請用此 Handle URI 來引用此文件：
http://tdr.lib.ntu.edu.tw/jspui/handle/123456789/80813完整後設資料紀錄
| DC 欄位 | 值 | 語言 |
|---|---|---|
| dc.contributor.advisor | 柯佳吟(Chia-Ying Ko) | |
| dc.contributor.author | Chih-Wei Tu | en |
| dc.contributor.author | 涂芝瑋 | zh_TW |
| dc.date.accessioned | 2022-11-24T03:17:31Z | - |
| dc.date.available | 2021-11-08 | |
| dc.date.available | 2022-11-24T03:17:31Z | - |
| dc.date.copyright | 2021-11-08 | |
| dc.date.issued | 2021 | |
| dc.date.submitted | 2021-10-05 | |
| dc.identifier.citation | Asch, R. G. (2015). Climate change and decadal shifts in the phenology of larval fishes. in the California Current ecosystem. Proceedings of the National Academy of Sciences of the United States of America, 112(30), 4065-4074. https://doi.org/10.1073/pnas.1421946112 Asch, R. G., Stock, C. A., Sarmiento, J. L. (2019). Climate change impacts on mismatches between phytoplankton blooms and fish spawning phenology. Global Change Biology, 25(8), 2544-2559. https://doi.org/10.1111/gcb.14650 Auth, T. D., Daly, E. A., Brodeur, R. D., Fisher, J. L. (2018). Phenological and distributional shifts in ichthyoplankton associated with recent warming in the northeast Pacific Ocean. Global Change Biology, 24(1), 259-272. https://doi.org/10.1111/gcb.13872 Boeing, W. J., Duffy-Anderson, J. T. (2008). Ichthyoplankton dynamics and biodiversity in the Gulf of Alaska: Responses to environmental change. Ecological Indicators, 8(3), 292-302. https://doi.org/10.1016/j.ecolind.2007.03.002 Brey, T., Pauly, D. (1986). Electronic length frequency analysis: a revised and expanded user's guide to ELEFAN 0, 1 and 2. Chiu, T., Chen, C. (2001). Growth and temporal variation of two Japanese anchovy cohorts during their recruitment to the East China Sea. Fisheries Research, 53(1), 1-15. https://doi.org/10.1016/S0165-7836(00)00294-0 Chiu, T.-S., Chen, C.-L., Young, S.-S. (1999). Age and growth of two co-occurred anchovy species (Encrasicholina puntifer and E. heteroloba) during autumn larval anchovy fishing season in I-lan Bay, NE Taiwan. Journal of the Fisheries Society of Taiwan, 26(4), 183-190. https://doi.org/10.29822/JFST.199912.0001 Gray, P. C., Bierlich, K. C., Mantell, S. A., Friedlaender, A. S., Goldbogen, J. A., Johnston, D. W. (2019). Drones and convolutional neural networks facilitate automated and accurate cetacean species identification and photogrammetry. Methods in Ecology and Evolution, 10(9), 1490-1500. https://doi.org/10.1111/2041-210X.13246 Gray, P. C., Fleishman, A. B., Klein, D. J., McKown, M. W., Bezy, V. S., Lohmann, K. J., Johnston, D. W. (2019). A convolutional neural network for detecting sea turtles in drone imagery. Methods in Ecology and Evolution, 10(3), 345-355. https://doi.org/10.1111/2041-210X.13132 Haibo, H., Garcia, E. A. (2009). Learning from Imbalanced Data. IEEE Transactions on Knowledge and Data Engineering, 21(9), 1263-1284. https://doi.org/10.1109/TKDE.2008.239 Hendrycks, D., Gimpel, K. (2016). A baseline for detecting misclassified and out-of-distribution examples in neural networks. arXiv preprint arXiv:1610.02136. Hordyk, A., Ono, K., Valencia, S., Loneragan, N., Prince, J. (2015). A novel length-based empirical estimation method of spawning potential ratio (SPR), and tests of its performance, for small-scale, data-poor fisheries. ICES Journal of Marine Science, 72(1), 217-231. https://doi.org/10.1093/icesjms/fsu004 Huang, C.-H., Chen, C.-S., Chiu, T.-S. (2018). Growth of buccaneer anchovy (Encrasicholina punctifer) during juvenile stage in the waters off southwestern Taiwan. Journal of Taiwan Fisheries Research, 26(1), 53-61. Maack, G., George, M. R. (1999). Contributions to the reproductive biology of Encrasicholina punctifer Fowler, 1938 (Engraulidae) from West Sumatra, Indonesia. Fisheries Research, 44(2), 113-120. https://doi.org/10.1016/S0165-7836(99)00084-3 Morales-Nin, B. (1989). Growth determination of tropical marine fishes by means of otolith interpretation and length frequency analysis. Aquatic Living Resources, 2(4), 241-253. https://doi.org/10.1051/alr:1989029 Nielsen, J. M., Rogers, L. A., Brodeur, R. D., Thompson, A. R., Auth, T. D., Deary, A. L., Duffy-Anderson, J.T., Galbraith, M., Koslow, J. A. Perry, R. I. (2021). Responses of ichthyoplankton assemblages to the recent marine heatwave and previous climate fluctuations in several Northeast Pacific marine ecosystems. Global Change Biology, 27(3), 506-520. https://doi.org/10.1111/gcb.15415 Schwamborn, R., Mildenberger, T., Taylor, M. (2019). Assessing sources of uncertainty in length-based estimates of body growth in populations of fishes and macroinvertebrates with bootstrapped ELEFAN. Ecological Modelling, 393, 37-51. https://doi.org/10.1016/j.ecolmodel.2018.12.001 Selvaraju, R. R., Cogswell, M., Das, A., Vedantam, R., Parikh, D., Batra, D. (2017). Grad-cam: Visual explanations from deep networks via gradient-based localization. Proceedings of the IEEE international conference on computer vision. 10.1007/s11263-019-01228-7 Simonyan, K., Vedaldi, A., Zisserman, A. (2013). Deep inside convolutional networks: Visualising image classification models and saliency maps. arXiv preprint arXiv:1312.6034. Somers, I. (1988). On a seasonally oscillating growth function. Fishbyte, 6(1), 8-11. Taylor, M. H., Mildenberger, T. K. (2017). Extending electronic length frequency analysis in R. Fisheries Management and Ecology, 24(4), 330-338. https://doi.org/10.1111/fme.12232 Tu, C.-Y., Tseng, Y.-H., Chiu, T.-S., Shen, M.-L., Hsieh, C.-H. (2012). Using coupled fish behavior-hydrodynamic model to investigate spawning migration of Japanese anchovy, Engraulis japonicus, from the East China Sea to Taiwan. Fisheries Oceanography, 21(4), 255-268. https://doi.org/10.1111/j.1365-2419.2012.00619.x Van der Maaten, L., Hinton, G. (2008). Visualizing data using t-SNE. Journal of machine learning research, 9(11). Wang, D., Shen, Z., Shao, J., Zhang, W., Xue, X., Zhang, Z. (2015). Multiple granularity descriptors for fine-grained categorization. IEEE International Conference on Computer Vision (ICCV), 2399-2406. https://doi.org/10.1109/ICCV.2015.276 Wang, K., Zhang, C., Sun, M., Xu, B., Ji, Y., Xue, Y., Ren, Y. (2021). Fishing pressure and lifespan affect the estimation of growth parameters using ELEFAN. Fisheries Research, 238. https://doi.org/10.1016/j.fishres.2021.105903 Wang, K., Zhang, C., Xu, B., Xue, Y., Ren, Y. (2020). Selecting optimal bin size to account for growth variability in Electronic LEngth Frequency ANalysis (ELEFAN). Fisheries Research, 225, 105474. https://doi.org/10.1016/j.fishres.2019.105474 Wang, Y.-C., Chen, W.-Y., Chen, Y.-K., Kuo, Y.-C., Lee, M.-A. (2018). Winter abundance and species composition of anchovy larvae associated with hydrological conditions in the coastal waers of Tanshul, Taiwan. Journal of Marine Science and Technology, 26, 455-474. https://doi.org/10.6119/JMST.2018.06_(3).0018 Wright, P. (1992). Ovarian development, spawning frequency and batch fecundity in Encrasicholina heteroloba (Ruppell, 1858). Journal of Fish Biology, 40(6), 833-844. https://doi.org/10.1111/j.1095-8649.1992.tb02630.x Yuxin, P., Xiangteng, H., Junjie, Z. (2018). Object-Part Attention Model for Fine-Grained Image Classification. IEEE Transactions on Image Processing, 27(3), 1487-1500. https://doi.org/10.1109/TIP.2017.2774041 Zhou, B., Khosla, A., Lapedriza, A., Oliva, A., Torralba, A. (2016). Learning deep features for discriminative localization. 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2921-2929. https://doi.org/10.1109/CVPR.2016.319. | |
| dc.identifier.uri | http://tdr.lib.ntu.edu.tw/jspui/handle/123456789/80813 | - |
| dc.description.abstract | 仔稚魚作為低營養階層以及環境敏感指標的物種，會透過由下而上的級聯反應(cascade effect)影響中、高階層的物種進而影響整個生態系；因此，仔稚魚對於了解海洋生態系統營養階層的動態以及與環境變化的關係至關重要。魩鱙漁業是亞洲地區重要且特殊的沿海漁業，考量如何平衡資源利用與永續發展，以及社會經濟與生態保育等多個面向，成為產官學界重要的課題。此外，混獲，即捕捉到非魩鱙漁業主要目標物種的鯷鯡科，也是另一個備受關注的議題。然而，長期的仔稚魚研究和即時的資源評估最大的困難是快速的物種識別和體長測量，因為辨識物種需要專業的辨識能力，且仔稚魚眾多的數量也導致測量需要花費大量時間。隨著科技的進步，神經網路(Neural Networks)成為自動化上述過程的一個可能性。因此，本研究的目的是採用深度學習的方法來建置仔稚魚物種辨識和體長測量的系統。我們收集了2018年至2020年台灣 35 種仔稚魚的影像資料共34580筆記錄作為數據集(Dataset)，並將數據集依照70%-20%-10%分割成訓練集(Training set)、驗證集(Validation set)及測試集(Testing set)且建立了一個卷積神經網絡(Convolutional Neural Networks)模型，接著結合生物及生態資訊來試圖提高模型的準確性(Accuracy)和精度(Precision)。此外，我們收集了另一筆獨立的仔稚魚影像資料用於測試最終模型的泛化性(Generalization)。最後，使用自動體長測量模型分析2020年宜蘭梗訪秋漁期的主要魩鱙漁獲魚種來展示模型的適用性(Applicability)，並進一步透過體長頻度資料計算物種成長參數。結果顯示，最終模型的準確率在物種和科別層級上，分別達到了95.5% 和99.6%。體長自動測量和手動測量相比下其精度誤差在 3.5%以內。在實際應用情況下，該模型在科別層級上仍然達到了99.8%的準確率，而在物種層級上只有77.9%的準確率。此外，使用自動體長測量模型分析時，異葉公鯷(Encrasicholina heteroba)和刺公鯷(Encrasicholina punctifer)的體長分佈顯示出週期性變化且產卵間隔約為 2.5週，日本鯷(Engraulis japonicus)則因為數量少無法觀察到週期性的變化。儘管在獨立數據集上的表現較差，我們的結果仍證明了使用深度學習技術來自動化仔稚魚物種辨識和體長測量的潛力，不僅可以有效地獲取基礎的生態資訊，如物種組成與族群體長分佈，並進一步提供即時的管理策略。但是，此系統仍然需要更多的研究來改善物種辨識中的較差泛化能力，並提高體長測量的精準度。 | zh_TW |
| dc.description.provenance | Made available in DSpace on 2022-11-24T03:17:31Z (GMT). No. of bitstreams: 1 U0001-0410202115281300.pdf: 7137184 bytes, checksum: 297be694844e6286195dcbda6b450972 (MD5) Previous issue date: 2021 | en |
| dc.description.tableofcontents | Acknowledgement i 中文摘要 ii Abstract iv Contents vii List of Figures xi List of Tables xiv Chapter 1 Introduction 1 Chapter 2 Preliminary 5 2.1 Basic components of neural network 5 2.1.1 Overview of (fully-connected) Neural Networks 5 2.1.2 Overview of convolutional neural networks 5 2.1.2.1 Convolution layers 6 2.1.2.2 Activation function 7 2.1.2.3 Pooling layers 8 2.1.2.4 Objective function 8 2.2 Classification and regression 9 2.3 Imbalanced Problem 10 2.4 Visualization 11 2.4.1 feature space visualization (t-SNE) 11 2.4.2 Model visualization (Saliency/Grad-cam) 12 Chapter 3 Materials and methods (system proposal) 14 3.1 Overview of workflow 14 3.2 Fishery images 14 3.2.1 Dataset 14 3.2.2 Dataset1 14 3.2.3 KenFong2020 15 3.3 Development environment 15 3.4 Model development - Classification design intuition 15 3.4.1 Transfer learning and finetune pretrained model 15 3.4.2 Hierarchical model 16 3.4.3 Regional features 17 3.5 Model development - Body length measurement 17 3.6 Model evaluation 18 3.7 Model generalization and model application 19 Chapter 4 Results 21 4.1 Larval species classification 21 4.1.1 Select and finetune baseline model 21 4.1.2 Hierarchical model 21 4.1.2.1 Engraulid classifier 22 4.1.2.2 Clupeid classifier 23 4.1.3 Regional features (domain model) 23 4.1.4 Final model proposal and visualization 24 4.2 Body length measurement 25 4.3 Generalization of the model 25 4.3.1 Dataset1 25 4.3.2 KenFong dataset 26 Chapter 5 Discussion and Futurework 28 5.1 System-related(classification) 28 5.1.1 t-SNE of Engraulidae/ Clupeidae using quintuplet loss 28 5.1.2 Quality of picture and automation of collecting image data 28 5.1.3 Class never seen before in the model 29 5.1.4 Suitability of regional features or its replacement 29 5.1.5 Transparency of tail 31 5.2 Body length distribution in KenFong 2020 32 5.2.1 Spawning interval 32 5.2.2 Growth model estimation 32 Chapter 6 Conclusion 34 Reference 35 Figures 40 Tables 80 | |
| dc.language.iso | en | |
| dc.subject | 仔稚魚 | zh_TW |
| dc.subject | 體長估計 | zh_TW |
| dc.subject | 物種辨識 | zh_TW |
| dc.subject | 卷積神經網路 | zh_TW |
| dc.subject | 魩鱙漁業 | zh_TW |
| dc.subject | larval fish | en |
| dc.subject | species classification | en |
| dc.subject | convolutional neural network | en |
| dc.subject | larva fisheries | en |
| dc.subject | body length measurement | en |
| dc.title | 仔稚魚物種辨識及體長測量自動化系統的建置 | zh_TW |
| dc.title | Establishment of specialized convolutional neural network for larval fish identification and body length measurement | en |
| dc.date.schoolyear | 109-2 | |
| dc.description.degree | 碩士 | |
| dc.contributor.oralexamcommittee | 邵皓強 (Hsin-Tsai Liu),蔡欣穆(Chih-Yang Tseng),丘臺生,陳仲吉 | |
| dc.subject.keyword | 仔稚魚,魩鱙漁業,卷積神經網路,物種辨識,體長估計, | zh_TW |
| dc.subject.keyword | larval fish,larva fisheries,convolutional neural network,species classification,body length measurement, | en |
| dc.relation.page | 93 | |
| dc.identifier.doi | 10.6342/NTU202103535 | |
| dc.rights.note | 同意授權(限校園內公開) | |
| dc.date.accepted | 2021-10-07 | |
| dc.contributor.author-college | 生命科學院 | zh_TW |
| dc.contributor.author-dept | 漁業科學研究所 | zh_TW |
| 顯示於系所單位： | 漁業科學研究所 | |
文件中的檔案：
| 檔案 | 大小 | 格式 | |
|---|---|---|---|
| U0001-0410202115281300.pdf 授權僅限NTU校內IP使用（校園外請利用VPN校外連線服務） | 6.97 MB | Adobe PDF |
系統中的文件，除了特別指名其著作權條款之外，均受到著作權保護，並且保留所有的權利。