【摘要】：隨著空間信息觀測技術(shù)的提高,人們已經(jīng)可以利用各種不同的觀測方式獲取地表空間信息。但是在觀測過程中,由于觀測方式本身的限制、觀測環(huán)境的影響和觀測平臺故障等多方面因素的影響,大量的空間觀測數(shù)據(jù)中都存在空間不連續(xù)的現(xiàn)象。這種“空間縫隙”為空間觀測數(shù)據(jù)的后續(xù)使用帶來了嚴重影響。因此,如何消除觀測數(shù)據(jù)中的無效信息,獲得空間無縫的空間觀測數(shù)據(jù)是一個具有重要意義的研究課題。 本文從時間信息互補、光譜信息互補、空間信息互補多個角度出發(fā),研究了： (1)基于譜段互補信息的遙感數(shù)據(jù)融合重建方法 針對高光譜或多光譜遙感影像中由與某一譜段傳感器故障或噪聲等因素引起的觀測信息空間不連續(xù)問題,利用影像數(shù)據(jù)中的光譜冗余信息,尋找多波段數(shù)據(jù)間的相似性,在此基礎(chǔ)上建立多波段數(shù)據(jù)間的關(guān)系,消除離群點現(xiàn)象的影響對空間不連續(xù)區(qū)域進行重建。 (2)基于時相互補信息的遙感數(shù)據(jù)融合重建方法 針對遙感影像中由觀測環(huán)境和傳感器故障等因素引起的觀測信息空間不連續(xù)問題,在對多時相數(shù)據(jù)的差異性進行分析的基礎(chǔ)上,建立多時相數(shù)據(jù)相似信息提取方法,克服復雜場景變化和離群點等因素帶來的負面影響,結(jié)合時域補充信息對空間不連續(xù)區(qū)域進行融合。 (3)基于空間互補信息的點-面融合方法 由于遙感數(shù)據(jù)觀測范圍廣但反演精度受限于多種因素,傳統(tǒng)的地表觀測數(shù)據(jù)精度雖高但觀測點往往過于離散難。針對此問題,本文擬通過分析遙感成像過程中各因素的相互作用,建立站點-遙感數(shù)據(jù)之間的關(guān)聯(lián)模型,研究基于地統(tǒng)計學的站點-遙感數(shù)據(jù)融合方法,并將其應用于京津冀地區(qū)大氣污染物監(jiān)測。
[Abstract]:With the improvement of spatial information observation technology, people can obtain surface spatial information by various observation methods. However, in the observation process, because of the limitation of the observation mode itself, the influence of the observation environment and the fault of the observation platform, there is spatial discontinuity in a large number of spatial observation data. This space gap has a serious impact on the subsequent use of space observation data. Therefore, how to eliminate the invalid information in the observation data and obtain the spatial observation data seamlessly is an important research topic. This paper starts from the complementary of time information, spectral information and spatial information. In this paper: (1) remote sensing data fusion and reconstruction based on spectral complementary information is applied to the observation of hyperspectral or multispectral remote sensing images caused by sensor fault or noise in a certain spectral segment. Discontinuity of information space, The spectral redundancy information in image data is used to search for the similarity between multi-band data. Based on this, the relationship between multi-band data is established, and the spatial discontinuous region is reconstructed by eliminating the influence of outliers. (2) the method of remote sensing data fusion reconstruction based on time-complementary information is aimed at the spatial discontinuity of observation information caused by observation environment and sensor fault in remote sensing image. Based on the analysis of the difference of multitemporal data, a method of extracting similar information from multitemporal data is established to overcome the negative effects of complex scene changes and outliers, etc. The spatial discontinuous region is fused with time domain supplementary information. (3) the point-surface fusion method based on spatial complementary information is difficult to discrete because of the wide observation range of remote sensing data and limited inversion accuracy due to many factors. In order to solve this problem, through analyzing the interaction of various factors in the process of remote sensing imaging, this paper proposes to establish the correlation model of site-remote sensing data, and to study the method of site-remote sensing data fusion based on geostatistics. It is applied to the monitoring of air pollutants in Beijing, Tianjin and Hebei.
【學位授予單位】：武漢大學
【學位級別】：博士
【學位授予年份】：2014
【分類號】：TP751

【參考文獻】

相關(guān)期刊論文 前2條

1 祿豐年;;多源遙感影像配準技術(shù)分析[J];測繪科學技術(shù)學報;2007年04期

2 楊存武;掃描線校正器的光學原理及分析[J];紅外研究(A輯);1985年01期

相關(guān)碩士學位論文 前1條

1 趙輝;基于點特征的圖像配準算法研究[D];山東大學;2006年

，

本文編號：2256417