共查询到15条相似文献，搜索用时 140 毫秒
1.
高分六号宽幅多光谱数据人工林类型分类 总被引：1，自引：0，他引：1
高分六号(GF-6)卫星于2018年成功发射,2019-03正式投入使用。由于GF-6宽幅相机的WFV(Wide Field of View)影像较GF-1的同类影像新增2个红边波段,将会提高对农业、林业、草原等资源监测能力。为了分析GF-6的WFV影像在人工林分类方面的能力,促进高分数据在林业领域的应用,本文选取广西高峰林场为研究区,以最新的GF-6 WFV影像为数据源,结合地面实测类型数据,进行广西南宁高峰林场的桉树,杉木等人工林类型提取。主要运用随机森林(random forests)的分层分类法:首先计算6种植被指数,并利用随机森林法进行植被指数的特征优选,然后确定4种波段组合数据集的分类方案,4种数据集分别为(1)无红边的前4个波段,(2)有红边的8个波段,(3) 8个波段加上未优化的植被指数特征组合,(4) 8个波段加上优化的植被指数特征组合。再进行WFV影像4种数据集的随机森林分类,随机森林采用分类回归树(CART)算法来生成分类树,结合了bagging和随机选择特征变量的优点,是一种有效的分类方法。最后比较4个方案的分类结果并进行精度验证。结果表明:方案2比方案1精度提高了4.99%,Kappa系数提高了0.058。说明包含红边的8波段数据比4个波段数据精度有大幅提升。方案4的8波段加上优化植被指数特征组合的分类精度最高,达到了85.38%,比方案2包含红边波段组和方案1无红边波段组的精度分别提高了3.98%,8.97%,Kappa系数分别提高了0.046,0.104。说明WFV影像加入红边波段比无红边波段精度明显增高。由结果可知,红边指数的引入,增强了植被信息,能够较准确地反映人工林类型特征差异,明显提升了人工林的分类精度。本研究方法可以有效改善广西人工林类型信息提取效果,为GF-6影像质量的评价及其在林业应用潜力提供科学参考依据。 相似文献
2.
面向对象与卷积神经网络模型的GF-6 WFV影像作物分类 总被引：1，自引：0，他引：1
GF-6 WFV影像是中国首颗带有红边波段的中高分辨率8波段多光谱卫星的遥感影像,对于其影像及红边波段对作物分类影响的研究利用亟待展开。本文结合面向对象和深度学习提出一种适用于GF-6 WFV红边波段的卷积神经网络(RE-CNN)遥感影像作物分类方法。首先采用多尺度分割和ESP工具选择最佳分割参数完成影像分割,通过面向对象的CART决策树消除椒盐现象的同时提取植被区域,并转化为卷积神经网络的输入数据,最后基于Python和Numpy库构建的卷积神经网络模型(RE-CNN)用于影像作物分类及精度验证。有无红边波段的两组分类实验结果表明:在红边波段组,卷积神经网络(RE-CNN)作物分类识别取得了较好的效果,总体精度高达94.38%,相比无红边波段组分类精度提高了2.83%,验证了GF-6 WFV红边波段对作物分类的有效性。为GF-6 WFV红边波段影像用于作物的分类研究提供技术参考和借鉴价值。 相似文献
3.
以山东省为研究区域,利用2009年9月MODIS的8 d合成波段反射率产品MOD09,选择特征变量植被指数(NDVI、EVI)、NDWI、NDMI、NDSI及辅助信息DEM,通过选取其中的影像特征组合来确定分类方案,构建各波段组合的CART决策树,对MODIS影像进行分类,得到CART决策树的最优波段组合。结果表明,特征变量DEM、NDVI、EVI对分类结果贡献较大;将CART决策树的分类结果与其相对应的最大似然分类结果进行比较可知,基于影像多特征的CART决策树分类方法能明显提高分类精度。 相似文献
4.
为研究我国首颗携带红边波段的高分六影像（GF-6）在林地与非林地上的识别贡献，本文选择复杂林地类型的安徽省黄山市作为研究区，采用特征优选（RFE）与随机森林（RF）相结合的方法开展了林地与非林地识别潜力研究。首先根据实地调查、Google Earth影像及林地"一张图"样本数据构建了样本库；然后基于DEM、多时相光谱特征、植被指数、红边指数等特征开展分类，并比较不同模型精度及不同变量的重要度。结果表明：GF-6红边信息对林地非林地识别较为重要，引入红边信息可将总体分类精度提升2%，其他新增波段及地形特征对林地与非林地识别贡献并不明显；多时相数据的运用相比单时相数据可整体提高林地类型的分类精度2.93%~4.1%，单时相分类结果6月最好，9月次之，12月最差；特征优选可以有效减少数据输入维数（46到15），并取得最高分类精度，在不牺牲精度的同时保证了运算数据量的减少且明确了不同变量的贡献，具有较强的应用意义。 相似文献
5.
针对信息冗余导致机器学习分类精度下降的问题,该文提出了一种结合Relief F和递归特征消除法(RFE)进行特征优选的桉树人工林面积提取方法。首先将Zhuhai-1和Sentinel-1卫星影像进行融合,通过结合Relief F和RFE进行特征优选,以减少信息冗余的影响。其次,基于光谱、红边指数、纹理特征、植被指数和后散射系数等特征,采用简单非迭代聚类(SNIC)面向对象和随机森林(RF)算法进行桉树人工林面积提取。最后,为验证Relief F-RFE模型在分类方面的性能表现,将其与无特征优选下基于Zhuhai-1和融合数据以及基于RF特征优选的分类结果进行对比分析。结果表明：利用Relief F-RFE优选特征能有效提高桉树人工林分类精度,总体精度达到96.43%,相比于无特征优选下基于Zhuhai-1和融合数据分类结果,总体精度分别提高14.95%和8.43%。在与RF特征选择方法进行对比时,总体精度有所增长,提高了7.55%。 相似文献
6.
高分六号红边特征的农作物识别与评估 总被引：3，自引：0，他引：3
红边作为植被敏感波段，其红边特征的运用是遥感识别农作物并实现精准农业的高新手段之一。以黑龙江松嫩平原北部为研究区，以国内首个提供红边波段的多光谱高分六号影像和玉米、大豆、水稻总计82859个作物样本同时作为研究对象，从以下几个方面研究了红边波段和红边指数波段等红边特征在农作物识别中的表现，并评估了农作物的识别精度。（1） 通过作物样本辐射亮度值的统计特征，初步显示了在两红边波段0.710 μm和0.750 μm处有比其他波段更好的区分；（2） 根据传统归一化植被指数形式构建了红边归一化植被指数NDVI710和NDVI750，综合两指数在J-M距离表征的作物样本类别区分度上比传统NDVI更显著；（3） 通过多种手段筛选了有效波段并且制定了支持向量机（SVM）框架下4种农作物识别的分类策略，分别在5∶5、6∶4、7∶3、8∶2、9∶1等5套随机样本分割方案下完成研究区域农作物的分类预测。在这20类分类精度中kappa系数均高于0.9609，总体精度高于0.9742；列向上5∶5分割方案的精度最高，8∶2的精度最低；横向上分类精度排序如下：SVM-RFE > SVM-RF > SVM-有红边波段 > SVM-无红边波段，该结果表明了红边指数和红边波段的参与显著地提高了作物的识别精度；（4） 由于水域等其他样本的缺少，SVM-RFE方法和SVM-RF方法的分类图像均存在少量错分现象。但从分类精度和图像细节展示上来看，SVM-RFE方法要优于SVM-RF方法，二者分类图像的交叉验证中kappa系数为0.8060，总体精度为0.8743。总之，高分六号红边特征在作物识别中表现优越，使得识别精度显著提高。后续研究者可开发更多与红边相关的植被指数，充分发挥红边特征在精准农业中的作用。 相似文献
7.
《北京测绘》2019,(12)
为研究辽东湾海冰类型特征,文中基于2016年2月9日的Landsat-8和Sentinel-1A数据,结合多光谱、归一化植被指数(NDVI)构建CART自动决策树(CART Automatic Decision Tree),并用灰度共生矩阵(GLCM)统计Sentinel-1ASAR数据的纹理信息特征,对比传统监督分类并验证两种遥感图像海冰分类精度。研究指出:对Landsat-8数据,基于CART自动决策树的分类精度最高,总精度达81.68%;而Sentinel-1ASAR数据,基于最大似然分类的总精度为73.88%,相比于CART自动决策树,其能获得更高的海冰分类精度。分析本研究可知,基于光学数据的CART自动决策树在海冰类型识别中占优,而最大似然分类在SAR数据中对海冰类型的识别度较好,本文为辽东湾海冰监测与预报提供了一种有希望的技术手段。 相似文献
8.
9.
10.
针对以光谱特征差异为依据,提取森林湿地信息精度低的问题,该文采用兼容多源数据的分类回归树(CART)提取方法,并以大沾河国家森林湿地进行实证研究。基于Landsat8遥感数据、Radarsat-2极化雷达数据和地形辅助数据,采用SPM软件分别构建3种特征变量组合的CART决策树模型,并获取分类规则,最后根据规则对研究区的森林湿地信息进行提取。结果表明:3种特征变量组合中,兼容光谱、纹理、雷达与地形辅助数据的CART决策树的森林湿地信息提取精度最高,用户精度和制图精度分别达到了88.46%和82.14%。研究结果体现了雷达数据与地形辅助数据有助于提取森林湿地信息。 相似文献
11.
Rubber (Hevea brasiliensis) tree cultivation is being continuously expanded northward by replacing evergreen forests and swidden-related regenerated vegetation across the uplands of mainland Southeast Asia (MSEA), e.g., Laos, a landlocked mountainous country. The non-native tree establishment in the northern tropical edge, or the non-traditional suitable planting area, provides stable supplies of natural latex, yet also leads to severe ecological degradation and environmental effects in water conservation, soil quality, rainforest fragmentation and biodiversity. Rubber plantations in the northern part of MSEA are normally characterized by periodic deciduous during the dry season, along with a lengthy defoliation-foliation duration, because of seasonal variations in temperature and precipitation. It thus lays a phenological and physiological base for dynamics monitoring with common multispectral (e.g., near-infrared and short-wave infrared bands) satellites, particularly Landsat. However, whether Sentinel-2 red-edge based algorithms are suitable for discriminating rubber plantations is not yet exclusively reported. Here, we developed a red-edge spectral indices (RESI) method through the normalization of three red-edge bands and applied it to identify and map rubber plantations in Luang Namtha Province of northern Laos, where a rubber boom begun in the mid-2000s. The RESI algorithm highlights the sensitivity of red-edge bands to the changes in moisture content and canopy density of rubber plantations. The area of mature rubber plantations was estimated to be 771.2 km2 in this province bordering southwest China in 2018, which was nearly twice as much as that of 2011, with the overall accuracy and kappa coefficient up to 92.50% and 0.91, respectively. Our phenology-based RESI approach not only indicates that Sentinel-2 imagery holds significant potential for monitoring rubber plantations, but also improves the remotely-sensed methods of rubber boom mapping via introducing the red-edge channel. 相似文献
12.
Model Building to Investigate the Role of Spatial Location in Classifying Satellite Image Using SVM,CART and mBACT: A Case Study 总被引：1，自引：0，他引：1
Reshu Agarwal 《Journal of the Indian Society of Remote Sensing》2017,45(4):569-578
This paper investigates the importance of spatial location of pixels in terms of row-column as an additional explanatory variable in classification along with available spectral bands of remotely sensed data. In view of this, a forward step-wise variable selection algorithm is used to select significant bands/variables and build an optimal model to extract the maximum accuracy. Author performed a case study on the area of town of Wolfville acquired by LANDSAT 5 TM data containing six 30 m resolution spectral bands and pixel location as an additional variable. Data are classified into seven classes using three advanced classifiers i.e. classification and regression trees (CART), support vector machines (SVM) and multi-class Bayesian additive classification tree (mBACT). Traditionally, it is assumed that addition of more explanatory variables always increase the accuracy of classified satellite images. However, results of this study show that adding more variables may sometimes confuse the classifier, that is, if selected carefully, fewer variables can provide the more accurate classification. Importance of row-column information turns out to be more beneficial for mBACT followed by SVM. Interestingly, spatial locations did not turn out to be useful for CART. Based on the findings of this study, mBACT appears to be a slightly better classifier than SVM and a substantially better than CART. 相似文献
13.
Evaluating the impact of red-edge band from Rapideye image for classifying insect defoliation levels
The prospect of regular assessments of insect defoliation using remote sensing technologies has increased in recent years through advances in the understanding of the spectral reflectance properties of vegetation. The aim of the present study was to evaluate the ability of the red edge channel of Rapideye imagery to discriminate different levels of insect defoliation in an African savanna by comparing the results of obtained from two classifiers. Random Forest and Support vector machine classification algorithms were applied using different sets of spectral analysis involving the red edge band. Results show that the integration of information from red edge increases classification accuracy of insect defoliation levels in all analysis performed in the study. For instance, when all the 5 bands of Rapideye imagery were used for classification, the overall accuracies increases about 19% and 21% for SVM and RF, respectively, as opposed to when the red edge channel was excluded. We also found out that the normalized difference red-edge index yielded a better accuracy result than normalized difference vegetation index. We conclude that the red-edge channel of relatively affordable and readily available high-resolution multispectral satellite data such as Rapideye has the potential to considerably improve insect defoliation classification especially in sub-Saharan Africa where data availability is limited. 相似文献
14.
基于Sentinel-2的潮间红树林提取方法 总被引：1，自引：0，他引：1
位于潮间带的红树林可能在高潮时被海水淹没的特点，使得传统的植被提取方法在红树林信息提取方面存在局限性。本文在对比分析了出露的红树林、高潮水位淹没的红树林、海水水体的光谱特征后，提出了一种利用归一化潮间红树林指数（NIMI）提取潮间带红树林的方法。该指数是由植被强吸收的红波段，强反射的两个红边波段和近红外波段组成的归一化表达式。利用该指数对福建省龙海九龙江口湿地的红树林进行了分类提取，提取结果与高分二号影像目视验证和现场调查结果进行了对照。结果显示，该方法提取红树林的用户精度达到93.98%，并显著优于利用归一化水体指数（NDWI）、归一化植被指数（NDVI）及随机森林的结果。 相似文献
15.
The aim of this study was to detect and map MSV using RapidEye multispectral sensor in Ofcolaco farm. To achieve this objective, the acquired RapidEye sensor was classified using the robust Random Forest algorithm. Furthermore, the variable importance technique was used to determine the influence of each spectral band and indices on the mapping accuracy. For better performance of image data, the value of the commonly used vegetation indices in improving the classification accuracy was tested. The results revealed that the use of RapidEye spectral bands in detection and mapping of MSV yielded good classification results with an overall accuracy of 82.75%. The inclusion of vegetation indices computed from RapidEye sensor improved the classification accuracies by 3.4%. The most important RapidEye spectral bands in classifying MSV were near infrared, blue and red-edge. On the other hand, the most important vegetation indices were the Soil adjusted vegetation index, Enhanced vegetation index, Red index and Normalized Vegetation Index. The current study recommends future studies to assess the importance of multi-temporal remote sensing applications in detecting and monitoring the spread of MSV. 相似文献