本文選題：光流法 + 流動顯示�。� 參考：《西南科技大學》2017年碩士論文

【摘要】：光流法已廣泛應(yīng)用于計算機視覺和圖像檢測領(lǐng)域。光流法涉及到圖像處理、人工智能和模式識別等學科,是當今圖像檢測領(lǐng)域的研究熱點之一,其成果已逐步推廣應(yīng)用到航天、軍事、工業(yè)、醫(yī)學等行業(yè)。本文圍繞基于光流法的流場測量顯示技術(shù)展開以下研究:(1)針對現(xiàn)有HS光流法的控制參數(shù)不能隨計算結(jié)果進行自優(yōu)化,故,本文將衡量光流精度的參數(shù)作為反饋控制參數(shù),作用于計算過程,同時將光流值插值回圖片進行多次迭代來提高估計精度。(2)針對光流出現(xiàn)的大位移問題,提出小波多分辨率分層細化處理方法,使得光流計算結(jié)果更加精確。實驗結(jié)果表明:改進后的光流算法較傳統(tǒng)HS算法具有更高的準確性。(3)開展基于光流法的流動顯示實驗研究。第一個是對機翼表面摩擦力線提取實驗,通過實驗顯示分析得到本文的光流算法可以顯示空氣流過機翼表面所產(chǎn)生的摩擦力線;第二個為兩種運動圖像速度場提取實驗,通過分析采集到的風洞背景紋影實驗數(shù)據(jù),驗證了本文算法可以顯示氣流的流動過程。(4)開展光流法在流場測量工程實踐中研究即對復雜流動結(jié)構(gòu)進行定量實驗研究。經(jīng)過試驗研究可以定量顯示空腔高速流動產(chǎn)生的波/渦/剪切層的位置、強弱,極其相互作用關(guān)系。本文算法對復雜流動結(jié)構(gòu)的定量測量具有一定的意義。
[Abstract]:Optical flow method has been widely used in computer vision and image detection. Optical flow method is one of the research hotspots in the field of image detection, which involves image processing, artificial intelligence and pattern recognition. Its achievements have been gradually applied to aerospace, military, industrial, medical and other industries. In this paper, the following research is carried out around the flow field measurement and display technology based on optical flow method: 1) since the control parameters of the existing HS optical flow method cannot be self-optimized with the calculation results, the parameters measuring the optical flow accuracy are taken as the feedback control parameters in this paper. In order to solve the problem of large displacement of optical flow, a wavelet multi-resolution hierarchical thinning method is proposed, which makes the calculation result of optical flow more accurate. The experimental results show that the improved optical flow algorithm has higher accuracy than the traditional HS algorithm. The first is the experiment of extracting the friction line of the wing surface, the optical flow algorithm of this paper can show the friction line produced by the air flowing through the wing surface, the second is the velocity field extraction experiment of two kinds of moving images. By analyzing the collected background schlieren data of wind tunnel, it is verified that the proposed algorithm can show the flow process of airflow. 4) the optical flow method is studied in the engineering practice of flow field measurement, that is, the quantitative experimental study on complex flow structure is carried out. The position, strength and interaction of the wave / vortex / shear layer produced by the high velocity flow in the cavity can be quantitatively demonstrated by the experimental study. This algorithm has certain significance for quantitative measurement of complex flow structures.
【學位授予單位】：西南科技大學
【學位級別】：碩士
【學位授予年份】：2017
【分類號】：TP391.41

【相似文獻】

相關(guān)期刊論文 前10條

1 孫正,郁道銀,陳曉冬,徐智,黃家祥,謝洪波;基于光流法的冠狀動脈造影圖像序列中血管運動的估計[J];工程圖學學報;2003年03期

2 韓雷;王洪慶;林隱靜;;光流法在強對流天氣臨近預報中的應(yīng)用[J];北京大學學報(自然科學版);2008年05期

3 龔大墉;汪治華;楊數(shù)強;劉巖;;基于目標的局部約束光流分析[J];重慶工學院學報(自然科學版);2009年07期

4 嚴強;黃增喜;曹麗萍;黃蓉剛;;光流法在機車安全行駛中的應(yīng)用[J];計算機應(yīng)用研究;2013年04期

5 馬鵬飛;楊金孝;;基于光流法的粒子圖像測速[J];科學技術(shù)與工程;2012年32期

6 張永亮;盧煥章;賀興華;謝耀華;;基于光流預測的直線對應(yīng)算法[J];信號處理;2010年05期

7 孫承志;熊田忠;吉順平;張家海;;基于差分的光流法在目標檢測跟蹤中的應(yīng)用[J];機床與液壓;2010年14期

8 王，

本文編號：1788213