本文選題：移動(dòng)測(cè)量系統(tǒng) + 路面裂縫檢測(cè)��； 參考：《北京建筑大學(xué)》2017年碩士論文

【摘要】：隨著公路建設(shè)的日益完善,相關(guān)公路的檢查與養(yǎng)護(hù)工作已經(jīng)成為公路管理中的重要任務(wù)。針對(duì)當(dāng)前人工檢測(cè)道路裂縫方式存在耗時(shí)長(zhǎng)、成本高和具有危險(xiǎn)性等問(wèn)題,文章研究一種基于數(shù)字圖像處理的方法檢測(cè)道路裂縫,可以較為快速、準(zhǔn)確地識(shí)別路面裂縫信息。實(shí)驗(yàn)檢測(cè)數(shù)據(jù)來(lái)源于移動(dòng)測(cè)量系統(tǒng)(Mobile Mapping System,簡(jiǎn)稱(chēng)MMS)采集到的路面影像——以常規(guī)車(chē)速在公路行駛的同時(shí)利用車(chē)頂架設(shè)的攝像機(jī)采集公路路面的可量測(cè)立體影像信息。因此,重點(diǎn)就基于MMS圖像的路面裂縫檢測(cè)與分類(lèi)算法進(jìn)行研究。由于MMS圖像存在硬件設(shè)備、光照不均和路面陰影等因素的影響,增加了圖像處理的復(fù)雜度。通過(guò)對(duì)比分析確定較為理想的圖像預(yù)處理步驟與方法:具體為采用加權(quán)平均值法進(jìn)行圖像灰度化;以降低計(jì)算量為目的的改進(jìn)型雙邊濾波圖像去噪算法;基于圖像飽和度和光照比例因子去除圖像陰影干擾;以及保證裂縫細(xì)節(jié)信息不被破壞的粗糙集理論算法增強(qiáng)圖像。其次,分析目前兩類(lèi)經(jīng)典的圖像分割算法。針對(duì)單一型圖像分割算法在檢測(cè)裂縫中均存在一定的局限性,研究采用一種融合閾值法與數(shù)學(xué)形態(tài)學(xué)法的圖像分割方法:即針對(duì)每幅MMS圖像特點(diǎn)利用最大類(lèi)間方差求出各圖像的自適應(yīng)閾值,并結(jié)合以多種方向的結(jié)構(gòu)元素為基礎(chǔ)的數(shù)學(xué)形態(tài)學(xué)算法實(shí)現(xiàn)道路裂縫的分割。此外,采用基于形態(tài)學(xué)生長(zhǎng)的算法對(duì)分割結(jié)果中斷裂的裂縫進(jìn)行邊緣連接,填補(bǔ)裂縫中未被提取到的灰度較低區(qū)域。最后,在路面裂縫檢測(cè)結(jié)果的基礎(chǔ)上,對(duì)路面裂縫特征進(jìn)行描述和類(lèi)別界定,并根據(jù)相關(guān)標(biāo)準(zhǔn)對(duì)路面損壞狀況進(jìn)行分析與評(píng)價(jià)。根據(jù)研究和實(shí)驗(yàn)驗(yàn)證,針對(duì)MMS圖像檢測(cè)道路裂縫的方法在保證檢測(cè)準(zhǔn)確性的同時(shí),也具有較高的路面裂縫檢測(cè)速度(通過(guò)優(yōu)化算法),有助于快速有效地發(fā)現(xiàn)公路路面裂縫,對(duì)進(jìn)一步實(shí)現(xiàn)道路裂縫檢測(cè)自動(dòng)化和實(shí)時(shí)化有著積極的推動(dòng)作用,提升公路養(yǎng)護(hù)工作的智能化水平。
[Abstract]:With the improvement of highway construction, the inspection and maintenance of highway has become an important task in highway management. Aiming at the problems of long time, high cost and dangerous in manual detection of road cracks, this paper studies a method based on digital image processing to detect road cracks, which can identify the information of road cracks more quickly and accurately. The experimental data come from the road image collected by the Mobile Mapping system (MMS), which uses the camera mounted on the top of the vehicle to collect the stereo image of the road surface with the normal speed while driving on the road. Therefore, the algorithm of pavement crack detection and classification based on MMS image is studied. The complexity of MMS image processing is increased because of the influence of hardware equipment, uneven illumination and road surface shadow. Through comparative analysis, the ideal image preprocessing steps and methods are determined: the weighted average method is used to grayscale the image, and the improved bilateral filtering image denoising algorithm is designed to reduce the computation cost. Based on image saturation and illumination ratio factor, shadow interference is removed, and rough set theory algorithm is used to enhance the image, which ensures that the crack details are not destroyed. Secondly, two classical image segmentation algorithms are analyzed. Aiming at the limitation of single image segmentation algorithm in detecting cracks, In this paper, an image segmentation method based on fusion threshold method and mathematical morphology method is proposed. According to the characteristics of each MMS image, the adaptive threshold of each image is obtained by using the maximum inter-class variance. Combined with the mathematical morphology algorithm based on the structural elements in many directions, the road crack segmentation is realized. In addition, an algorithm based on morphological growth is used to connect the broken cracks in the segmentation results to fill the lower gray level areas that have not been extracted from the cracks. Finally, based on the detection results of pavement cracks, the characteristics of pavement cracks are described and classified, and the pavement damage status is analyzed and evaluated according to the relevant standards. According to the research and experimental verification, the method of road crack detection based on MMS image not only ensures the accuracy of the detection, but also has a higher detection speed (through the optimization algorithm), which is helpful to quickly and effectively find the road surface crack. It plays an active role in realizing the automation and realtime of road crack detection and improving the intelligent level of highway maintenance.
【學(xué)位授予單位】：北京建筑大學(xué)
【學(xué)位級(jí)別】：碩士
【學(xué)位授予年份】：2017
【分類(lèi)號(hào)】：U418.6;TP391.41

【參考文獻(xiàn)】

相關(guān)期刊論文 前10條

1 邱立英;;一種改進(jìn)的邊界掃描的路面裂縫檢測(cè)算法[J];鹽城工學(xué)院學(xué)報(bào)(自然科學(xué)版);2015年03期

2 王德方;曾衛(wèi)明;王倪傳;;基于改進(jìn)K-means算法的不均勻光照下道路裂縫檢測(cè)[J];計(jì)算機(jī)應(yīng)用與軟件;2015年07期

3 王見(jiàn)紅;;淺談移動(dòng)測(cè)量系統(tǒng)未來(lái)發(fā)展[J];北京測(cè)繪;2014年06期

4 許艷;;圖像的頻域?yàn)V波增強(qiáng)研究與仿真[J];電子技術(shù);2013年12期

5 徐威;唐振民;呂建勇;;基于圖像顯著性的路面裂縫檢測(cè)[J];中國(guó)圖象圖形學(xué)報(bào);2013年01期

6 王平讓;黃宏偉;薛亞?wèn)|;;基于圖像局部網(wǎng)格特征的隧道襯砌裂縫自動(dòng)識(shí)別[J];巖石力學(xué)與工程學(xué)報(bào);2012年05期

7 朱平哲;黎蔚;;基于主動(dòng)生長(zhǎng)的斷裂裂縫塊的連接方法[J];計(jì)算機(jī)應(yīng)用;2011年12期

8 馬常霞;趙春霞;狄峰;李e，

本文編號(hào)：2060299