{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,9,12]],"date-time":"2026-09-12T05:07:13Z","timestamp":1789189633040,"version":"build-2803163510"},"reference-count":30,"publisher":"MDPI AG","issue":"18","license":[{"start":{"date-parts":[[2026,9,10]],"date-time":"2026-09-10T00:00:00Z","timestamp":1788998400000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"funder":[{"DOI":"10.13039\/501100011705","name":"Korea Institute of Marine Science & Technology Promotion","doi-asserted-by":"crossref","id":[{"id":"10.13039\/501100011705","id-type":"DOI","asserted-by":"crossref"}]},{"id":[{"id":"https:\/\/ror.org\/04cqyfy93","id-type":"ROR","asserted-by":"publisher"}]},{"DOI":"10.13039\/501100011705","name":"Korea Coast Guard","doi-asserted-by":"publisher","award":["RS-2025-02313423"],"award-info":[{"award-number":["RS-2025-02313423"]}],"id":[{"id":"10.13039\/501100011705","id-type":"DOI","asserted-by":"publisher"}]}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":["Sensors"],"abstract":"<jats:p>Rapid maritime incident response requires visual recognition near the point of observation, but practical deployment must balance detection accuracy, inference latency, hardware-conversion constraints, and limited communication capacity. This study benchmarks the visual-compute component of an Event-Centric Edge Intelligence (ECEI) architecture on two maritime edge tiers: a Jetson AGX Orin 64 GB Developer Kit and a reComputer AI R2000-12 based on Raspberry Pi 5. Three tasks were selected to represent distinct maritime event categories: Human Detection for person-overboard response, six-class Ship Detection for vessel monitoring and collision awareness, and eleven-class Marine Trash Detection for floating-debris and pollution monitoring. YOLO26, YOLO11, and YOLOv8 were evaluated across 54 model\u2013dataset\u2013device configurations. Repeated on-device validation used 100 deterministically selected images per task and 20 measured repetitions for each checkpoint, yielding 108,000 timed predictions. On the paired S\/N subset, the median reComputer\/AGX mean-latency ratio was 4.40\u00d7; mean latency ranged from 19.84\u201357.09 ms on AGX and 68.89\u2013172.11 ms on reComputer. The highest AGX mAP50\u201395 values were 0.331, 0.813, and 0.639 for Human, Ship, and Marine Trash Detection, respectively; the highest reComputer values were 0.318, 0.806, and 0.558. Across the evaluated event-message configuration, event-candidate JSON bytes were 98.20% lower than raw JPEG bytes. These results support deployment-aware visual-tier selection and an auditable detector-to-event interface for maritime edge intelligence.<\/jats:p>","DOI":"10.3390\/s26185764","type":"journal-article","created":{"date-parts":[[2026,9,10]],"date-time":"2026-09-10T15:54:19Z","timestamp":1789055659000},"page":"5764","update-policy":"https:\/\/doi.org\/10.3390\/mdpi_crossmark_policy","source":"Crossref","is-referenced-by-count":0,"title":["Benchmarking Visual Detection Tiers for Event-Centric Edge Intelligence in Maritime Disaster Response"],"prefix":"10.3390","volume":"26","author":[{"given":"Sobirjon","family":"Habibullaev","sequence":"first","affiliation":[{"name":"GreenBlue Corporate Research Institute, GreenBlue Inc., 40 Imi-ro, Uiwang-si 16006, Gyeonggi-do, Republic of Korea"}],"role":[{"vocabulary":"crossref","role":"author"},{"role":"software","vocabulary":"credit"},{"role":"investigation","vocabulary":"credit"},{"role":"methodology","vocabulary":"credit"},{"role":"writing-original-draft","vocabulary":"credit"},{"role":"validation","vocabulary":"credit"}]},{"ORCID":"https:\/\/orcid.org\/0009-0005-3877-3740","authenticated-orcid":false,"given":"Juno","family":"Choi","sequence":"additional","affiliation":[{"name":"GreenBlue Corporate Research Institute, GreenBlue Inc., 40 Imi-ro, Uiwang-si 16006, Gyeonggi-do, Republic of Korea"}],"role":[{"vocabulary":"crossref","role":"author"},{"role":"conceptualization","vocabulary":"credit"},{"role":"writing-review-editing","vocabulary":"credit"},{"role":"validation","vocabulary":"credit"},{"role":"project-administration","vocabulary":"credit"},{"role":"resources","vocabulary":"credit"}]}],"member":"1968","published-online":{"date-parts":[[2026,9,10]]},"reference":[{"key":"ref_1","unstructured":"Korea Coast Guard (2026, September 03). Maritime Distress Accident Detailed Data Status. Public Data Portal, Available online: https:\/\/www.data.go.kr\/data\/3043630\/fileData.do."},{"key":"ref_2","unstructured":"Korea Coast Guard (2026, September 03). Marine Pollution Accident Detailed Data Status. Public Data Portal, Available online: https:\/\/www.data.go.kr\/data\/15113375\/fileData.do."},{"key":"ref_3","doi-asserted-by":"crossref","first-page":"637","DOI":"10.1109\/JIOT.2016.2579198","article-title":"Edge Computing: Vision and Challenges","volume":"3","author":"Shi","year":"2016","journal-title":"IEEE Internet Things J."},{"key":"ref_4","doi-asserted-by":"crossref","first-page":"30","DOI":"10.1109\/MC.2017.9","article-title":"The Emergence of Edge Computing","volume":"50","author":"Satyanarayanan","year":"2017","journal-title":"Computer"},{"key":"ref_5","doi-asserted-by":"crossref","first-page":"348","DOI":"10.1016\/j.sigpro.2014.07.029","article-title":"Measurement Level AIS\/Radar Fusion","volume":"106","author":"Habtemariam","year":"2015","journal-title":"Signal Process."},{"key":"ref_6","doi-asserted-by":"crossref","first-page":"114198","DOI":"10.1016\/j.oceaneng.2023.114198","article-title":"Improving Maritime Traffic Surveillance in Inland Waterways Using the Robust Fusion of AIS and Visual Data","volume":"275","author":"Qu","year":"2023","journal-title":"Ocean Eng."},{"key":"ref_7","doi-asserted-by":"crossref","first-page":"111630","DOI":"10.1016\/j.measurement.2022.111630","article-title":"A New Multi-Sensor Fusion Approach for Integrated Ship Motion Perception in Inland Waterways","volume":"200","author":"Wu","year":"2022","journal-title":"Measurement"},{"key":"ref_8","doi-asserted-by":"crossref","first-page":"122643","DOI":"10.1016\/j.oceaneng.2025.122643","article-title":"Real-Time Vessel Geolocation and Automatic Identification via Video\u2013AIS Data Fusion","volume":"341","author":"Gan","year":"2025","journal-title":"Ocean Eng."},{"key":"ref_9","doi-asserted-by":"crossref","first-page":"111118","DOI":"10.1016\/j.engappai.2025.111118","article-title":"Multi-Source Perception Data Fusion of Vessels in Visual Occlusion Scenarios: Leveraging Prior Knowledge of Vessel Motion","volume":"156","author":"He","year":"2025","journal-title":"Eng. Appl. Artif. Intell."},{"key":"ref_10","doi-asserted-by":"crossref","first-page":"15","DOI":"10.1016\/j.patrec.2025.02.022","article-title":"Multimodal AI-Enhanced Ship Detection for Mapping Fishing Vessels and Informing on Suspicious Activities","volume":"191","author":"Galdelli","year":"2025","journal-title":"Pattern Recognit. Lett."},{"key":"ref_11","doi-asserted-by":"crossref","first-page":"184","DOI":"10.1145\/3555802","article-title":"Edge Computing with Artificial Intelligence: A Machine Learning Perspective","volume":"55","author":"Hua","year":"2023","journal-title":"ACM Comput. Surv."},{"key":"ref_12","doi-asserted-by":"crossref","first-page":"10908","DOI":"10.1038\/s41598-026-46453-6","article-title":"Review of Large YOLOv8 and RT-DETR Energy Efficiency on Edge Devices for Real-Time Detection","volume":"16","year":"2026","journal-title":"Sci. Rep."},{"key":"ref_13","doi-asserted-by":"crossref","unstructured":"Aljami, H.M., Alrowais, N.A., AlAwajy, A.M., Alhrgan, S.O., Aldwaani, R.A., Alsawadi, M.S., Saqib, N.U., Alam, S.S., and Alsubaie, R. (2026). Benchmarking YOLOv8 Variants for Object Detection Efficiency on Jetson Orin NX for Edge Computing Applications. Computers, 15.","DOI":"10.3390\/computers15020074"},{"key":"ref_14","doi-asserted-by":"crossref","first-page":"5875","DOI":"10.1038\/s41598-026-36862-y","article-title":"Object Detection on Low-Compute Edge SoCs: A Reproducible Benchmark and Deployment Guidelines","volume":"16","author":"Kong","year":"2026","journal-title":"Sci. Rep."},{"key":"ref_15","doi-asserted-by":"crossref","unstructured":"Moosbauer, S., K\u00f6nig, D., J\u00e4kel, J., and Teutsch, M. (2019, January 16\u201320). A Benchmark for Deep Learning Based Object Detection in Maritime Environments. Proceedings of the IEEE\/CVF CVPR Workshops, Long Beach, CA, USA.","DOI":"10.1109\/CVPRW.2019.00121"},{"key":"ref_16","doi-asserted-by":"crossref","unstructured":"Kiefer, B., \u017dust, L., Muhovi\u010d, J., Kristan, M., Per\u0161, J., Ter\u0161ek, M., Mudenagudi, U., Desai, C., Wiliem, A., and Kreis, M. (March, January 28). 3rd Workshop on Maritime Computer Vision (MaCVi) 2025: Challenge Results. Proceedings of the WACV Workshops, Tucson, AZ, USA.","DOI":"10.1109\/WACVW65960.2025.00170"},{"key":"ref_17","unstructured":"Kiefer, B., Augustin, J.L., Muhovi\u010d, J., Jeong, M., Wiliem, A., Pers, J., Kristan, M., Li, A.Q., Ter\u0161ek, M., and \u0160ari\u0107, J. (2026, January 2\u20136). 4th Workshop on Maritime Computer Vision (MaCVi): Challenge Overview. Proceedings of the IEEE\/CVF CVPR Workshops, Denver, CO, USA."},{"key":"ref_18","doi-asserted-by":"crossref","first-page":"110798","DOI":"10.1016\/j.engappai.2025.110798","article-title":"A Real-Time Lightweight Object Detection Algorithm Based on Improved You Only Look Once Version 8 for Unmanned Surface Vehicle","volume":"152","author":"Gong","year":"2025","journal-title":"Eng. Appl. Artif. Intell."},{"key":"ref_19","doi-asserted-by":"crossref","unstructured":"Kong, L. (2025, January 23\u201325). Securing the Seas, Strengthening the State: Leveraging Multi-Scale Joint Attention Networks for Intelligent Ship Detection. Proceedings of the 2025 IEEE 12th Joint International Information Technology and Artificial Intelligence Conference (ITAIC), Chongqing, China.","DOI":"10.1109\/ITAIC64559.2025.11163094"},{"key":"ref_20","unstructured":"Ultralytics (2026, July 29). YOLO26 Documentation. Available online: https:\/\/docs.ultralytics.com\/models\/yolo26\/."},{"key":"ref_21","unstructured":"Jocher, G., and Qiu, J. (2026, July 29). Ultralytics YOLO11. Available online: https:\/\/docs.ultralytics.com\/models\/yolo11\/."},{"key":"ref_22","unstructured":"Jocher, G., Chaurasia, A., and Qiu, J. (2026, July 29). Ultralytics YOLOv8. Available online: https:\/\/docs.ultralytics.com\/models\/yolov8\/."},{"key":"ref_23","doi-asserted-by":"crossref","unstructured":"Knocker, M., Sontheimer, M., Goncalves Machado, G., and Chou, S.-Y. (2025, January 7\u201311). Democratizing UAV-Based Search and Rescue: Real-Time Human Detection on Affordable Edge AI Hardware. Proceedings of the 32nd International Conference on Transdisciplinary Engineering, Monterrey, Mexico.","DOI":"10.3233\/ATDE251079"},{"key":"ref_24","unstructured":"(2026, July 29). NVIDIA Jetson AGX Orin Technical Specifications. Available online: https:\/\/www.nvidia.com\/en-eu\/autonomous-machines\/embedded-systems\/jetson-orin\/."},{"key":"ref_25","unstructured":"Seeed Studio (2026, July 29). reComputer AI R2000 Series: Raspberry Pi 5 Edge AI Computer with Hailo-8. Available online: https:\/\/www.seeedstudio.com.cn\/product\/recomputer-ai-r2000-series-raspberry-pi-high-performance-edge-ai-computer-based-on-raspberry-pi-5-hailo-8-ai-acceleration-module."},{"key":"ref_26","unstructured":"Roboflow Universe (2026, August 04). Human Detection in Water Dataset, Version 1; License: CC BY 4.0. Available online: https:\/\/universe.roboflow.com\/otd-erfan-n9dzd\/human-detecion-in-water\/dataset\/1."},{"key":"ref_27","unstructured":"Roboflow Universe (2026, August 04). seaShip Dataset, Version 1. License: CC BY 4.0. Available online: https:\/\/universe.roboflow.com\/datasets-ol6bv\/seaship-m74ww\/dataset\/1."},{"key":"ref_28","unstructured":"AI Hub (2026, August 04). Marine Trash Detection Dataset (Korean title: \u201c\ud574\uc591\uce68\uc801\uc4f0\ub808\uae30 \uc774\ubbf8\uc9c0 \ub370\uc774\ud130 \uace0\ub3c4\ud654\u201d), version 1.1. Available online: https:\/\/www.aihub.or.kr\/aihubdata\/data\/view.do?aihubDataSe=data&dataSetSn=71340."},{"key":"ref_29","doi-asserted-by":"crossref","unstructured":"Lin, T.-Y., Maire, M., Belongie, S., Hays, J., Perona, P., Ramanan, D., Doll\u00e1r, P., and Zitnick, C.L. (2014, January 6\u201312). Microsoft COCO: Common Objects in Context. Proceedings of the ECCV 2014, Zurich, Switzerland.","DOI":"10.1007\/978-3-319-10602-1_48"},{"key":"ref_30","unstructured":"Ultralytics (2026, July 29). Object Detection Performance Metrics. Available online: https:\/\/docs.ultralytics.com\/guides\/yolo-performance-metrics\/."}],"container-title":["Sensors"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/www.mdpi.com\/1424-8220\/26\/18\/5764\/pdf","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,9,12]],"date-time":"2026-09-12T04:27:29Z","timestamp":1789187249000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.mdpi.com\/1424-8220\/26\/18\/5764"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,9,10]]},"references-count":30,"journal-issue":{"issue":"18","published-online":{"date-parts":[[2026,9]]}},"alternative-id":["s26185764"],"URL":"https:\/\/doi.org\/10.3390\/s26185764","relation":{},"ISSN":["1424-8220"],"issn-type":[{"value":"1424-8220","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,9,10]]}}}