本文選題：遙感影像 + 分類(lèi)�。� 參考：《新疆大學(xué)》2017年碩士論文

【摘要】：高精度的遙感影像分類(lèi)是遙感技術(shù)應(yīng)用于其他領(lǐng)域的重要前提。伴隨著遙感技術(shù)的飛速發(fā)展,利用遙感技術(shù)對(duì)地物信息進(jìn)行分類(lèi)一直是遙感領(lǐng)域的研究熱點(diǎn)。近年來(lái),越來(lái)越多的學(xué)者將機(jī)器學(xué)習(xí)算法應(yīng)用于模式識(shí)別領(lǐng)域。其中,淺層機(jī)器學(xué)習(xí)算法在遙感影像分類(lèi)方面取得了一定的成果,但是由于淺層機(jī)器學(xué)習(xí)的學(xué)習(xí)能力有限且泛化能力比較差,其效率并不能很好的滿(mǎn)足海量遙感影像分類(lèi)的需求。深度學(xué)習(xí)的興起及迅速發(fā)展,為海量遙感影像分類(lèi)的研究開(kāi)辟了新的道路。分布式計(jì)算框架的提出為遙感影像分類(lèi)的高效并行提供了有效的解決方案。本文從淺層機(jī)器學(xué)習(xí)、深度學(xué)習(xí)和并行計(jì)算三個(gè)方面對(duì)遙感影像分類(lèi)進(jìn)行研究,主要研究?jī)?nèi)容有以下三方面:(1)構(gòu)建了一種基于BP神經(jīng)網(wǎng)絡(luò)的遙感影像棉花識(shí)別模型。為了簡(jiǎn)化使用遙感影像進(jìn)行棉花識(shí)別的過(guò)程,提高棉花識(shí)別的精度,本文提出了基于BP神經(jīng)網(wǎng)絡(luò)的棉花識(shí)別方法。從高分一號(hào)8米分辨率影像中提取歸一化植被指數(shù)、差值植被指數(shù)、比值植被指數(shù)等多個(gè)特征指數(shù),并將不同的特征指數(shù)進(jìn)行組合作為BP神經(jīng)網(wǎng)絡(luò)的輸入,對(duì)網(wǎng)絡(luò)進(jìn)行訓(xùn)練。實(shí)驗(yàn)表明:該模型能有效提高遙感影像棉花識(shí)別的精度。(2)構(gòu)建了一種基于卷積神經(jīng)網(wǎng)絡(luò)的遙感影像草地分類(lèi)模型。針對(duì)淺層機(jī)器學(xué)習(xí)在處理海量遙感數(shù)據(jù)時(shí)存在的局限性,本文將深度學(xué)習(xí)思想引入遙感影像分類(lèi)的研究之中,提出了基于卷積神經(jīng)網(wǎng)絡(luò)的遙感影像分類(lèi)模型。該模型可以直接將遙感圖像作為網(wǎng)絡(luò)的輸入,通過(guò)逐層的網(wǎng)絡(luò)訓(xùn)練,提取有用的特征用以實(shí)現(xiàn)遙感影像的自動(dòng)分類(lèi)。實(shí)驗(yàn)結(jié)果表明:迭代次數(shù)的增加及訓(xùn)練樣本的增加均可提高模型的分類(lèi)精度;基于卷積神經(jīng)網(wǎng)絡(luò)的遙感影像草地分類(lèi)模型的分類(lèi)精度為96.35%,比支持向量機(jī)(SVM)和BP神經(jīng)網(wǎng)絡(luò)方法分別提高8.90%和4.78%。(3)基于Spark分布式內(nèi)存計(jì)算平臺(tái)實(shí)現(xiàn)了遙感影像分類(lèi)的并行化。針對(duì)卷積神經(jīng)網(wǎng)絡(luò)模型在進(jìn)行遙感影像草地分類(lèi)過(guò)程中的高時(shí)耗問(wèn)題,本文將并行計(jì)算思想引入到遙感影像分類(lèi)的研究之中。借助Hadoop分布式文件系統(tǒng)(HDFS)對(duì)遙感影像進(jìn)行分布式存儲(chǔ),通過(guò)Spark分布式內(nèi)存計(jì)算平臺(tái)實(shí)現(xiàn)遙感影像分類(lèi)的高效并行。在分布式計(jì)算集群上采用數(shù)據(jù)并行方式實(shí)現(xiàn)卷積神經(jīng)網(wǎng)絡(luò)的并行化模型。通過(guò)三組實(shí)驗(yàn)驗(yàn)證了并行化卷積神經(jīng)網(wǎng)絡(luò)模型的性能。實(shí)驗(yàn)結(jié)果表明:并行化的卷積神經(jīng)網(wǎng)絡(luò)模型在保證遙感影像草地分類(lèi)精度的同時(shí),還提高了草地分類(lèi)的效率,實(shí)現(xiàn)了真正意義上的高效并行計(jì)算。
[Abstract]:High precision classification of remote sensing images is an important prerequisite for the application of remote sensing technology in other fields. With the rapid development of remote sensing technology, the classification of ground object information using remote sensing technology has been a hot spot in the field of remote sensing. In recent years, more and more scholars apply machine learning algorithm to pattern recognition. Among them, the shallow machine learning algorithm has made some achievements in remote sensing image classification, but the learning ability of shallow machine learning is limited and the generalization ability is relatively poor. Its efficiency can not meet the needs of massive remote sensing image classification. The rise and rapid development of deep learning has opened a new way for the classification of massive remote sensing images. The proposed distributed computing framework provides an effective solution for efficient parallel classification of remote sensing images. In this paper, the classification of remote sensing images is studied from three aspects: shallow machine learning, deep learning and parallel computing. The main research contents are as follows: 1) A cotton recognition model based on BP neural network is constructed. In order to simplify the process of cotton recognition using remote sensing image and improve the precision of cotton recognition, a cotton recognition method based on BP neural network is proposed in this paper. The normalized vegetation index, difference vegetation index and ratio vegetation index were extracted from the 8m resolution image of Gao Fen 1, and the different characteristic indexes were combined as the input of BP neural network to train the network. Experiments show that the model can effectively improve the precision of cotton recognition in remote sensing images. (2) A remote sensing image grassland classification model based on convolution neural network is constructed. Aiming at the limitation of shallow machine learning in processing massive remote sensing data, this paper introduces the idea of deep learning into the research of remote sensing image classification, and puts forward a classification model of remote sensing image based on convolution neural network. The model can directly take remote sensing images as the input of the network and extract useful features to realize the automatic classification of remote sensing images through the training of the network layer by layer. The experimental results show that the classification accuracy of the model can be improved by increasing the number of iterations and training samples. The classification accuracy of remote sensing image grassland classification model based on convolution neural network is 96.35, which is 8.90% and 4.78% higher than support vector machine (SVM) and BP neural network, respectively. The parallel classification of remote sensing image is realized based on Spark distributed memory computing platform. Aiming at the problem of high time consumption in remote sensing image grassland classification based on convolution neural network model, this paper introduces the idea of parallel computing into the research of remote sensing image classification. Remote sensing images are stored distributed by Hadoop distributed file system (HDFS), and the efficient parallel classification of remote sensing images is realized by Spark distributed memory computing platform. The parallel model of convolutional neural network is realized by data parallelism in distributed computing cluster. The performance of parallel convolution neural network model is verified by three groups of experiments. The experimental results show that the parallel convolution neural network model not only ensures the accuracy of grassland classification in remote sensing images, but also improves the efficiency of grassland classification.
【學(xué)位授予單位】：新疆大學(xué)
【學(xué)位級(jí)別】：碩士
【學(xué)位授予年份】：2017
【分類(lèi)號(hào)】：TP751

【參考文獻(xiàn)】

相關(guān)期刊論文 前10條

1 李彥冬;郝宗波;雷航;;卷積神經(jīng)網(wǎng)絡(luò)研究綜述[J];計(jì)算機(jī)應(yīng)用;2016年09期

2 滿(mǎn)衛(wèi)東;王宗明;劉明月;路春燕;賈明明;毛德華;任春穎;;1990-2013年?yáng)|北地區(qū)耕地時(shí)空變化遙感分析[J];農(nóng)業(yè)工程學(xué)報(bào);2016年07期

3 曹洪濤;張拯寧;李明;李器宇;陳浩;;基于C4.5決策樹(shù)的多特征遙感分類(lèi)方法[J];測(cè)繪工程;2016年03期

4 陳冠宇;安凱;李向;;基于卷積神經(jīng)網(wǎng)絡(luò)的不良地質(zhì)體識(shí)別與分類(lèi)[J];地質(zhì)科技情報(bào);2016年01期

5 王詔遠(yuǎn);王宏杰;邢煥來(lái);李天瑞;;基于Spark的蟻群優(yōu)化算法[J];計(jì)算機(jī)應(yīng)用;2015年10期

6 王知音;禹龍;田生偉;錢(qián)育蓉;丁建麗;楊柳;;基于棧式自編碼的水體提取方法[J];計(jì)算機(jī)應(yīng)用;2015年09期

7 劉煥軍;孟令華;張新樂(lè);Susan Ustin;寧東浩;孫思雨;;基于時(shí)間序列Landsat影像的棉花估產(chǎn)模型[J];農(nóng)業(yè)工程學(xué)報(bào);2015年17期

8 呂啟;竇勇;牛新;徐佳慶;夏飛;;基于DBN模型的遙感圖像分類(lèi)[J];計(jì)算機(jī)研究與發(fā)展;2014年09期

9 柴瑞敏;曹振基;;基于Gabor小波與深度信念網(wǎng)絡(luò)的人臉識(shí)別方法[J];計(jì)算機(jī)應(yīng)用;2014年09期

10 肖錦成;歐維新;符海月;;基于BP神經(jīng)網(wǎng)絡(luò)與ETM+遙感數(shù)據(jù)的鹽城濱海自然濕地覆被分類(lèi)[J];生態(tài)學(xué)報(bào);2013年23期

相關(guān)博士學(xué)位論文 前3條

1 張雁;基于機(jī)器學(xué)習(xí)的遙感圖像分類(lèi)研究[D];北京林業(yè)大學(xué);2014年

2 程付超;面向海量遙感數(shù)據(jù)的分布式混合計(jì)算技術(shù)研究[D];成都理工大學(xué);2014年

3 周海芳;遙感圖像并行處理算法的研究與應(yīng)用[D];國(guó)防科學(xué)技術(shù)大學(xué);2003年

相關(guān)碩士學(xué)位論文 前10條

1 楊志偉;基于Spark平臺(tái)推薦系統(tǒng)研究[D];中國(guó)科學(xué)技術(shù)大學(xué);2015年

2 孫科;基于Spark的機(jī)器學(xué)習(xí)應(yīng)用框架研究與實(shí)現(xiàn)[D];上海交通大學(xué);2015年

3 王恒歡;基于深度學(xué)習(xí)的圖像識(shí)別算法研究[D];北京郵電大學(xué);2015年

4 唐振坤;基于Spark的機(jī)器學(xué)習(xí)平臺(tái)設(shè)計(jì)與實(shí)現(xiàn)[D];廈門(mén)大學(xué);2014年

5 陳先昌;基于卷積神經(jīng)網(wǎng)絡(luò)的深度學(xué)習(xí)算法與應(yīng)用研究[D];浙江工商大學(xué);2014年

6 安源;遙感數(shù)字圖像處理技術(shù)在地質(zhì)填圖中的應(yīng)用研究[D];東北師范大學(xué);2013年

7 張博;高分辨率遙感影像多尺度分類(lèi)方法研究[D];電子科技大學(xué);2013年

8 邰建華;Hadoop平臺(tái)下的海量數(shù)據(jù)存儲(chǔ)技術(shù)研究[D];東北石油大學(xué);2012年

9 周婷婷;遙感影像輻射校正研究與應(yīng)用[D];福建師范大學(xué);2010年

10 黃曉云;基于HDFS的云存儲(chǔ)服務(wù)系統(tǒng)研究[D];大連海事大學(xué);2010年

，

本文編號(hào)：1938212