基于激光點(diǎn)云和近景影像的數(shù)字拓片生成方法
發(fā)布時(shí)間:2018-08-20 13:32
【摘要】:傳統(tǒng)的拓片制作過(guò)程復(fù)雜,天氣、光照、環(huán)境等客觀條件都會(huì)影響拓片質(zhì)量,而且手工拓片會(huì)對(duì)文物本體造成傷害。隨著信息化技術(shù)的發(fā)展,提出了利用數(shù)碼影像制作數(shù)字拓片。由于一般的數(shù)碼影像是中心投影且沒(méi)有可量測(cè)性,生成的數(shù)字拓片也沒(méi)有實(shí)際的應(yīng)用價(jià)值,且基于多張影像的數(shù)字拓片還存在圖像投影、配準(zhǔn)等諸多問(wèn)題。針對(duì)上述情況,本文提出了利用三維激光點(diǎn)云和近景影像生成數(shù)字拓片的解決方案:首先利用點(diǎn)云和影像配準(zhǔn)生成正射影像,然后采用圖像增強(qiáng)的方法消除光照不均,再將圖像局部自適應(yīng)閾值二值化,最后采用圖像融合的方法將邊緣模糊,生成有傳統(tǒng)拓片效果的數(shù)字拓片。通過(guò)實(shí)驗(yàn)證明,本文制作數(shù)字拓片的方法具有一定的可行性,且制作的文物拓片具有可量測(cè)的優(yōu)越性,為文物數(shù)字化存檔和研究提供了更豐富的數(shù)據(jù)。
[Abstract]:The traditional process of making rubbings is complex, such as weather, light, environment and other objective conditions will affect the quality of rubbings, and manual rubbings will cause damage to the body of cultural relics. With the development of information technology, a digital topology using digital image is proposed. Because the general digital image is a central projection and has no measurability, the generated digital topology has no practical application value, and the digital topology based on multiple images still has many problems, such as image projection, registration and so on. In view of the above situation, this paper proposes a solution to generate digital topology using 3D laser point cloud and close-range image: firstly, point cloud and image registration are used to generate orthophoto image, and then image enhancement method is used to eliminate uneven illumination. Finally, the image edge is blurred by image fusion method, and the digital topology with traditional topology effect is generated. It is proved by experiments that the method of making digital rubbings in this paper is feasible, and that the rubbings made have the advantages of being measurable, which provides more abundant data for the digitized archiving and research of cultural relics.
【作者單位】: 北京建筑大學(xué)測(cè)繪與城市空間信息學(xué)院;北京建筑遺產(chǎn)精細(xì)重構(gòu)與健康監(jiān)測(cè)重點(diǎn)實(shí)驗(yàn)室;山東正元數(shù)字城市建設(shè)有限公司;測(cè)繪出版社;
【基金】:國(guó)家自然科學(xué)基金(41401536);國(guó)家自然科學(xué)基金(41501495);國(guó)家自然科學(xué)基金(41601409) 北京市教委科技面上項(xiàng)目(KM201510016007) 北京市委組織部?jī)?yōu)秀人才項(xiàng)目D類(2013D005017000002)
【分類號(hào)】:K854;TP391.41
本文編號(hào):2193796
[Abstract]:The traditional process of making rubbings is complex, such as weather, light, environment and other objective conditions will affect the quality of rubbings, and manual rubbings will cause damage to the body of cultural relics. With the development of information technology, a digital topology using digital image is proposed. Because the general digital image is a central projection and has no measurability, the generated digital topology has no practical application value, and the digital topology based on multiple images still has many problems, such as image projection, registration and so on. In view of the above situation, this paper proposes a solution to generate digital topology using 3D laser point cloud and close-range image: firstly, point cloud and image registration are used to generate orthophoto image, and then image enhancement method is used to eliminate uneven illumination. Finally, the image edge is blurred by image fusion method, and the digital topology with traditional topology effect is generated. It is proved by experiments that the method of making digital rubbings in this paper is feasible, and that the rubbings made have the advantages of being measurable, which provides more abundant data for the digitized archiving and research of cultural relics.
【作者單位】: 北京建筑大學(xué)測(cè)繪與城市空間信息學(xué)院;北京建筑遺產(chǎn)精細(xì)重構(gòu)與健康監(jiān)測(cè)重點(diǎn)實(shí)驗(yàn)室;山東正元數(shù)字城市建設(shè)有限公司;測(cè)繪出版社;
【基金】:國(guó)家自然科學(xué)基金(41401536);國(guó)家自然科學(xué)基金(41501495);國(guó)家自然科學(xué)基金(41601409) 北京市教委科技面上項(xiàng)目(KM201510016007) 北京市委組織部?jī)?yōu)秀人才項(xiàng)目D類(2013D005017000002)
【分類號(hào)】:K854;TP391.41
【相似文獻(xiàn)】
相關(guān)碩士學(xué)位論文 前1條
1 劉亞楠;點(diǎn)云處理方法研究及在文物虛擬復(fù)原中的應(yīng)用[D];西北大學(xué);2015年
,本文編號(hào):2193796
本文鏈接:http://sikaile.net/shekelunwen/kgx/2193796.html
最近更新
教材專著