【摘要】：隨著工業(yè)的迅速發(fā)展,激光超聲無損檢測技術(shù)的應(yīng)用范圍越來越廣闊。目前,激光超聲無損檢測的對象已不再局限于平面和圓筒圓柱等具有規(guī)則形狀的工件,在航空航天及核電等工業(yè)領(lǐng)域,也通常需要采用激光超聲無損檢測技術(shù)對一些具有自由曲面形狀的曲面工件進(jìn)行一系列的檢測工作。為了解決復(fù)雜曲面工件的激光超聲自動探測問題,國家質(zhì)檢總局開展了基于激光超聲技術(shù)的高溫曲面工件自動探測設(shè)備研制這一公益性科研專項。在激光超聲曲面自動探測系統(tǒng)研制背景下,為了保證激光超聲探測時需保證檢測激光沿被測物體表面法向方向掃查,以提高系統(tǒng)的檢測精度,本文采用了基于單目相機(jī)多輪廓遮擋技術(shù)的物體三維模型重構(gòu)方案來實現(xiàn)目標(biāo)曲面法矢量提取。本文建立了以CMOS相機(jī)、鏡頭、自動轉(zhuǎn)臺及計算機(jī)為基礎(chǔ)的圖像采集系統(tǒng),利用張正友標(biāo)定法及數(shù)學(xué)計算的形式對系統(tǒng)進(jìn)行內(nèi)外參標(biāo)定,并通過計算機(jī)完成目標(biāo)物體多角度圖像序列的自動采集;對采集到的圖像序列進(jìn)行二值化、濾波去噪、去小面積等處理,提取了圖像中目標(biāo)物體的輪廓數(shù)據(jù),并將多角度輪廓序列及其對應(yīng)的光心點坐標(biāo)數(shù)據(jù)進(jìn)行了坐標(biāo)轉(zhuǎn)換,實現(xiàn)了數(shù)據(jù)在世界坐標(biāo)系下的統(tǒng)一;與3D-Rhino軟件結(jié)合,根據(jù)體相交法原理獲取目標(biāo)物體近似三維模型;針對基于遮擋輪廓原理體相交法得到的三維模型大于目標(biāo)物體實際尺寸這一問題進(jìn)行了相關(guān)理論和實驗研究,得到了模型尺寸與光心點到物體的距離之間的線性變化規(guī)律,提出了基于模型尺寸相較于目標(biāo)物體實際尺寸的擴(kuò)大倍數(shù)反向縮小的方法將所得模型還原到目標(biāo)物體實際尺寸,并進(jìn)行了實驗驗證;在縮小后的目標(biāo)物體三維模型上通過點坐標(biāo)分析的方式提取了目標(biāo)曲面法矢量。同時通過兩個外形由簡單到復(fù)雜的物體作為實驗對象驗證了論文方案的可行性。最后描述了如何將所獲取的目標(biāo)物體法矢量用于激光超聲自動檢測系統(tǒng)進(jìn)行探頭掃描方位的控制。論文實現(xiàn)了三維重構(gòu)技術(shù)與激光超聲無損檢測技術(shù)的結(jié)合,法矢量的成功提取解決了激光超聲復(fù)雜曲面工件自動探測系統(tǒng)的探頭掃描方向的控制問題,由此為激光超聲曲面自動探測系統(tǒng)的研制提供了依據(jù)。
[Abstract]:With the rapid development of industry, laser ultrasonic nondestructive testing technology is more and more widely used. At present, the objects of laser ultrasonic nondestructive testing are no longer confined to regular workpieces, such as plane and cylinder cylinders, but also in the fields of aerospace, nuclear power and other industries. Laser ultrasonic nondestructive testing (NDT) is also usually used to detect some curved surfaces with free-form surfaces. In order to solve the problem of laser ultrasonic automatic detection of complex curved surface workpiece, the State Administration of quality Supervision, Quarantine and Quarantine has carried out the development of the high temperature surface workpiece automatic detection equipment based on laser ultrasonic technology, which is a public welfare scientific research project. Under the background of the development of laser ultrasonic surface automatic detection system, in order to ensure the laser scanning along the normal direction of the object surface, in order to improve the accuracy of the system, In this paper, the object 3D model reconstruction scheme based on the multi-contour occlusion technology of monocular camera is used to extract the normal vector of the target surface. In this paper, an image acquisition system based on CMOS camera, lens, automatic turntable and computer is established. The calibration method of Zhang Zhengyou and mathematical calculation are used to calibrate the internal and external parameters of the system. The multi-angle image sequence of the target object is automatically collected by computer. The contour data of the target object in the image are extracted by binary processing, filtering and denoising, de-noising, and the coordinate transformation of the multi-angle contour sequence and its corresponding optical center point coordinate data are carried out. The unification of data in the world coordinate system is realized. Combined with 3D-Rhino software, the approximate three-dimensional model of object is obtained according to the principle of volume intersection. Aiming at the problem that the 3D model based on the principle of shading contour is larger than the actual size of the object, the linear variation law between the model size and the distance from the center of light to the object is obtained. Based on the inverse reduction of the expansion multiple of the model size compared with the actual size of the target object, the model is restored to the actual size of the target object, and the experimental results are verified. The normal vector of the object surface is extracted by point coordinate analysis on the reduced object 3D model. At the same time, the feasibility of the scheme is verified by two objects with simple to complex shapes. At last, how to use the normal vector of the target object to control the scanning azimuth of the probe in the laser ultrasonic automatic detection system is described. The paper realizes the combination of 3D reconstruction technology and laser ultrasonic nondestructive testing technology. The normal vector is extracted successfully to solve the problem of scanning direction control of the probe in the automatic detecting system of laser ultrasonic complex curved surface workpiece. This provides the basis for the development of laser ultrasonic surface automatic detection system.
【學(xué)位授予單位】：南京航空航天大學(xué)
【學(xué)位級別】：碩士
【學(xué)位授予年份】：2017
【分類號】：TB559

【參考文獻(xiàn)】

相關(guān)期刊論文 前10條

1 孫廣開;周正干;陳曦;;激光超聲技術(shù)在先進(jìn)復(fù)合材料無損檢測中的應(yīng)用研究[J];失效分析與預(yù)防;2016年05期

2 曹建樹;曹振;趙龍飛;徐寶東;劉強(qiáng);姬保平;;激光超聲管道表面裂紋檢測技術(shù)[J];光電工程;2016年03期

3 田苗;郝向陽;劉松林;張小東;;基于點離散度的張正友標(biāo)定法[J];全球定位系統(tǒng);2015年06期

4 李濤濤;何宇;;最小二乘法在張正友標(biāo)定法中的應(yīng)用[J];信息與電腦(理論版);2015年23期

5 鄭德根;陳華斌;王繼鋒;林濤;陳善本;;基于激光超聲的鋁合金攪拌摩擦焊典型缺陷檢測及分析[J];焊接學(xué)報;2015年07期

6 馬官兵;袁書現(xiàn);徐清國;李明;王韋強(qiáng);;超聲仿真技術(shù)和激光超聲波可視化技術(shù)在核反應(yīng)堆壓力容器安全端窄間隙焊縫超聲波檢測能力驗證中的應(yīng)用[J];無損檢測;2015年06期

7 曾偉;王海濤;田貴云;胡國星;汪文;;研究激光激發(fā)的聲表面波與材料近表面缺陷的振蕩效應(yīng)[J];物理學(xué)報;2015年13期

8 楊依光;趙鋒;王飛;焦書浩;;管道壁厚及焊縫缺陷激光超聲檢測技術(shù)[J];油氣儲運;2015年07期

9 郭佳;李四海;寧寧;李閔行;張昭;王丹;;激光超聲技術(shù)在無損檢測中的應(yīng)用[J];航空工程進(jìn)展;2014年04期

10 王宇凡;崔省安;李瑞君;高兵;王建利;;超聲波在圓柱形工件中的波型變化[J];無損檢測;2014年07期

相關(guān)博士學(xué)位論文 前2條

1 趙艷;流—固界面和彎曲表面上激光超聲波的理論與實驗研究[D];南京理工大學(xué);2007年

2 吳思源;曲面工件超聲自動檢測中若干關(guān)鍵技術(shù)研究[D];浙江大學(xué);2006年

相關(guān)碩士學(xué)位論文 前6條

1 馮樹彪;基于圖像的三維重建[D];江南大學(xué);2012年

2 王方;利用數(shù)碼相機(jī)重構(gòu)物體三維形體的方法研究[D];西安理工大學(xué);2008年

3 程琳;激光超聲無損檢測技術(shù)研究[D];電子科技大學(xué);2007年

4 何麗;多視角背影輪廓3D測量方法的研究[D];天津大學(xué);2006年

5 李剛;激光超聲技術(shù)及其在金屬無損檢測方面的應(yīng)用[D];華中師范大學(xué);2004年

6 余峰;基于圖像的三維造型算法研究[D];浙江大學(xué);2004年

，

本文編號：2407040