Please use this identifier to cite or link to this item:
標題: 機器學習應用於種雞辨識系統之研究
Study on Machine Learning for Recognition System of Breeder
作者: 李祐承
Yu-Cheng Li
關鍵字: 機器學習;卷積神經網路;種雞;Machine Learning;Convolutional Neural Network;Breeder
引用: 1.李淵百。2005。台灣土雞的育種改良與產業趨勢。農業生技產業季刊第二期。 2.林大貴。2017。TensorFlow+Keras深度學習人工智慧實務應用。博碩出版社。 3.泛科技。2017。2017 The AI Summit London 倫敦人工智慧峰會。醫療新境界。網址。上網日期:2018-03-08。 4.國立中興大學動物科學系家禽育種研究室。2009。台灣的土雞。台中:國立中興大學動物科學系家禽育種研究室。網址。上網日期:2018-01-14。 5.趙清賢、林旻蓉、賴元亮、蘇夢蘭、何玉珍、陳志峰、李淵百。2005。中國畜牧學會會誌第34卷(65-77)。 6.農委會。2010。土雞產業之輔導成果及展望。農政與農情,(220)。 7.農委會林務局。2003。台灣的雞畜牧場。台北 : 行政院農業委員會林務局。網址。上網日期:2018-01-14。 8.農委會。2006。土雞選拔模式。台北 : 行政院農業委員會。網址。上網日期:2018-01-14。 9.數位時代。2017。電商拚服務力跟風導入人工智慧。網址 上網日期:2018-03-08。 10.斎藤康毅。2017。Deep Learning:用Python進行深度學習的基礎理論實作。歐萊禮出版社。 11.Szegedy, C., W. Liu, Y. Jia, P. Sermanet, S. Reed, D. Anguelov, D. Erhan, V. Vanhoucke, and A. Rabinovich, 2014, in Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pp. 1-9. 12.Rosenblatt, F., 1958, The Perceptron: A Probabilistic Model for Information Storage and Organization in the Brain, Cornell Aeronautical Laboratory, Psychological Review, v65, No. 6, pp. 386–408. 13.Krizkevsky, A., I. Sutskever, and G. E. Hinton, 2012, Imagined classification with deep convolutional neural networks, In Advances in neural information processing systems, pp. 1097-1105. 14.He, K, X. Zhang, S. Ren, and J. Sun, 2016, Deep Residual Learning for Image Recognition, in IEEE Conference on Computer Vision and Pattern Recognition, pp. 770-778 15.Dumoulin, V.,and F. Visin, 2016, A guide to convolution arithmetic for deep learning, arXiv preprint arXiv:1603.07285. 16.LeCun, Y., L. Bottou, Y. Bengio, and P. Haffner, Gradient-based learning applied to document recognition, Proceedings of the IEEE, vol. 86, no. 11, pp. 2278-2324, 1998.
本研究之目的為改善業者種雞選拔系統,建立一套透過影像資料進行種雞篩選的種雞選拔人工智慧(Artificial Intelligence, AI)系統,可以幫助有色雞業者發展有效且快速的種雞選拔系統,系統除選出生育率性能良好的種雞,並由即早淘汱不適任的種雞,而大幅降低種雞選拔成本,如飼料成本,同時減少作業人力的需求與時間的消耗,本系統利用Google團隊開發的Tensorflow與Keras學習框架進行卷積神經網路的辨識演算法,期望能找出種雞與非種雞之間的差異性,為取得足夠且具代表性的雞隻影像,本研究同時開發了利用誘食的方式所建立的種雞選拔校正試驗系統。在實驗中將種雞與非種雞的影像到卷積神經網路影像共600張,做出的結果辨識率最高達60.03 %。

The purpose of this study was to establish a breeders selection system by adopting an artificial intelligence (AI) system by filtering breeders through machine vision. It can help the colored chicken industry to improve its productive efficiency by adopting this effective and rapid breeder selection system. In addition to selecting breeders based on the fertility performance via the system, and eliminating the breeders what were ineligibility as soon as possible will dramatically reduce the cost of breeder selection, such as feed costs, and also reduce the labor demand and time consumption. The system used the Tensorflow and Keras learning frameworks is developed by the Google team to proceed the identification algorithm of Convolutional Neural Network. The developed system is used to find the difference between breeders and non-breeders. In order to obtain sufficient representative images of chicken samples, this study developed a breeder calibration and test system, which by altering feed position periodically on both ends of the image taking tunnel in the calibrating system. In the experiment, 600 images of breeder and non-breeder were taken and sent to Convolutional Neural Network. So far, the recognition rate of the developed system was up to 60.03%.
Rights: 不同意授權瀏覽/列印電子全文服務
Appears in Collections:生物產業機電工程學系

Files in This Item:
File SizeFormat Existing users please Login
nchu-107-7105040105-1.pdf3.04 MBAdobe PDFThis file is only available in the university internal network    Request a copy
Show full item record

Google ScholarTM


Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.