本文關(guān)鍵詞： 地理國情普查 地表覆蓋 影像分割 面向?qū)ο蠓诸?高分辨率遙感影像　出處：《遼寧工程技術(shù)大學(xué)》2013年碩士論文　論文類型：學(xué)位論文

【摘要】：地理國情普查是獲取國情國力信息的重要手段,是掌握自然資源、生態(tài)環(huán)境以及人類活動基本情況的綜合性、基礎(chǔ)性工作。本文從地理國情普查的應(yīng)用角度出發(fā),重點研究面向地理國情普查的高分辨率遙感影像的解譯技術(shù)和方法,制定地表覆蓋解譯流程,研制面向地理國情普查的地表覆蓋解譯軟件,并開展相關(guān)的實驗。主要包括以下工作： (1)引入了MPI技術(shù),研究了影像的并行多尺度分割方法,16個進程下的運行速度是單個進程的8倍,提高了影像分割的效率；研究DLG數(shù)據(jù)輔助的分割方法,對分割邊界進行約束。 (2)探索出“分層逐級分類”的分類策略,根據(jù)遙感影像和地表覆蓋地物要素的復(fù)雜性,綜合考慮影像光譜、形狀和紋理特征,構(gòu)建分層的分類樹,按照“先易后難”的原則,實現(xiàn)基于決策樹和SVM的面向?qū)ο蠓诸悺?(3)針對地理國情普查的特點和當前影像解譯精度不高、后編輯工作量大等問題,制定面向地理國情普查的地表覆蓋解譯技術(shù)流程,完成了臨潼區(qū)的WorldView-2的影像的解譯,解譯精度達到了81%以上,驗證了地表覆蓋解譯流程的可行性。
[Abstract]:The general survey of geographical conditions is an important means to obtain the information of national conditions and national strength. It is also a comprehensive way to master the natural resources, ecological environment and the basic situation of human activities. This paper focuses on the interpretation technology and method of high-resolution remote sensing image for geographical situation census, and formulates the interpretation process of ground cover. To develop a ground cover interpretation software for GIS census, and to carry out related experiments. The main work includes the following: 1) the parallel multi-scale image segmentation method is studied by introducing MPI technology. The speed of 16 processes is 8 times that of a single process, and the efficiency of image segmentation is improved. The DLG data aided segmentation method is studied, and the segmentation boundary is constrained. According to the complexity of the elements of remote sensing image and ground cover, and considering the spectral, shape and texture features of the image, the hierarchical classification tree is constructed. 2) the classification strategy of "hierarchical classification" is explored, and the hierarchical classification tree is constructed according to the complexity of the elements of remote sensing image and ground cover. According to the principle of "easy before hard", the object-oriented classification based on decision tree and SVM is realized. 3) aiming at the problems such as the characteristics of the general survey of geographical conditions, the low precision of image interpretation at present, and the heavy workload of post-editing, the technical flow of surface coverage interpretation for the survey of geographical conditions is formulated. The interpretation of WorldView-2 images in Lintong District is completed, and the interpretation accuracy is more than 81%, which verifies the feasibility of the interpretation process of ground cover.
【學(xué)位授予單位】：遼寧工程技術(shù)大學(xué)
【學(xué)位級別】：碩士
【學(xué)位授予年份】：2013
【分類號】：P237

【參考文獻】

相關(guān)期刊論文 前10條

1 王惠英;蘇偉;周軍其;;基于多分類器組合面向?qū)ο筮b感影像信息提取方法[J];測繪工程;2009年05期

2 駱劍承,王欽敏,馬江洪,周成虎,梁怡;遙感圖像最大似然分類方法的EM改進算法[J];測繪學(xué)報;2002年03期

3 童磊;鄒崢嶸;;基于高分辨率衛(wèi)星影像的城市用地信息提取研究[J];測繪與空間地理信息;2009年02期

4 馬子睿;;基于數(shù)學(xué)形態(tài)學(xué)的醫(yī)學(xué)圖像分割方法研究[J];電腦與信息技術(shù);2009年02期

5 王艷萍;姜紀沂;林玲玲;;高分辨率遙感影像中道路震害信息的識別方法[J];計算機工程與應(yīng)用;2012年03期

6 張芳;李鑫;陳佳楠;;衛(wèi)星遙感影像目視解譯在森林資源調(diào)查中的應(yīng)用分析[J];內(nèi)蒙古林業(yè)調(diào)查設(shè)計;2009年04期

7 譚國強;王周龍;王霞;馬金衛(wèi);;基于知識的遙感影像居民地信息提取——以山東省蓬萊市為例[J];山東國土資源;2011年04期

8 駱劍承,周成虎,楊艷;具有部分監(jiān)督的遙感影像模糊聚類方法研究及應(yīng)用[J];遙感技術(shù)與應(yīng)用;1999年04期

9 杜鳳蘭,田慶久,夏學(xué)齊;遙感圖像分類方法評析與展望[J];遙感技術(shù)與應(yīng)用;2004年06期

10 周春艷;王萍;張振勇;齊成濤;;基于面向?qū)ο笮畔⑻崛〖夹g(shù)的城市用地分類[J];遙感技術(shù)與應(yīng)用;2008年01期

相關(guān)博士學(xué)位論文 前2條

1 高偉;基于特征知識庫的遙感信息提取技術(shù)研究[D];中國地質(zhì)大學(xué);2010年

2 劉煒;土地利用/覆被變化信息遙感圖像自動分類識別與提取方法研究[D];西北農(nóng)林科技大學(xué);2012年

，

本文編號：1464009