本文選題：光斑檢測(cè)　切入點(diǎn)：圖像處理　出處：《北京交通大學(xué)》2017年碩士論文

【摘要】：隨著計(jì)算機(jī)技術(shù)、數(shù)字信號(hào)處理技術(shù)和光電探測(cè)器件的不斷發(fā)展和進(jìn)步,基于光纖光斑檢測(cè)的各項(xiàng)技術(shù)的復(fù)雜程度和成本大幅降低,促進(jìn)了光纖光斑檢測(cè)在傳感等研究領(lǐng)域的發(fā)展。多模光纖由于具有較大的芯徑,支持大量對(duì)外界變化非常敏感的高階模式,使基于多模光纖結(jié)構(gòu)的光斑檢測(cè)技術(shù)具有很高的應(yīng)用價(jià)值。本文對(duì)多模光纖光斑檢測(cè)技術(shù)新的應(yīng)用場(chǎng)景和光斑圖像處理技術(shù)展開(kāi)了研究,獲得的主要研究成果如下:針對(duì)目前模式分析領(lǐng)域,解析法對(duì)波導(dǎo)結(jié)構(gòu)要求很高,而數(shù)值分析方法存在著計(jì)算精度和計(jì)算機(jī)性能之間相互制約的局限性,本文將圖像處理技術(shù)中灰度共生矩陣引入光斑處理算法中,提出了基于光斑圖像灰度共生矩陣對(duì)波導(dǎo)模式數(shù)量進(jìn)行分析的方法,詳細(xì)分析了多模光纖中支持不同模式數(shù)量時(shí)光纖端面光斑與圖像灰度共生矩陣特征值的關(guān)系,得到了模式數(shù)量和灰度共生矩陣特征值能量(energy)、一致性(homogeneity)之間的近似線性關(guān)系,并進(jìn)行了實(shí)驗(yàn)驗(yàn)證,之后將這一結(jié)論應(yīng)用于對(duì)光纖內(nèi)模式數(shù)量的定性分析。通過(guò)利用單模-多模結(jié)構(gòu)中兩種光纖偏移量的不同來(lái)模擬進(jìn)行微位移傳感的方式,對(duì)光斑圖像中灰度共生矩陣特征值與位移量的關(guān)系進(jìn)行了研究。并利用此方法同時(shí)對(duì)具有不同彎曲半徑的光纖的輸出光斑進(jìn)行了研究。由于兩種光纖不同偏移對(duì)接情況和不同彎曲情況下在多模光纖中激發(fā)出的模式數(shù)量不同,根據(jù)多模光纖光斑的灰度共生矩陣特征值能量和一致性與光纖中激發(fā)出的模式數(shù)量近似呈線性關(guān)系的結(jié)論,對(duì)偏移量和彎曲半徑與光斑圖像灰度共生矩陣特征值之間的關(guān)系進(jìn)行了仿真分析與實(shí)驗(yàn)驗(yàn)證,獲得了兩個(gè)特征值能量和一致性與不同光纖相對(duì)位移量和不同彎曲半徑呈近似線性的關(guān)系,從而表明,可以通過(guò)對(duì)多模光纖光斑圖像灰度矩陣特征值能量和一致性對(duì)位移量和彎曲半徑進(jìn)行定量分析來(lái)實(shí)現(xiàn)光纖微位移傳感和光纖彎曲傳感的研究。
[Abstract]:With the continuous development and progress of computer technology, digital signal processing technology and photodetectors, the complexity and cost of various technologies based on optical fiber spot detection have been greatly reduced. It promotes the development of optical spot detection in sensing and other fields. Because of its large core diameter, multimode optical fiber supports a large number of high order modes which are sensitive to external changes. It makes the spot detection technology based on multimode optical fiber structure have high application value. In this paper, the new application scene and image processing technology of multimode optical fiber spot detection technology are studied. The main research results obtained are as follows: in the field of mode analysis, the analytical method requires the waveguide structure very high, but the numerical analysis method has the limitation of the mutual restriction between the calculation precision and the computer performance. In this paper, the gray level co-occurrence matrix in image processing is introduced into the spot processing algorithm, and a method to analyze the number of waveguide modes based on the gray level co-occurrence matrix of the spot image is proposed. In this paper, the relationship between the number of optical fiber facets and the eigenvalues of gray level co-occurrence matrix in multimode optical fiber is analyzed in detail, and the approximate linear relationship between the number of patterns and eigenvalues of gray co-occurrence matrix energy energy, consistency homogeneityis obtained. This conclusion is applied to the qualitative analysis of the number of internal modes in the fiber. By using the difference between the two optical fiber offsets in the single-mode and multi-mode structure, the method of micro-displacement sensing is simulated. The relationship between the eigenvalues of gray level co-occurrence matrix and displacement in spot image is studied. The output spot of fiber with different bending radius is also studied by using this method. The number of modes excited in multimode optical fibers varies from case to case and under different bending conditions, According to the conclusion that the eigenvalue energy and consistency of gray level co-occurrence matrix of multimode optical fiber spot are approximately linear with the number of modes excited in the optical fiber. The relationship between offset and bending radius and the eigenvalue of gray level co-occurrence matrix of spot image is simulated and verified by experiments. The energy and consistency of the two eigenvalues are approximately linear with the relative displacement and bending radius of different optical fibers. The research of fiber micro displacement sensor and fiber bending sensor can be realized by quantitative analysis of displacement and bending radius through the energy and consistency of grayscale matrix eigenvalue of multimode optical fiber spot image.
【學(xué)位授予單位】：北京交通大學(xué)
【學(xué)位級(jí)別】：碩士
【學(xué)位授予年份】：2017
【分類號(hào)】：TP391.41

【參考文獻(xiàn)】

相關(guān)期刊論文 前3條

1 李強(qiáng);黃澤鋏;徐雅芹;張凌云;史驥;王智;;基于單模-多模-單模光纖模間干涉的傳感系統(tǒng)[J];紅外與激光工程;2014年05期

2 何偉剛;;基于改進(jìn)的四叉樹(shù)Sobel算子圖像分割研究[J];計(jì)算機(jī)仿真;2012年04期

3 楊春;張華勇;張敏;廖延彪;;光纖水聽(tīng)器時(shí)分復(fù)用檢測(cè)電路帶寬與串?dāng)_分析[J];光電工程;2010年07期

相關(guān)碩士學(xué)位論文 前6條

1 吳紅梅;基于CCD的激光光斑檢測(cè)系統(tǒng)的設(shè)計(jì)[D];黑龍江大學(xué);2014年

2 劉友江;VCH-F2腹腔切片圖像分割技術(shù)研究[D];重慶交通大學(xué);2013年

3 申曉軍;粒子群算法在混合尺寸泡沫圖像分割中的應(yīng)用[D];中南大學(xué);2010年

4 吳雷;基于多模光纖聲發(fā)射檢測(cè)的理論與實(shí)驗(yàn)研究[D];哈爾濱工程大學(xué);2009年

5 葉艷芳;城區(qū)復(fù)雜背景條件下的車(chē)輛檢測(cè)算法研究[D];重慶大學(xué);2008年

6 焦鳳萍;合成孔徑雷達(dá)圖像的預(yù)處理方法研究[D];安徽大學(xué);2007年

，

本文編號(hào)：1697744