【摘要】：遙感影像的獲取與分類技術(shù)是遙感監(jiān)測(cè)過(guò)程中的基礎(chǔ)和關(guān)鍵。以往的航天航空遙感雖然能快速獲取大面積遙感影像,但是空間、時(shí)間分辨率較低。隨著當(dāng)前無(wú)人機(jī)和輕小型傳感器的發(fā)展,使高空間、時(shí)間分辨率的低空影像的獲取成為可能。當(dāng)前多光譜影像在應(yīng)用中易產(chǎn)生“同譜異物”、“同物異譜”等現(xiàn)象,更高的光譜分辨率也讓相鄰波段間的相關(guān)性大大增加,使得計(jì)算復(fù)雜度和時(shí)間復(fù)雜度大幅提高。因此對(duì)多(高)光譜遙感影像數(shù)據(jù)進(jìn)行降維處理是目前應(yīng)用研究中的一個(gè)難點(diǎn)問(wèn)題。本文面向低空多光譜地物分類,基于最佳波段指數(shù)和影像的光譜特征、紋理特征進(jìn)行最佳波段組合的選擇,然后利用支持向量機(jī)和最小二乘支持向量機(jī)分類方法構(gòu)建了多組分類模型進(jìn)行分類對(duì)比實(shí)驗(yàn)。主要工作及相關(guān)研究成果如下:(1)利用大型固定翼無(wú)人機(jī)搭載輕型多光譜相機(jī)搭建無(wú)人機(jī)遙感影像采集平臺(tái)并獲取了地面分辨率為22.6cm的12波段無(wú)人機(jī)多光譜遙感影像,再通過(guò)Pix4D Mapper對(duì)原始圖像進(jìn)行基于特征的配準(zhǔn)和特征級(jí)的融合得到研究區(qū)域的正射影像。(2)針對(duì)無(wú)人機(jī)多光譜影像數(shù)據(jù)的空間分辨率高、波段間相關(guān)性大等特點(diǎn),綜合影像的植被及水體指數(shù)等光譜信息,影像主成分分析和灰度共生矩陣計(jì)算得到的紋理特征信息及最佳波段指數(shù)法篩選的原始波段得到了最佳波段組合來(lái)進(jìn)行地物分類。(3)針對(duì)研究區(qū)初始得到的波段組合,設(shè)計(jì)監(jiān)督分類和非監(jiān)督分類的對(duì)比實(shí)驗(yàn)。相對(duì)原始波段組合,其中研究區(qū)域A的1,6,11,NDVI,NDWI,Mean波段組合的IsoData分類精度從83.57%提高到89.80%,SVM分類精度95.58%提高到99.76%。實(shí)驗(yàn)證明此波段組合不僅包含較多的波段信息且波段間相關(guān)性系數(shù)較低,同時(shí)反映了地物的光譜信息和紋理信息,可選擇其作為Micro MCA12 Snap的最佳波段組合。(4)針對(duì)實(shí)驗(yàn)得到的最佳波段組合,分別使用粒子群優(yōu)化和網(wǎng)格搜索算法進(jìn)行參數(shù)尋優(yōu)并使用交叉驗(yàn)證的方法對(duì)研究區(qū)域進(jìn)行SVM和LSSVM對(duì)比實(shí)驗(yàn)。實(shí)驗(yàn)結(jié)果表明,以粒子群優(yōu)化進(jìn)行參數(shù)尋優(yōu)得到的LSSVM分類模型,相對(duì)SVM粒子群優(yōu)化分類精度從97.833%提高到99.854%;相對(duì)LSSVM網(wǎng)格搜索分類精度從99.762%提高到99.854%。同時(shí)LSSVM粒子群優(yōu)化在一定程度上提高了分類的速度,是針對(duì)本文最佳波段組合在地物分類上的理想分類模型。
[Abstract]:The acquisition and classification of remote sensing images is the basis and key in the process of remote sensing monitoring. Although space and aviation remote sensing can obtain large area remote sensing image quickly, the spatial and temporal resolution is low. With the development of UAV and light sensor, it is possible to obtain low-altitude images with high spatial and temporal resolution. At present, multi-spectral images are easy to produce such phenomena as "isospectral foreign bodies" and "isospectral spectra". The higher spectral resolution also increases the correlation between adjacent bands, which greatly increases the computational complexity and time complexity. Therefore, dimensionality reduction of multispectral remote sensing image data is a difficult problem in application research. In this paper, we choose the best band combination based on the best spectral index and the spectral feature of the image, and the texture feature is used to select the best band combination for the low altitude multi-spectral feature classification. Then, support vector machine (SVM) and least squares support vector machine (LS-SVM) are used to construct multi-group classification models for classification comparison. The main work and related research results are as follows: (1) using a large fixed-wing UAV with a light multi-spectral camera to build a UAV remote sensing image acquisition platform and obtain 12-band UAV multi-spectral remote sensing image with ground resolution of 22.6cm. Then, the orthophoto image of the study area is obtained by the feature based registration and feature level fusion of the original image by Pix4D Mapper. (2) aiming at the characteristics of the UAV multi-spectral image data, such as high spatial resolution and large correlation between bands, etc. Spectral information such as vegetation and water body index of integrated image, The texture feature information obtained by principal component analysis and gray level co-occurrence matrix calculation and the original wave band selected by the best band index method are selected to obtain the best band combination for ground object classification. (3) for the initial band combination in the study area, Design the contrast experiment between supervised classification and unsupervised classification. Compared with the original band combination, the IsoData classification accuracy of the study area A is improved from 83.57% to 89.80% from 83.57% to 99.76%. The experimental results show that the band combination not only contains more band information, but also reflects the spectral information and texture information. It can be chosen as the best band combination of Micro MCA12 Snap. (4) for the best band combination obtained from the experiment, Particle swarm optimization (PSO) and mesh search algorithm are used to optimize the parameters, and the cross-validation method is used to carry out the SVM and LSSVM comparative experiments in the study area. The experimental results show that the classification accuracy of relative SVM particle swarm optimization is improved from 97.833% to 99.854, and that of relative LSSVM mesh search and classification is improved from 99.762% to 99.854. At the same time, LSSVM particle swarm optimization improves the speed of classification to a certain extent, which is an ideal classification model for the best band combination in this paper.
【學(xué)位授予單位】：石河子大學(xué)
【學(xué)位級(jí)別】：碩士
【學(xué)位授予年份】：2017
【分類號(hào)】：P237

【參考文獻(xiàn)】

相關(guān)期刊論文 前10條

1 楊龍;孫中宇;唐光良;林志文;陳燕喬;黎喻;李勇;;基于微型無(wú)人機(jī)遙感的亞熱帶林冠物種識(shí)別[J];熱帶地理;2016年05期

2 任哲;陳懷亮;王連喜;李穎;李琪;;利用交叉驗(yàn)證的小麥LAI反演模型研究[J];國(guó)土資源遙感;2015年04期

3 楊釗霞;鄒崢嶸;陶超;田彥平;何小飛;;空-譜信息與稀疏表示相結(jié)合的高光譜遙感影像分類[J];測(cè)繪學(xué)報(bào);2015年07期

4 高林;楊貴軍;王寶山;于海洋;徐波;馮海寬;;基于無(wú)人機(jī)遙感影像的大豆葉面積指數(shù)反演研究[J];中國(guó)生態(tài)農(nóng)業(yè)學(xué)報(bào);2015年07期

5 李曉東;姜琦剛;;基于多時(shí)相遙感數(shù)據(jù)的農(nóng)田分類提取[J];農(nóng)業(yè)工程學(xué)報(bào);2015年07期

6 王聰;杜華強(qiáng);周國(guó)模;徐小軍;孫少波;高國(guó)龍;;基于幾何光學(xué)模型的毛竹林郁閉度無(wú)人機(jī)遙感定量反演[J];應(yīng)用生態(tài)學(xué)報(bào);2015年05期

7 馮家莉;劉凱;朱遠(yuǎn)輝;李勇;柳林;蒙琳;;無(wú)人機(jī)遙感在紅樹(shù)林資源調(diào)查中的應(yīng)用[J];熱帶地理;2015年01期

8 程多祥;林家元;;寬河道線性基元無(wú)人機(jī)遙感影像鑲嵌配準(zhǔn)方法[J];測(cè)繪科學(xué);2014年12期

9 李宗南;陳仲新;王利民;劉佳;周清波;;基于小型無(wú)人機(jī)遙感的玉米倒伏面積提取[J];農(nóng)業(yè)工程學(xué)報(bào);2014年19期

10 汪沛;羅錫文;周志艷;臧英;胡煉;;基于微小型無(wú)人機(jī)的遙感信息獲取關(guān)鍵技術(shù)綜述[J];農(nóng)業(yè)工程學(xué)報(bào);2014年18期

相關(guān)博士學(xué)位論文 前2條

1 賈銀江;無(wú)人機(jī)遙感圖像拼接關(guān)鍵技術(shù)研究[D];東北農(nóng)業(yè)大學(xué);2016年

2 潘家志;基于光譜和多光譜數(shù)字圖像的作物與雜草識(shí)別方法研究[D];浙江大學(xué);2007年

相關(guān)碩士學(xué)位論文 前9條

1 裴松年;基于機(jī)器學(xué)習(xí)的分類算法研究[D];中北大學(xué);2016年

2 李毅;遺傳改進(jìn)粒子群優(yōu)化特征選擇的研究與應(yīng)用[D];云南大學(xué);2015年

3 丁雷龍;基于無(wú)人機(jī)影像的微山湖地區(qū)宜居性評(píng)價(jià)[D];中國(guó)地質(zhì)大學(xué)(北京);2015年

4 范永東;模型選擇中的交叉驗(yàn)證方法綜述[D];山西大學(xué);2013年

5 蔣韜;基于遺傳粒子群優(yōu)化算法的遙感圖像分類方法研究與應(yīng)用[D];首都師范大學(xué);2013年

6 徐秋輝;無(wú)控制點(diǎn)的無(wú)人機(jī)遙感影像幾何校正與拼接方法研究[D];南京大學(xué);2013年

7 何少林;基于無(wú)人機(jī)遙感影像的土地信息提取及專題圖制作研究[D];西南交通大學(xué);2013年

8 李靜;高光譜遙感影像降維及分類方法研究[D];中南大學(xué);2012年

9 項(xiàng)霞;基于最小二乘支持向量機(jī)的多光譜遙感影像分類[D];武漢大學(xué);2005年

，

本文編號(hào)：2192293