近年來，氣候變遷加劇與水體優養化導致藍綠菌異常增生，其所形成的懸浮藻毯對水處理系統與飲用水安全造成嚴重威脅。傳統的人工採樣與新型衛星遙測方法存在即時監測能力及局部空間解析度有限等問題，為建立更具時效性的預警機制，本研究建立一套基於人工智慧的自動化懸浮藻毯監測與預警系統 。
研究蒐集來自天然渠道與淨水廠的現地影像以建立專屬資料集，並採用YOLOv11 (You Only Look Once v11) 與RF-DETR (Roboflow Detection Transformer) 兩種先進深度學習物件偵測模型進行訓練與效能評估。為突破單一架構模型的侷限，本研究導入加權框融合 (Weighted Boxes Fusion, WBF) 演算法以整合雙模型之優勢，並結合DeepSORT多目標追蹤演算法實現懸浮藻毯的即時動態追蹤。此外，系統透過計算影像中的藻毯覆蓋率推估藻毯潛在的2-MIB濃度，並建立當濃度超過10 ng/L時的即時預警機制。
研究結果顯示，基於Transformer架構的RF-DETR在測試集上的整體表現（F1-score為70%，mAP@0.5為70%）優於基於CNN架構的YOLOv11。經WBF雙模型融合後，模型的Recall成功提升至62%，F1-score達70%，有效降低模型遺漏偵測目標物的機率 。在實務應用成效上，WBF融合模型結合DeepSORT於天然渠道中能穩定追蹤藻毯的漂移軌跡，其基於面積計算之Recall高達92.8% ；而於淨水廠中，其基於面積之Recall更達到100%。本研究所建立之系統能有效提供即時的藻毯覆蓋率及2-MIB濃度推估數據，為水處理單位提供具高度實務參考價值的早期預警與管理決策依據。

This study develops an AI-driven automated monitoring and early warning system to mitigate the severe threats posed by floating algae mats to water quality and treatment facilities. Traditional monitoring methods, such as manual sampling and satellite remote sensing, suffer from delayed responses and insufficient spatial resolution. To address these limitations, this research established a customized dataset from natural channels and a water treatment plant. Two advanced object detection models, the CNN-based YOLOv11 and the Transformer-based RF-DETR, were trained. To overcome the constraints of single architecture, the Weighted Boxes Fusion (WBF) algorithm was implemented to integrate the predictive strengths of both models. Furthermore, the DeepSORT algorithm was incorporated to achieve real-time dynamic tracking of the algae mats.
The system calculates the real-time spatial coverage of the algae to estimate the concentration of the odorous compound 2-MIB. It triggers an automatic early warning when the estimated concentration exceeds the 10 ng/L threshold. Results demonstrated that the WBF-fused model successfully enhanced detection performance, achieving a Recall of 62% and an F1-score of 70%. Field applications confirmed that the integrated system stably tracked algae trajectories, achieving an area-based recall of 92.8% in natural channels and 100% in treatment plants. Ultimately, this system provides water management authorities with a highly practical, real-time decision-support tool for proactive water quality management.

Abdullah, Ali, S., Khan, Z., Hussain, A., Athar, A., Kim, H.-C., 2022. Computer Vision Based Deep Learning Approach for the Detection and Classification of Algae Species Using Microscopic Images, Water, p. 2219.
Al-Hejri, A.M., Al-Tam, R.M., Sable, A.H., Almuhaya, B., Alshamrani, S.S., Alshmrany, K.M., 2025. A hybrid vision transformer with ensemble CNN framework for cervical cancer diagnosis. BMC Med Inform Decis Mak 25, 411.
Asada, Y., Hayasaka, S., Miyoshi, T., Tokuyasu, M., Akiba, M., 2023. Effects of raw water quality on the adsorptive removal of 2-methylisoborneol by powdered activated carbon under non-equilibrium conditions. AQUA - Water Infrastructure, Ecosystems and Society 72, 1084-1095.
Bewley, A., Zongyuan, G., Lionel, O., Fabio, R., Ben, U., 2016. Simple online and realtime tracking.
Bodla, N., Singh, B., Chellappa, R., Davis, L.S., 2017. Soft-NMS — Improving Object Detection with One Line of Code. 2017 IEEE International Conference on Computer Vision (ICCV), 5562-5570.
Bresciani, M., Giardino, C., Rosaria, L., Matta, E., Cazzaniga, I., Pinardi, M., Lami, A., Austoni, M., Viaggiu, E., Congestri, R., Morabito, G., 2016. Earth observation for monitoring and mapping of cyanobacteria blooms. Case studies on five Italian lakes. Journal of Limnology 76.
Brust, C.-A., Käding, C., Denzler, J., 2019. Active Learning for Deep Object Detection.
Carmichael, W.W., Boyer, G.L., 2016. Health impacts from cyanobacteria harmful algae blooms: Implications for the North American Great Lakes. Harmful Algae 54, 194-212.
Chen, L., Zhu, J., 2024. Water surface garbage detection based on lightweight YOLOv5. Scientific Reports 14.
Chorus, I., 1999. Toxic Cyanobacteria in Water - A guide to their public health consequences, monitoring and management.
Chorus, I., Welker, M., 2021. Toxic Cyanobacteria in Water: A Guide to Their Public Health Consequences, Monitoring and Management (2nd ed.).
Chourib, I., 2025. From Detection to Diagnosis: An Advanced Transfer Learning Pipeline Using YOLO11 with Morphological Post-Processing for Brain Tumor Analysis for MRI Images. Journal of Imaging 11, 282.
Chu, Y.E., 2026. The Contribution and Simulation of Two Earthy and Musty Compounds of Cyanobacterial Mat in Channels, Department of Environmental Engineering. National Cheng Kung University, National Digital Library of Theses and Dissertations in Taiwan.
Clark, J.M., Schaeffer, B.A., Darling, J.A., Urquhart, E.A., Johnston, J.M., Ignatius, A.R., Myer, M.H., Loftin, K.A., Werdell, P.J., Stumpf, R.P., 2017. Satellite monitoring of cyanobacterial harmful algal bloom frequency in recreational waters and drinking water sources. Ecological Indicators 80, 84-95.
Cullen, J.J., Ciotti, Á.M., Davis, R.F., Lewis, M.R., 1997. Optical detection and assessment of algal blooms. Limnology and Oceanography 42, 1223-1239.
Dahiya, N., Prakash, D., Kundu, S., Kuttan, S.R., Suwalka, I., Ayadi, M., Dubale, M., Hashmi, A., 2025. Optimised RFO tuned RF-DETR model for precision urine microscopy for renal and systemic disease diagnosis. Scientific Reports 15, 25842.
Dantas Borges, H., Cicerelli, R., Almeida, T., Roig, H., Olivetti, D., 2019. Monitoring cyanobacteria occurrence in freshwater reservoirs using semi-analytical algorithms and orbital remote sensing. Marine and Freshwater Research 71, 569-578.
Devi, A., Chiu, Y.-T., Hsueh, H.-T., Lin, T.-F., 2021. Quantitative PCR based detection system for cyanobacterial geosmin/2-methylisoborneol (2-MIB) events in drinking water sources: Current status and challenges. Water Research 188, 116478.
Di, J., Xi, K., Yang, Y., 2025. An enhanced YOLOv8 model for accurate detection of solid floating waste. Sci Rep 15, 25015.
Diptho, R.A., Basak, S., 2025. Enhancing Dermatological Diagnosis Through Medical Image Analysis: How Effective Is YOLO11 Compared to Leading CNN Models? NDT 3, 11.
Dodds, W.K., Bouska, W.W., Eitzmann, J.L., Pilger, T.J., Pitts, K.L., Riley, A.J., Schloesser, J.T., Thornbrugh, D.J., 2008. Eutrophication of U.S. Freshwaters: Analysis of Potential Economic Damages. Environmental Science & Technology 43, 12-19.
Du, Y., Zhao, Z., Song, Y., Zhao, Y., Su, F., Gong, T., Meng, H., 2023. StrongSORT: Make DeepSORT Great Again. IEEE Transactions on Multimedia 25, 8725-8737.
Everingham, M., Van Gool, L., Williams, C.K.I., Winn, J., Zisserman, A., 2010. The Pascal Visual Object Classes (VOC) Challenge. International Journal of Computer Vision 88, 303-338.
Ganaie, M.A., Hu, M., Malik, A.K., Tanveer, M., Suganthan, P.N., 2022. Ensemble deep learning: A review. Engineering Applications of Artificial Intelligence 115, 105151.
Gholamy, A., Kreinovich, V., Kosheleva, O., 2018. Why 70/30 or 80/20 Relation Between Training and Testing Sets: A Pedagogical Explanation.
Gibson, J.J., 1957. Optical motions and transformations as stimuli for visual perception. Psychol Rev 64, 288-295.
Gong, X., Ma, C., Sun, B., Zhang, J., 2023. An Efficient Self-Organized Detection System for Algae, Sensors, p. 1609.
Goodfellow, I., Bengio, Y., Courville, A., Bengio, Y., 2016. Deep learning. MIT press Cambridge.
He, L.-h., Zhou, Y.-z., Liu, L., Cao, W., Ma, J.-h., 2025. Research on object detection and recognition in remote sensing images based on YOLOv11. Scientific Reports 15, 14032.
He, L., Zhou, Y., Liu, L., Ma, J., 2024. Research and Application of YOLOv11-Based Object Segmentation in Intelligent Recognition at Construction Sites. Buildings 14, 3777.
He, X., Liu, Y.L., Conklin, A., Westrick, J., Weavers, L.K., Dionysiou, D.D., Lenhart, J.J., Mouser, P.J., Szlag, D., Walker, H.W., 2016. Toxic cyanobacteria and drinking water: Impacts, detection, and treatment. Harmful Algae 54, 174-193.
Häder, D.-P., Sinha, R.P., Mishra, S., Kumari, N., 2022. Cyanobacterial Blooms and Their Implications in the Changing Environment. Advances in Environmental and Engineering Research 03, 011.
Hosang, J., Benenson, R., Schiele, B., 2017. Learning non-maximum suppression.
Ingrid Chorus, I.R.F.H.J.S.J.B., 2000. HEALTH RISKS CAUSED BY FRESHWATER CYANOBACTERIA IN RECREATIONAL WATERS. Journal of Toxicology and Environmental Health, Part B 3, 323-347.
Jenkins, W.M., Masterton, R.B., 1982. Sound localization: effects of unilateral lesions in central auditory system. J Neurophysiol 47, 987-1016.
Jocher, G., Munawar, R., Derrenger, P., 2024. Ultralytics YOLO11.
Kalman, R.E., 1960. A New Approach to Linear Filtering and Prediction Problems. Journal of Basic Engineering 82, 35-45.
Khanam, R., Hussain, M., 2024. YOLOv11: An Overview of the Key Architectural Enhancements.
Kong, Y., Shang, X., Jia, S., 2024. Drone-DETR: Efficient Small Object Detection for Remote Sensing Image Using Enhanced RT-DETR Model, Sensors, p. 5496.
Kuhn, H.W., 1955. The Hungarian method for the assignment problem. Naval Research Logistics Quarterly 2, 83-97.
Kumar, M.S., Babu, N.R., 2026. Underwater waste detection via quadruple neutrosophic image enhancement for marine environmental monitoring. Marine Pollution Bulletin 227, 119478.
Kuo, L.C., Tai, C.C., 2022. Robust Image-Based Water-Level Estimation Using Single-Camera Monitoring. IEEE Transactions on Instrumentation and Measurement 71, 1-11.
Kuznetsova, A., Rom, H., Alldrin, N., Uijlings, J., Krasin, I., Pont-Tuset, J., Kamali, S., Popov, S., Malloci, M., Kolesnikov, A., Duerig, T., Ferrari, V., 2020. The Open Images Dataset V4: Unified Image Classification, Object Detection, and Visual Relationship Detection at Scale. International Journal of Computer Vision 128, 1956-1981.
Li, J., Zheng, S., Chen, Y., Luo, K., 2026. Adaptive Multi-Scale Fusion Enhanced RT-DETR for Efficient Cyanobacteria Detection in Microscopic Images. Biology 15, 970.
Liang, Y., Han, Y., Jiang, F., 2022. Deep Learning-based Small Object Detection: A Survey, Proceedings of the 8th International Conference on Computing and Artificial Intelligence. Association for Computing Machinery, Tianjin, China, pp. 432–438.
Lin, T.-F., Watson, S., Dietrich, A., Suffet, M., 2019. Taste and Odour in Source and Drinking Water: Causes, Controls, and Consequences.
Lin, T.-Y., Maire, M., Belongie, S., Hays, J., Perona, P., Ramanan, D., Dollár, P., Zitnick, C.L., 2014. Microsoft COCO: Common Objects in Context, in: Fleet, D., Pajdla, T., Schiele, B., Tuytelaars, T. (Eds.), Computer Vision – ECCV 2014. Springer International Publishing, Cham, pp. 740-755.
Liu, C., Ersan, M.S., Plewa, M.J., Amy, G., Karanfil, T., 2018. Formation of regulated and unregulated disinfection byproducts during chlorination of algal organic matter extracted from freshwater and marine algae. Water Research 142, 313-324.
Liu, D., Wang, P., Cheng, Y., Bi, H., 2022. An improved algae-YOLO model based on deep learning for object detection of ocean microalgae considering aquacultural lightweight deployment. Frontiers in Marine Science Volume 9 - 2022.
Liu, Y., He, M., Hui, B., 2025. ESO-DETR: An Improved Real-Time Detection Transformer Model for Enhanced Small Object Detection in UAV Imagery, Drones, p. 143.
Loshchilov, I., Hutter, F., 2017. Decoupled Weight Decay Regularization, International Conference on Learning Representations.
Luo, W., Xing, J., Milan, A., Zhang, X., Liu, W., Kim, T.-K., 2021. Multiple object tracking: A literature review. Artificial Intelligence 293, 103448.
Lv, Z., Dong, S., Xia, Z., He, J., Zhang, J., 2025. Enhanced real-time detection transformer (RT-DETR) for robotic inspection of underwater bridge pier cracks. Automation in Construction 170, 105921.
Matthews, M.W., 2014. Eutrophication and cyanobacterial blooms in South African inland waters: 10years of MERIS observations. Remote Sensing of Environment 155, 161-177.
Miller, M.A., Kudela, R.M., Mekebri, A., Crane, D., Oates, S.C., Tinker, M.T., Staedler, M., Miller, W.A., Toy-Choutka, S., Dominik, C., Hardin, D., Langlois, G., Murray, M., Ward, K., Jessup, D.A., 2010. Evidence for a novel marine harmful algal bloom: cyanotoxin (microcystin) transfer from land to sea otters. PLoS One 5.
Moreira, C., Vasconcelos, V., Antunes, A., 2022. Cyanobacterial Blooms: Current Knowledge and New Perspectives. Earth 3, 127-135.
O’Neil, J.M., Davis, T.W., Burford, M.A., Gobler, C.J., 2012. The rise of harmful cyanobacteria blooms: The potential roles of eutrophication and climate change. Harmful Algae 14, 313-334.
Padilla, R., Netto, S., da Silva, E., 2020a. A Survey on Performance Metrics for Object-Detection Algorithms.
Padilla, R., Netto, S.L., Silva, E.A.B.d., 2020b. A Survey on Performance Metrics for Object-Detection Algorithms, 2020 International Conference on Systems, Signals and Image Processing (IWSSIP), pp. 237-242.
Paerl, H., Huisman, J., 2008. Blooms Like It Hot. Science (New York, N.Y.) 320, 57-58.
Park, J., Baek, J., Kim, J., You, K., Kim, K., 2022. Deep Learning-Based Algal Detection Model Development Considering Field Application. Water 14, 1275.
Park, J., Patel, K., Lee, W.H., 2024. Recent advances in algal bloom detection and prediction technology using machine learning. Science of The Total Environment 938, 173546.
Powers, D., 2008. Evaluation: From Precision, Recall and F-Factor to ROC, Informedness, Markedness & Correlation. Mach. Learn. Technol. 2.
Priyadharshini, S., Bhoopalan, R., Manikandan, D., Ramaswamy, K., 2025. A successive framework for brain tumor interpretation using Yolo variants. Scientific Reports 15, 27973.
Pyo, J., Duan, H., Baek, S., Kim, M.S., Jeon, T., Kwon, Y.S., Lee, H., Cho, K.H., 2019. A convolutional neural network regression for quantifying cyanobacteria using hyperspectral imagery. Remote Sensing of Environment 233, 111350.
Rahima, K., Muhammad, H., 2024. YOLOv11: An Overview of the Key Architectural Enhancements.
Redmon, J., Divvala, S., Girshick, R., Farhadi, A., 2016a. You Only Look Once: Unified, Real-Time Object Detection.
Redmon, J., Divvala, S., Girshick, R., Farhadi, A., 2016b. You Only Look Once: Unified, Real-Time Object Detection, 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), pp. 779-788.
Ripley, B.D., 1996. Pattern Recognition and Neural Networks. Cambridge University Press, Cambridge.
Robinson, I., Robicheaux, P., Popov, M., 2025. RF-DETR: Real-Time SOTA Object Detection.
Robinson, I., Robicheaux, P., Popov, M., Ramanan, D., Peri, N., 2026. RF-DETR: Neural Architecture Search for Real-Time Detection Transformers.
Roboflow, 2024. Roboflow: Computer Vision Data Management and Annotation Platform (Version 1.0).
Russakovsky, O., Deng, J., Su, H., Krause, J., Satheesh, S., Ma, S., Huang, Z., Karpathy, A., Khosla, A., Bernstein, M., Berg, A.C., Fei-Fei, L., 2015. ImageNet Large Scale Visual Recognition Challenge. International Journal of Computer Vision 115, 211-252.
Samantaray, A., Yang, B., Dietz, J.E., Min, B.-C., 2018. Algae Detection Using Computer Vision and Deep Learning. ArXiv abs/1811.10847.
Sazak, H., Kotan, M., 2025. Automated Blood Cell Detection and Classification in Microscopic Images Using YOLOv11 and Optimized Weights. Diagnostics 15, 22.
Scholin, C., Doucette, G., Jensen, S., Roman, B., Pargett, D., Iii, R., Preston, C., Jones, W., Feldman, J., Everlove, C., Harris, A., Alvarado, N., Massion, E., Birch, J., Greenfield, D., Wheeler, K., Vrijenhoek, R., Mikulski, C., Jones, K., 2009. Remote Detection of Marine Microbes, Small Invertebrates, Harmful Algae, and Biotoxins using the Environmental Sample Processor (ESP). Oceanography 22, 158-167.
Shao, H., Wang, S., Cui, W., 2026. CER-DETR: Channel-aware and edge-aligned detection of floating debris under complex weather. Marine Pollution Bulletin 224, 119098.
Shi, K., Zhang, Y., Qin, B., Zhou, B., 2019. Remote sensing of cyanobacterial blooms in inland waters: present knowledge and future challenges. Science Bulletin 64, 1540-1556.
Shorten, C., Khoshgoftaar, T.M., 2019. A survey on Image Data Augmentation for Deep Learning. Journal of Big Data 6, 60.
Smith, L., 2018. A disciplined approach to neural network hyper-parameters: Part 1 -- learning rate, batch size, momentum, and weight decay.
Solovyev, R., Wang, W., Gabruseva, T., 2021. Weighted boxes fusion: Ensembling boxes from different object detection models. Image and Vision Computing 107, 104117.
Srinivasan, R., Sorial, G.A., 2011. Treatment of taste and odor causing compounds 2-methyl isoborneol and geosmin in drinking water: A critical review. Journal of Environmental Sciences 23, 1-13.
State Administration for Market Regulation, National Standardization Administration, 2022. National Standard GB 5749-2022 for Drinking Water Quality. Standards Press of China, Beijing.
Suffet, M., Braithwaite, S., Zhou, Y., Bruchet, A., 2019. The drinking water taste-and-odour wheel after 30 years.
Trinh, Q.N., 2025. AI-Based Tracking of Floating Algal Mats for Drinking Water Quality Management, Department of Environmental Engineering. National Cheng Kung University, National Digital Library of Theses and Dissertations in Taiwan.
Vaičiūtė, D., Bučas, M., Bresciani, M., Dabulevičienė, T., Gintauskas, J., Mėžinė, J., Tiškus, E., Umgiesser, G., Morkūnas, J., De Santi, F., Bartoli, M., 2021. Hot moments and hotspots of cyanobacteria hyperblooms in the Curonian Lagoon (SE Baltic Sea) revealed via remote sensing-based retrospective analysis. Science of The Total Environment 769, 145053.
Vicino, F., Popeo, G., Santoro, F., Pascuzzi, S., Paciolla, F., 2026. Farmer-Friendly Approach for Table Grape Bunch Detection Using the Roboflow Platform, Agriculture, p. 218.
Wang, F., 2025. Improving YOLOv11 for marine water quality monitoring and pollution source identification. Scientific Reports 15, 21367.
Wang, Y., Liu, S., He, Y., Zhang, Y., 2025. YOLO11-YX: An efficient algorithm for marine debris target detection. Marine Pollution Bulletin 221, 118511.
Wang, Z., Zhang, H., Guo, Z., Fan, Y., Xu, L., Li, X., Wang, C., 2026. CDANet: A context-detail aware network for marine oil spill detection in SAR imagery. Journal of Hazardous Materials 501, 140683.
Wei, X., Yin, L., Zhang, L., Wu, F., 2024. DV-DETR: Improved UAV Aerial Small Target Detection Algorithm Based on RT-DETR, Sensors, p. 7376.
Welch, G., Bishop, G., 2006. An Introduction to the Kalman Filter.
Wert, E.C., Rosario-Ortiz, F.L., 2013. Intracellular Organic Matter from Cyanobacteria as a Precursor for Carbonaceous and Nitrogenous Disinfection Byproducts. Environmental Science & Technology 47, 6332-6340.
Wojke, N., Bewley, A., Paulus, D., 2017. Simple online and realtime tracking with a deep association metric.
Wu, D., Li, R., Zhang, F., Liu, J., 2019. A review on drone-based harmful algae blooms monitoring. Environmental Monitoring and Assessment 191, 211.
Wu, J., Cao, Y., Wu, S., Parajuli, S., Zhao, K., Lee, J., 2025. Current Capabilities and Challenges of Remote Sensing in Monitoring Freshwater Cyanobacterial Blooms: A Scoping Review. Remote Sensing 17, 918.
Xue, B., Cheng, Y., Ding, K., Pan, C., Xiang, S., 2026. USVTrack: A Benchmark for Multi-Object Tracking in Complex Water Surface Scenes. IEEE Transactions on Circuits and Systems for Video Technology 36, 1029-1044.
Yilmaz, A., Javed, O., Shah, M., 2006. Object tracking: A survey. ACM Comput. Surv. 38, 13–es.
Young, W.F., Horth, H., Crane, R., Ogden, T., Arnott, M., 1996. Taste and odour threshold concentrations of potential potable water contaminants. Water Research 30, 331-340.
Zahir, M., Su, Y., Shahzad, M.I., Ayub, G., Rahman, S.U., Ijaz, J., 2024. A review on monitoring, forecasting, and early warning of harmful algal bloom. Aquaculture 593, 741351.
Zamyadi, A., Choo, F., Newcombe, G., Stuetz, R., Henderson, R.K., 2016. A review of monitoring technologies for real-time management of cyanobacteria: Recent advances and future direction. TrAC Trends in Analytical Chemistry 85, 83-96.
Zhang, Y., Sun, P., Jiang, Y., Yu, D., Weng, F., Yuan, Z., Luo, P., Liu, W., Wang, X., 2022. ByteTrack: Multi-Object Tracking by Associating Every Detection Box.
Zhao, Y., Wenyu, L., Shangliang, X., Jinman, W., Guanzhong, W., Qingqing, D., Yi, L., Jie, C., 2024. DETRs Beat YOLOs on Real-time Object Detection.
Zou, Y., Wang, X., Wang, L., Chen, K., Ge, Y., Zhao, L., 2022. A High-Quality Instance-Segmentation Network for Floating-Algae Detection Using RGB Images. Remote Sensing 14, 6247.