以電子觀察員取代派遣實體觀察員是未來的趨勢，但是目前電子觀察員系統，存在著無法即時處理影像及需要大量人力截取漁獲資訊的缺點，因此本研究將探討如何自動並且即時將魚獲資料，透過衛星通訊即時傳輸到岸端伺服器，降低人力成本。以電子觀察員系統為基礎，將智慧化偵測、定位影像中的魚體位置即時化，並且增加魚體長度自動量測，電子標籤、衛星通訊即時傳輸硬體部分進行系統整合，達到智慧型的電子觀察員系統，本研究建立一個初步的系統環境進行模擬。
本研究基於魚體輪廓進行魚體長度量測，以下提出三種方法進行；方法一—最小包覆圓法，利用一個能夠包覆輪廓的最小外接圓，外接圓的直徑即為魚體長度。由於影像品質、燈光亮度、甲板地毯等等因素，使得魚體輪廓完整度不如預期。最小包覆圓無法確定框選到的點是尾鰭、尾叉或是誤判斷，僅能進行初步的長度判斷。利用方法一分別對鮪魚、鬼頭刀與鯊魚各100張實船上魚體影像進行量測，量測誤差分別為8.68%、4.69%與3.96%。
為了改善方法一提出單一角度量測法。將魚體輪廓轉正為水平，利用通過魚體輪廓重心的水平線，水平線與輪廓線的交點視為魚吻點與尾叉點，量測兩點之間的距離。因為上述實船環境影像缺點，實船影像在轉正時會產生誤差，量測誤差分別為15.23%、6.87%與21.52%。
為了將參考線方法更有效率的應用於實際環境，多角度量測法不再進行影像轉正，而是延伸單一角度參考線的基礎，增加參考線數量，找尋魚吻部與尾叉部進行長度的量測。應用在模擬環境影像的量測誤差為1.46%，實船影像量測誤差分別為13.21%、5.22%與18.62%。
將魚體偵測、漁獲計數、魚長量測、魚種辨識與電子標籤軟硬體部分，進行系統整合與測試，並且討論系統測試結果。

The electric observer system (EOS) will be the mainstream in the future. However, it consists of disadvantages in this system. Including it cannot process image immediately, needing amount of human resource to get the fish data. In this research, trying to find out how to get fish data automaticed. This system is from prototype of version one EOS. In order to develop intelligent system, it should combine with intelligence fish detection, automatic fish length measurement, electronic label and satellite transmission. This research sets up an environment to simulate this system.
This research measures fish length base on contour. First, using a circumcircle to cover whole contour. The radius of circumcircle, which is the fish length. However, there are many factor contribute to defect of fish contour, including image quality, brightness…Method 1 can only measure preliminary.
In order to optimize method 1, method 2 rotates the contoue into horizontal. Using a baseline which through the centroid of contour to detect intersation. Considering the length of intersations to be fish length. The results of method 2 are not as expected because of the defect of fish contour.
For the purpose of applying baseline detection to real environment. The method 3 increases the number of baseline with different angle instead of rotating the image, and the error decreases effective.
Integrating fish detection, length measurement, category identification, electronic label and satellite transmission to the system. This research will discuss with the effectiveness of the system.

[1] Chan, D., Hockaday, S., Tillett, R., & Ross, L. A trainable n-tuple pattern classifier and its application for monitoring fish underwater. Paper presented at the Image Processing And Its Applications, 1999. Seventh International Conference on (Conf. Publ. No. 465), 1999.
[2] Ching-Lu Hsieh、Hsiang-Yun Chang、Fei-Hung Chen、Jhao-Huei Liou、Shui-Kai Chang、Ta-Te Lin,”A simple and effective digital imaging approach for tuna fish length measurement compatible with fishing operations”, Computers and Electronics in Agriculture,Vol.75, p.44-51, 2010.
[3] C. Tomasi andR. Manduchi, Bilateral Filtering for Gray and Color Image. Computer Vision, Sixth International Conference, 1988.
[4] Howard McElderry、Maria Jose Pria、Morgan Dyas、Randy McVeigh,” A PILOT STUDY USING EM IN THE HAWAIIAN LONGLINE FISHERY”,Archipelago Marine Research Ltd,2010。
[5] J. Koenderink and A. J. van Doorn, “Surface shape and curvature scales,” Image and Vision Computing, vol. 10, pp. 557-565, 1992.
[6] Julie Bonney,Katy McGauley,”Testing the Use of Electronic Monitoring to Quantify At‐sea Halibut Discards in the Central Gulf of Alaska Rockfish Fishery”,EFP 07‐02 Final Report,2008。
[7] Kaiming He, Xiangyu Zhang, Shaoqing Ren, Jian Sun, “Deep Residual Learning for Image Recognition. ICML, 2016.
[8] Lee, D., Redd, S., Schoenberger, R., Xu, X., & Zhan, P. An automated fish species classification and migration monitoring system. Paper presented at the Industrial Electronics Society, 2003. IECON'03. The 29th Annual Conference of the IEEE, 2003.
[9] Lee, D.-J., Schoenberger, R. B., Shiozawa, D., Xu, X., & Zhan, P. Contour matching for a fish recognition and migration-monitoring system. Paper presented at the Optics East,2004
[10] Lee, J.-H., Wu, M.-Y., & Guo, Z.-C. A tank fish recognition and tracking system-using computer vision techniques. Paper presented at the Computer Science and Information Technology (ICCSIT), 2010 3rd IEEE International Conference on, 2010.
[11] Spampinato, C., Giordano, D., Di Salvo, R., Chen-Burger, Y.-H. J., Fisher, R. B., & Nadarajan, G. Automatic fish classification for underwater species behavior understanding. Paper presented at the Proceedings of the first ACM international workshop on Analysis and retrieval of tracked events and motion in imagery streams, 2010.
[12] Steve Dunn and Ian Knuckey, Preliminary Report to the Technical and Compliance Committee. Western and Central Pacific Fisheries Commission, 2013.
[13] Thanh Nguyen Duc, P. Ogunbona, Wanqing Li, “Human detection based on weighted template matching,”IEEE Int. Conf. on Multimedia and Expo, pp. 634-637, June 2009.
[14] Yao, H., Duan, Q., Li, D., & Wang, J. An improved K-means clustering algorithm for fish image segmentation. Mathematical and Computer Modelling, 58(3), 790-798, 2013.
[15] Zion, B., Alchanatis, V., Ostrovsky, V., Barki, A., & Karplus, I. Real-time underwater sorting of edible fish species. Computers and Electronics in Agriculture, 56(1), 34-45, 2007.
[16] 吳俊樓，“運用影像處理技術擷取延繩釣揚繩作業影片中之漁獲資訊”，碩士論文，國立成功大學，2012。
[17] 吳庭瑋，“適用於中小型延繩釣漁船之自動監控與資訊擷取系統之開發”，碩士論文，國立成功大學，2014。
[18] 陳飛宏，“鮪釣船觀察員自動攝影系統之開發”，碩士論文，國立屏東科技大學，2007。
[19] 郭俊宏，“基於影像差分直方圖於揚繩作業影像中之魚體偵測”，碩士論文，國立成功大學，2016。
[20] 魏子量，“基於摺積類神經網路之魚類偵測與辨識”，碩士論文，國立成功大學，2016。