{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,8,3]],"date-time":"2026-08-03T14:01:45Z","timestamp":1785765705638,"version":"3.56.0"},"reference-count":32,"publisher":"Optica Publishing Group","issue":"23","license":[{"start":{"date-parts":[[2026,8,3]],"date-time":"2026-08-03T00:00:00Z","timestamp":1785715200000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/doi.org\/10.1364\/OA_License_v2#VOR"},{"start":{"date-parts":[[2026,8,3]],"date-time":"2026-08-03T00:00:00Z","timestamp":1785715200000},"content-version":"tdm","delay-in-days":0,"URL":"https:\/\/opg.optica.org\/policies\/opg-tdm-policy.json"}],"funder":[{"name":"Independent Innovation Science Fund of National University of Defense Technology","award":["24-ZZCX-JDZ-29"],"award-info":[{"award-number":["24-ZZCX-JDZ-29"]}]},{"DOI":"10.13039\/501100001809","name":"National Natural Science Foundation of China","doi-asserted-by":"crossref","award":["52305594"],"award-info":[{"award-number":["52305594"]}],"id":[{"id":"10.13039\/501100001809","id-type":"DOI","asserted-by":"crossref"},{"id":"https:\/\/ror.org\/01h0zpd94","id-type":"ROR","asserted-by":"publisher"}]}],"content-domain":{"domain":["opg.optica.org"],"crossmark-restriction":false},"short-container-title":["Appl. Opt."],"published-print":{"date-parts":[[2026,8,10]]},"abstract":"<jats:p>Achieving reliable object perception in underwater environments characterized by turbidity, low-light, and spectral distortion remains a critical challenge for vision-based applications. To address this, we propose a novel, to our knowledge, dynamic adaptive object detection model. Our core innovation lies in a collaborative optimization framework that integrates a dynamic feature fusion module (utilizing DySample for upsampling) and a dynamic convolution module (DyC2f) to adaptively enhance multi-scale feature representation. Furthermore, we design a tailored loss function (FWNWD) that combines a WIoU dynamic focusing mechanism, a Focaler-IoU weighting strategy, and the normalized Wasserstein distance (NWD) to significantly improve localization robustness, particularly for blurry and small objects. Extensive evaluations on the large-scale DUO dataset (containing 74,515 instances) demonstrate the effectiveness of our approach. The model achieves a state-of-the-art mAP@0.5 of 91.7% with only 6.6G FLOPs, and runs at 23 FPS on an embedded platform (Orange Pi Aipro), successfully balancing high accuracy with real-time efficiency. Ablation studies confirm the contribution of each component, notably showing a 2.3% increase in recall for small objects attributable to the FWNWD loss. This work provides a practical and efficient solution for high-precision underwater perception, and its dynamically tunable architecture offers valuable insights for vision systems in other complex, degraded environments.<\/jats:p>","DOI":"10.1364\/ao.597480","type":"journal-article","created":{"date-parts":[[2026,7,9]],"date-time":"2026-07-09T16:00:10Z","timestamp":1783612810000},"page":"7758","update-policy":"https:\/\/doi.org\/10.1364\/crossmark_policy","source":"Crossref","is-referenced-by-count":0,"title":["Toward high-accuracy underwater object detection in optically challenging environments via dynamic adaptive optimization"],"prefix":"10.1364","volume":"65","author":[{"given":"Shucheng","family":"Li","sequence":"first","affiliation":[{"id":[{"id":"https:\/\/ror.org\/05d2yfz11","id-type":"ROR","asserted-by":"publisher"}]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Xinjie","family":"Zhao","sequence":"additional","affiliation":[{"id":[{"id":"https:\/\/ror.org\/05d2yfz11","id-type":"ROR","asserted-by":"publisher"}]},{"name":"National Key Laboratory of Equipment State Sensing and Smart Support"},{"name":"Hunan Provincial Key Laboratory of Ultra-Precision Machining Technology"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Zihao","family":"Zheng","sequence":"additional","affiliation":[{"id":[{"id":"https:\/\/ror.org\/05d2yfz11","id-type":"ROR","asserted-by":"publisher"}]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Xing","family":"Peng","sequence":"additional","affiliation":[{"id":[{"id":"https:\/\/ror.org\/05d2yfz11","id-type":"ROR","asserted-by":"publisher"}]},{"name":"National Key Laboratory of Equipment State Sensing and Smart Support"},{"name":"Hunan Provincial Key Laboratory of Ultra-Precision Machining Technology"}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"285","published-online":{"date-parts":[[2026,8,3]]},"reference":[{"key":"ao-65-23-7758-R1","doi-asserted-by":"publisher","first-page":"4962","DOI":"10.3390\/s20174962","type":"journal-article","volume":"20","author":"Liu","year":"2020","journal-title":"Sensors"},{"key":"ao-65-23-7758-R2","first-page":"886","article-title":"Histograms of oriented gradients for human detection","volume-title":"Proceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition","author":"Dalal","year":"2005"},{"key":"ao-65-23-7758-R3","first-page":"I","article-title":"Rapid object detection using a boosted cascade of simple features","volume-title":"Proceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition","author":"Viola","year":"2001"},{"key":"ao-65-23-7758-R4","first-page":"580","article-title":"Rich feature hierarchies for accurate object detection and semantic segmentation","volume-title":"Proceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition","author":"Girshick","year":"2014"},{"key":"ao-65-23-7758-R5","doi-asserted-by":"publisher","first-page":"35479","DOI":"10.1109\/ACCESS.2023.3266093","type":"journal-article","volume":"11","author":"Amjoud","year":"2023","journal-title":"IEEE Access"},{"key":"ao-65-23-7758-R6","first-page":"1440","article-title":"Fast R-CNN","volume-title":"Proceedings of the IEEE International Conference on Computer Vision","author":"Girshick","year":"2015"},{"key":"ao-65-23-7758-R7","doi-asserted-by":"publisher","first-page":"1137","DOI":"10.1109\/TPAMI.2016.2577031","type":"journal-article","volume":"39","author":"Ren","year":"2017","journal-title":"IEEE Trans. Pattern Anal. Mach. Intell."},{"key":"ao-65-23-7758-R8","first-page":"21","article-title":"SSD: single shot multibox detector","volume-title":"European Conference on Computer Vision","author":"Liu","year":"2016"},{"key":"ao-65-23-7758-R9","first-page":"779","article-title":"You only look once: unified, real-time object detection","volume-title":"Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition","author":"Redmon","year":"2016"},{"key":"ao-65-23-7758-R10","doi-asserted-by":"publisher","first-page":"103514","DOI":"10.1016\/j.dsp.2022.103514","type":"journal-article","volume":"126","author":"Zaidi","year":"2022","journal-title":"Digit. Signal Process."},{"key":"ao-65-23-7758-R11","doi-asserted-by":"publisher","first-page":"34831","DOI":"10.1109\/ACCESS.2025.3544061","type":"journal-article","volume":"13","author":"Luo","year":"2025","journal-title":"IEEE Access"},{"key":"ao-65-23-7758-R12","doi-asserted-by":"publisher","first-page":"26","DOI":"10.1007\/s00530-024-01617-0","type":"journal-article","volume":"31","author":"Zheng","year":"2025","journal-title":"Multimed. Syst."},{"key":"ao-65-23-7758-R13","doi-asserted-by":"publisher","first-page":"14057","DOI":"10.1007\/s11227-024-06020-0","type":"journal-article","volume":"80","author":"Chen","year":"2024","journal-title":"J. Supercomput."},{"key":"ao-65-23-7758-R14","doi-asserted-by":"publisher","first-page":"065108","DOI":"10.1063\/5.0206734","type":"journal-article","volume":"95","author":"Zhang","year":"2024","journal-title":"Rev. Sci. Instrum."},{"key":"ao-65-23-7758-R15","doi-asserted-by":"publisher","first-page":"1127","DOI":"10.3390\/jmse12071127","type":"journal-article","volume":"12","author":"Lu","year":"2024","journal-title":"J. Mar. Sci. Eng."},{"key":"ao-65-23-7758-R16","doi-asserted-by":"publisher","first-page":"423","DOI":"10.1016\/j.aej.2024.07.126","type":"journal-article","volume":"104","author":"Liu","year":"2024","journal-title":"Alex. Eng. J."},{"key":"ao-65-23-7758-R17","doi-asserted-by":"publisher","first-page":"1348883","DOI":"10.3389\/fmars.2024.1348883","type":"journal-article","volume":"11","author":"Cheng","year":"2024","journal-title":"Front. Mar. Sci."},{"key":"ao-65-23-7758-R18","doi-asserted-by":"publisher","first-page":"229","DOI":"10.1108\/RIA-07-2023-0090","type":"journal-article","volume":"44","author":"Luo","year":"2024","journal-title":"Robotic Intell. Autom."},{"key":"ao-65-23-7758-R19","doi-asserted-by":"publisher","first-page":"2326","DOI":"10.3390\/jmse12122326","type":"journal-article","volume":"12","author":"Zheng","year":"2024","journal-title":"J. Mar. Sci. Eng."},{"key":"ao-65-23-7758-R20","doi-asserted-by":"publisher","first-page":"294","DOI":"10.3390\/fishes9080294","type":"journal-article","volume":"9","author":"Feng","year":"2024","journal-title":"Fishes"},{"key":"ao-65-23-7758-R21","doi-asserted-by":"publisher","first-page":"135","DOI":"10.3390\/jmse13010135","type":"journal-article","volume":"13","author":"Hu","year":"2025","journal-title":"J. Mar. Sci. Eng."},{"key":"ao-65-23-7758-R22","doi-asserted-by":"publisher","author":"Khanam","year":"2024","DOI":"10.48550\/arXiv.2410.17725","type":"preprint"},{"key":"ao-65-23-7758-R23","doi-asserted-by":"publisher","first-page":"105057","DOI":"10.1016\/j.imavis.2024.105057","type":"journal-article","volume":"147","author":"Kang","year":"2024","journal-title":"Image Vis. Comput."},{"key":"ao-65-23-7758-R24","first-page":"6027","article-title":"Learning to upsample by learning to sample","volume-title":"Proceedings of the IEEE\/CVF International Conference on Computer Vision","author":"Liu","year":"2023"},{"key":"ao-65-23-7758-R25","first-page":"15751","article-title":"ParameterNet: parameters are all you need for large-scale visual pretraining of mobile networks","volume-title":"Proceedings of the IEEE\/CVF Conference on Computer Vision and Pattern Recognition","author":"Han","year":"2024"},{"key":"ao-65-23-7758-R26","first-page":"85","article-title":"IOU loss for 2D\/3D object detection","volume-title":"International Conference on 3D Vision","author":"Zhou","year":"2019"},{"key":"ao-65-23-7758-R27","first-page":"12993","article-title":"Distance-IoU loss: faster and better learning for bounding box regression","volume-title":"Proceedings of the AAAI Conference on Artificial Intelligence","author":"Zheng","year":"2020"},{"key":"ao-65-23-7758-R28","doi-asserted-by":"publisher","author":"Tong","year":"2023","DOI":"10.48550\/arXiv.2301.10051","type":"preprint"},{"key":"ao-65-23-7758-R29","doi-asserted-by":"publisher","author":"Zhang","year":"2024","DOI":"10.48550\/arXiv.2401.10525","type":"preprint"},{"key":"ao-65-23-7758-R30","doi-asserted-by":"publisher","author":"Wang","year":"2021","DOI":"10.48550\/arXiv.2110.13389","type":"preprint"},{"key":"ao-65-23-7758-R31","first-page":"1","article-title":"A dataset and benchmark of underwater object detection for robot picking","volume-title":"IEEE International Conference on Multimedia & Expo Workshops","author":"Liu","year":"2021"},{"key":"ao-65-23-7758-R33","doi-asserted-by":"publisher","first-page":"2831","DOI":"10.1109\/TCSVT.2021.3100059","type":"journal-article","volume":"32","author":"Liu","year":"2022","journal-title":"IEEE Trans. Circuits Syst. Video Technol."}],"container-title":["Applied Optics"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/opg.optica.org\/viewmedia.cfm?URI=ao-65-23-7758&seq=0","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,8,3]],"date-time":"2026-08-03T13:33:34Z","timestamp":1785764014000},"score":1,"resource":{"primary":{"URL":"https:\/\/opg.optica.org\/abstract.cfm?URI=ao-65-23-7758"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,8,3]]},"references-count":32,"journal-issue":{"issue":"23","published-online":{"date-parts":[[2026]]},"published-print":{"date-parts":[[2026]]}},"URL":"https:\/\/doi.org\/10.1364\/ao.597480","relation":{},"ISSN":["1559-128X","2155-3165"],"issn-type":[{"value":"1559-128X","type":"print"},{"value":"2155-3165","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,8,3]]},"assertion":[{"value":"Optica Publishing Group","name":"publisher","label":"This article is maintained by"},{"value":"https:\/\/doi.org\/10.1364\/AO.597480","name":"articlelink","label":"Crossref DOI link to publisher maintained version"},{"value":"research-article","name":"content_type","label":"Article type"},{"value":"Screened by Similarity Check","name":"cross_check","label":"Similarity check"},{"value":"Yes","order":0,"name":"peer_reviewed","label":"Peer reviewed","group":{"name":"peer_review","label":"Peer review"}},{"value":"Single blind","order":1,"name":"review_process","label":"Review process","group":{"name":"peer_review","label":"Peer review"}},{"value":"9 March 2026","order":0,"name":"received","label":"Received","group":{"name":"publication_history","label":"Publication History"}},{"value":"16 June 2026","order":1,"name":"accepted","label":"Accepted","group":{"name":"publication_history","label":"Publication History"}},{"value":"3 August 2026","order":2,"name":"published","label":"Published","group":{"name":"publication_history","label":"Publication History"}},{"value":"\u00a9 2026 Optica Publishing Group. All rights, including for text and data mining (TDM), Artificial Intelligence (AI) training, and similar technologies, are reserved.","name":"copyright","label":"Copyright"}]}}