本文選題：空間插值 + 克里金插值算法��； 參考：《成都理工大學》2013年碩士論文

【摘要】：隨著社會經濟的不斷發(fā)展，科學技術的不斷進步，使得人類對空間環(huán)境的作用越來越強。一方面，人們對于資源地質信息預測和分析的精度要求越來越高，范圍要求越來越廣；另一方面，礦產預測與新興科學技術的聯(lián)系越來越緊密，并不斷有新方法、新技術的提出，特別是遙感技術的快速發(fā)展，傳遞給我們的信息越來越多，人們也越來越認識到信息技術對資源獲取所起到的重大作用。 另一方面，有時會因為多方面的原因以至于不能獲取完整有效的遙感數(shù)據(jù)。利用臨近的已知空間數(shù)據(jù)對未知空間數(shù)據(jù)值進行估計和推測，是解決缺省或無效空間數(shù)據(jù)非常有效的手段，即空間插值。 本文詳細介紹了常用空間插值算法，在克里金（kriging）插值算法和分形插值算法研究的基礎上，將以上兩種算法應用于適當調整之后的高光譜遙感數(shù)據(jù)。主要工作包括以下幾個方面： （1）簡要介紹了空間常用插值算法理論； （2）克里金插值算法的研究； （3）分形插值算法的研究； （4）利用克里金插值和分形插值算法對高光譜遙感影像數(shù)據(jù)缺省波段進行插值運算，并通過峰值信噪比等圖像信息參量對試驗結果進行驗證評估。 本文的主要創(chuàng)新點有： （1）因為克里金插值算法和分形插值算法雖然在地質領域運用很廣，但是運用于高光譜遙感影像光譜曲線插值還非常少，本文通過對高光譜遙感數(shù)據(jù)立方體進行空間坐標分解之后，運用克里金插值和分形插值對其假設缺省波段進行插值模擬； （2）在應用插值算法的時候，為了保證算法精度的無偏性和平穩(wěn)性，本文將波段數(shù)據(jù)中空值坐標區(qū)域全部換為波段均值，在插值模擬完成之后，，再將對應坐標位置的象元值替換為0，保證和原始數(shù)據(jù)空間結構一致。 論文最后給出了克里金插值和分形插值對文中高光譜數(shù)據(jù)插值的對比結果，包括克里金多種模型和多個波段的實驗結果。實驗結果表明，除極少數(shù)跳躍波段外，本文中所應用的克里金（kriging）插值和分形插值算法對高光譜影像數(shù)據(jù)具有較好的模擬效果，具有一定的應用價值，并對論文工作進行了總結，對空間插值算法進行了深入的分析與展望。
[Abstract]:With the development of social economy and the progress of science and technology, human beings play a more and more important role in space environment.On the one hand, the accuracy and scope of prediction and analysis of resource geological information are becoming more and more demanding; on the other hand, the relationship between mineral prediction and emerging science and technology is becoming closer and closer, and new methods are constantly available.With the development of new technology, especially the rapid development of remote sensing technology, more and more information has been transmitted to us, and more and more people realize that information technology plays an important role in resource acquisition.On the other hand, remote sensing data can not be obtained for many reasons.The estimation and estimation of unknown spatial data using adjacent known spatial data is a very effective method to solve the default or invalid spatial data, namely spatial interpolation.Based on the research of Kriging interpolation algorithm and fractal interpolation algorithm, the above two algorithms are applied to the hyperspectral remote sensing data after proper adjustment.The main tasks include the following:Firstly, the theory of spatial interpolation algorithm is introduced briefly.2) the research of Kriging interpolation algorithm;3) Fractal interpolation algorithm;(4) Kriging interpolation and fractal interpolation are used to interpolate the default band of hyperspectral remote sensing image data, and the experimental results are verified and evaluated by image information parameters such as peak signal-to-noise ratio (PSNR).The main innovations of this paper are:Because Kriging interpolation algorithm and fractal interpolation algorithm are widely used in geological field, but they are still very few in hyperspectral remote sensing image spectral curve interpolation.After the hyperspectral remote sensing data cube is decomposed into spatial coordinates, the Kriging interpolation and fractal interpolation are used to simulate the hypothetical default band of hyperspectral remote sensing data cube.In order to ensure the accuracy of the interpolation algorithm, in order to ensure the accuracy of the algorithm unbiased and stationary, this paper changes the band data hollow value coordinate region to the band mean value, and after the interpolation simulation is completed,Then the pixel value of the corresponding coordinate position is replaced with 0, which is consistent with the original data spatial structure.At the end of this paper, the comparison between Kriging interpolation and fractal interpolation for hyperspectral data interpolation is given, including the experimental results of several Kriging models and multiple bands.The experimental results show that, except for a few jump bands, the Kriging Kriging-based interpolation and fractal interpolation algorithms used in this paper have a good simulation effect on hyperspectral image data, and have certain application value, and the work of this paper is summarized.The spatial interpolation algorithm is analyzed and prospected.
【學位授予單位】：成都理工大學
【學位級別】：碩士
【學位授予年份】：2013
【分類號】：P237

【參考文獻】

相關期刊論文 前8條

1 朱求安,張萬昌,余鈞輝;基于GIS的空間插值方法研究[J];江西師范大學學報(自然科學版);2004年02期

2 曾懷恩;黃聲享;;基于Kriging方法的空間數(shù)據(jù)插值研究[J];測繪工程;2007年05期

3 馬軒龍;李春娥;陳全功;;基于GIS的氣象要素空間插值方法研究[J];草業(yè)科學;2008年11期

4 李新,程國棟,盧玲;空間內插方法比較[J];地球科學進展;2000年03期

5 王靖波,潘懋,張緒定;基于Kriging方法的空間散亂點插值[J];計算機輔助設計與圖形學學報;1999年06期

6 李德仁,李清泉;論地球空間信息技術與通信技術的集成[J];武漢大學學報(信息科學版);2001年01期

7 劉剛,胡遠來,賈s

本文編號：1735834