查詢結果分析
來源資料
頁籤選單縮合
題名 | 應用類神經網路在流場影像質點運動之辨識=Using Artificial Neural Network to Identify Particle Motions in Fluids |
---|---|
作者 | 丁肇隆; 林銘崇; 曾慶深; Ting, Chao-lung; Lin, Ming-chung; Tsang, Hing-sum; |
期刊 | 國立臺灣大學工程學刊 |
出版日期 | 20021000 |
卷期 | 86 2002.10[民91.10] |
頁次 | 頁53-59 |
分類號 | 440.137 |
語文 | chi |
關鍵詞 | 無監督式類神經網路; 影像辨識; 瞬時速度; Unsupervised artificial neural network; Image identification; Instantaneous velocity; |
中文摘要 | 本研究要探討應用無監督式類神經網路,利用Particle Imaging Velocimetry (P.I.V)的方法所拍攝流場之影像當作試驗案例,再應用類神經網路方法作影像中質點運動之辨識,進而求得各影像之瞬間流場。為瞭解類神經網路方法之優缺點,流場影像也以傳統之cross-correlation 方式分析作比較,計算其影像之瞬間速度場,結果顯示類神經網路方法之計算速度較傳統方式為快,而且其質點之辨識率也較傳統之方式為好。 |
英文摘要 | The purpose of this research is to apply unsupervised artificial neural network to identify particle motions in a plunging breaker. The images of particle motions were captured in water using Particles imaging Velocimetry (PIV). The artificial neural network is applied to identify these particle movements. The identified particles can be used to calculate their instantaneous velocities. To demonstrate the advantages and disadvantages of neural network method, traditional cross-correlation method was used to find particle motions also. The results show that artificial neural network method can compute particle velocities faster than cross-correlation method can. Furthermore, the artificial neural network method has a higher identification rate. |
本系統之摘要資訊系依該期刊論文摘要之資訊為主。