGARSS46834.2022.9884596"},{"key":"ref_44","doi-asserted-by":"crossref","first-page":"4504005","DOI":"10.1109\/LGRS.2022.3145996","article-title":"Wavelength-resolution SAR speckle model","volume":"19","author":"Xu","year":"2022","journal-title":"IEEE Geosci. Remote Sens. Lett."},{"key":"ref_45","doi-asserted-by":"crossref","first-page":"6","DOI":"10.1109\/MGRS.2013.2277512","article-title":"A tutorial on speckle reduction in synthetic aperture radar images","volume":"1","author":"Argenti","year":"2013","journal-title":"IEEE Geosci. Remote Sens. Mag."},{"key":"ref_46","doi-asserted-by":"crossref","first-page":"2232","DOI":"10.1109\/TGRS.2003.815240","article-title":"Polarimetric SAR speckle noise model","volume":"41","author":"Fabregas","year":"2003","journal-title":"IEEE Trans. Geosci. Remote Sens."},{"key":"ref_47","doi-asserted-by":"crossref","first-page":"2724","DOI":"10.1109\/TGRS.2011.2107915","article-title":"SAR image despeckling based on local homogeneous-region segmentation by using pixel-relativity measurement","volume":"49","author":"Feng","year":"2011","journal-title":"IEEE Trans. Geosci. Remote Sens."},{"key":"ref_48","doi-asserted-by":"crossref","first-page":"8029705","DOI":"10.1109\/LGRS.2022.3211399","article-title":"Background noise suppression for DAS-VSP records using GC-AB-Unet","volume":"19","author":"Zhao","year":"2022","journal-title":"IEEE Geosci. Remote Sens. Lett."},{"key":"ref_49","doi-asserted-by":"crossref","unstructured":"Zhang, Y., Li, K., Li, K., Wang, L., Zhong, B., and Fu, Y. (2018, January 8\u201314). Image super-resolution using very deep residual channel attention networks. Proceedings of the European Conference on Computer Vision (ECCV), Munich, Germany.","DOI":"10.1007\/978-3-030-01234-2_18"},{"key":"ref_50","doi-asserted-by":"crossref","unstructured":"Woo, S., Park, J., Lee, J.Y., and Kweon, I.S. (2018, January 8\u201314). Cbam: Convolutional block attention module. Proceedings of the European Conference on Computer Vision (ECCV), Munich, Germany.","DOI":"10.1007\/978-3-030-01234-2_1"},{"key":"ref_51","doi-asserted-by":"crossref","unstructured":"Yang, Y., and Newsam, S. (2010, January 2\u20135). Bag-of-visual-words and spatial extensions for land-use classification. Proceedings of the 18th SIGSPATIAL International Conference on Advances in Geographic Information Systems, San Jose, CA, USA.","DOI":"10.1145\/1869790.1869829"},{"key":"ref_52","doi-asserted-by":"crossref","first-page":"4010905","DOI":"10.1109\/LGRS.2023.3315839","article-title":"Efficient Maximum-Likelihood Estimation of Equivalent Number of Looks for PolSAR Image","volume":"20","author":"Qin","year":"2023","journal-title":"IEEE Geosci. Remote Sens. Lett."},{"key":"ref_53","doi-asserted-by":"crossref","unstructured":"Sheeny, M., De Pellegrin, E., Mukherjee, S., Ahrabian, A., Wang, S., and Wallace, A. (2021). Radiate: A radar dataset for automotive perception in bad weather. Proceedings of the 2021 IEEE International Conference on Robotics and Automation (ICRA), IEEE.","DOI":"10.1109\/ICRA48506.2021.9562089"},{"key":"ref_54","doi-asserted-by":"crossref","first-page":"2661","DOI":"10.1109\/TIP.2009.2029593","article-title":"Iterative Weighted Maximum Likelihood Denoising With Probabilistic Patch-Based Weights","volume":"18","author":"Deledalle","year":"2009","journal-title":"IEEE Trans. Image Process."},{"key":"ref_55","doi-asserted-by":"crossref","first-page":"606","DOI":"10.1109\/TGRS.2011.2161586","article-title":"A Nonlocal SAR Image Denoising Algorithm Based on LLMMSE Wavelet Shrinkage","volume":"50","author":"Parrilli","year":"2012","journal-title":"IEEE Trans. Geosci. Remote Sens."}],"container-title":["Sensors"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/www.mdpi.com\/1424-8220\/26\/10\/2976\/pdf","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,5,11]],"date-time":"2026-05-11T15:45:18Z","timestamp":1778514318000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.mdpi.com\/1424-8220\/26\/10\/2976"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,5,9]]},"references-count":55,"journal-issue":{"issue":"10","published-online":{"date-parts":[[2026,5]]}},"alternative-id":["s26102976"],"URL":"https:\/\/doi.org\/10.3390\/s26102976","relation":{},"ISSN":["1424-8220"],"issn-type":[{"value":"1424-8220","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,5,9]]}}}