本文選題：裂紋檢測(cè) + 分形�。� 參考：《長(zhǎng)安大學(xué)》2014年碩士論文

【摘要】：目前國(guó)內(nèi)外道路的養(yǎng)護(hù)與管理工作主要根據(jù)人工檢測(cè)和自動(dòng)檢測(cè)而來的數(shù)據(jù)信息來實(shí)施。因此國(guó)內(nèi)外也都投入巨大費(fèi)用用于研發(fā)各種路面自動(dòng)采集設(shè)備與數(shù)據(jù)處理系統(tǒng)，一般都是應(yīng)用多功能道路檢測(cè)車系統(tǒng)獲取路面破損圖像、車轍、平整度以及道路周圍的環(huán)境信息，后續(xù)對(duì)采集的數(shù)據(jù)進(jìn)行分析與識(shí)別處理，從而提供養(yǎng)護(hù)與維修報(bào)告。路面破損數(shù)據(jù)的信息處理由于受到路面自身的復(fù)雜性以及數(shù)據(jù)在采集過程中所遇到的外部環(huán)境影響，使得路面破損圖像的數(shù)據(jù)處理工作量巨大而花費(fèi)時(shí)間很長(zhǎng)。因此關(guān)于路面破損的圖像自動(dòng)識(shí)別系統(tǒng)研究意義重大。 本課題研究的內(nèi)容是對(duì)設(shè)備采集的路面破損圖像實(shí)現(xiàn)自動(dòng)識(shí)別，主要是對(duì)路面破損圖像中的各種裂紋進(jìn)行自動(dòng)識(shí)別。為了很好識(shí)別出破損圖像中的裂紋信息，本文提出了兩種分形的方法對(duì)路面破損圖像中的裂紋信息進(jìn)行識(shí)別，一種是分形的自相似性對(duì)路面破損圖像的識(shí)別，另外一種是基于分形理論中離散分?jǐn)?shù)布朗運(yùn)動(dòng)模型進(jìn)行識(shí)別。研究了如何把破損圖像中的標(biāo)志標(biāo)線和車轍等信息去除以及對(duì)破損路面圖像中裂紋信息的增強(qiáng)與圖像分割，最終實(shí)現(xiàn)了路面破損圖像中的裂紋識(shí)別的研究目的。 本論文主要研究成果建立的分形模型能很好的描述具有紋理的圖像；對(duì)采集的路面圖像進(jìn)行了預(yù)處理以達(dá)到改善圖像質(zhì)量的目的，為最后路面破損圖像裂紋的邊緣檢測(cè)與自動(dòng)識(shí)別做準(zhǔn)備；針對(duì)路面破損圖像的自相似和自仿射性的特點(diǎn)，，根據(jù)分形理論對(duì)路面破損圖像中裂紋進(jìn)行邊緣識(shí)別檢測(cè)。根據(jù)路面破損圖像的分形特性，提出了基于分形的自相似性和DFBR模型對(duì)路面破損圖像的裂紋識(shí)別，從而識(shí)別出路面破損圖像的裂紋信息。
[Abstract]:At present, the maintenance and management of roads both at home and abroad are carried out mainly according to the data and information from manual detection and automatic testing. Therefore, great costs are put into the research and development of various road automatic acquisition equipment and data processing systems at home and abroad. The smoothness and environmental information around the road, followed by analysis and recognition of the collected data, so as to provide maintenance and maintenance reports. The information processing of pavement damage data processing, due to the complexity of the road surface itself and the influence of the external environment of the data during the collection process, makes the data processing of the damaged image of the pavement. The workload is huge and takes a long time. Therefore, it is of great significance to study the automatic image recognition system for pavement distress.
The content of this study is to automatically recognize the damaged image of the pavement which is collected by the equipment. It is mainly to recognize all kinds of cracks in the damaged image. In order to identify the crack information in the damaged image, this paper proposes two fractal methods to identify the crack information in the damaged image of the pavement. The fractal self similarity is used to recognize the damaged image of the pavement, and the other is based on the discrete fractional Brown motion model in fractal theory. The information of marking and rutting in the damaged image and the enhancement of the crack information in the damaged road image and the image segmentation are studied, and the pavement damage is finally realized. The purpose of research on crack identification in images.
The fractal model of the main research results in this paper can describe the image with texture well, preprocessing the image of the collected pavement to improve the image quality, and prepare the edge detection and automatic recognition for the crack of the broken road surface, and the self similarity and self affine of the damaged road surface. On the basis of fractal theory, the fractal theory is used to detect the cracks in the damaged image of the pavement. According to the fractal characteristics of the damaged image of the pavement, a fractal based self similarity and DFBR model is proposed to identify the cracks in the damaged image of the pavement, thus identifying the crack information of the damaged image of the pavement.
【學(xué)位授予單位】：長(zhǎng)安大學(xué)
【學(xué)位級(jí)別】：碩士
【學(xué)位授予年份】：2014
【分類號(hào)】：U495;U416.0

【參考文獻(xiàn)】

相關(guān)期刊論文 前3條

1 高浩軍,杜宇人;中值濾波在圖像處理中的應(yīng)用[J];電子工程師;2004年08期

2 張洪光;王祁;;基于人工種群和Agent的路面裂紋檢測(cè)算法[J];哈爾濱工業(yè)大學(xué)學(xué)報(bào);2007年01期

3 王華;朱寧;王祁;;應(yīng)用計(jì)盒維數(shù)方法的路面裂縫圖像分割[J];哈爾濱工業(yè)大學(xué)學(xué)報(bào);2007年01期

本文編號(hào)：2081255