Time Series Of Multispectral Images Research Articles

Remote Sensing Monitoring on the Territory of the Volga Carbon Polygon (Republic of Tatarstan, Russia)

Remote sensing technology and geographic information systems allow the assessment of terrestrial carbon stocks in large areas. The article considers the procedure of remote monitoring organization in the forest and water parts of the Volga Carbon polygon (Republic of Tatarstan, Russia). The main purpose of the article is to present the structure of remote monitoring of the polygon, show the results of the first year of remote research, and discuss the need to use other remote sensing methods. This study reviews and highlights the advantages and limitations of various remote sensing methods and sensors, including optical, multispectral, radar, and lidar, which are widely used for above-ground biomass (AGB) and carbon stocks (CS) estimation. The first results of field studies by unmanned aerial vehicle (UAV) at the Volga Carbon polygon are presented: orthophotomaps, digital terrain and relief models, bathymetric map and multispectral image time-series. Different remote sensing methods from echolocation to the multispectral survey are used, and the most optimal remote sensing data and processing methods are determined. For the first time for the Republic of Tatarstan, an integrated carbon balance monitoring system has been developed.

Mapping Plant Species in a Former Industrial Site Using Airborne Hyperspectral and Time Series of Sentinel-2 Data Sets

Industrial activities induce various impacts on ecosystems that influence species richness and distribution. An effective way to assess the resulting impacts on biodiversity lies in vegetation mapping. Species classification achieved through supervised machine learning algorithms at the pixel level has shown promising results using hyperspectral images and multispectral, multitemporal images. This study aims to determine whether airborne hyperspectral images with a high spatial resolution or phenological information obtained by spaceborne multispectral time series (Sentinel-2) are suitable to discriminate species and assess biodiversity in a complex impacted context. The industrial heritage of the study site has indeed induced high spatial heterogeneity in terms of stressors and species over a reduced scale. First, vegetation indices, derivative spectra, continuum removed spectra, and components provided by three feature extraction techniques, namely, Principal Component Analysis, Minimal Noise Fraction, and Independent Component Analysis, were calculated from reflectance spectra. These features were then analyzed through Sequential Floating Feature Selection. Supervised classification was finally performed using various machine learning algorithms (Random Forest, Support Vector Machines, and Regularized Logistic Regression) considering a probability-based rejection approach. Biodiversity metrics were derived from resulted maps and analyzed considering the impacts. Average Overall Accuracy (AOA) reached up to 94% using the hyperspectral image and Regularized Logistic Regression algorithm, whereas the time series of multispectral images never exceeded 72% AOA. From all tested spectral transformations, only vegetation indices applied to the time series of multispectral images increased the performance. The results obtained with the hyperspectral image degraded to the specifications of Sentinel-2 emphasize the importance of fine spatial and spectral resolutions to achieve accurate mapping in this complex context. While no significant difference was found between impacted and reference sites through biodiversity metrics, vegetation mapping highlighted some differences in species distribution.

Potential of N-CovSel for Variable Selection: A Case Study on Time-Series of Multispectral Images

Multispectral image time-series have been promising for some years; yet, the substantial advance of the technology involved, with unprecedented combinations of spatial, temporal, and spectral capabilities for remote sensing applications, raises new challenges, in particular, the need for methodologies that can process the different dimensions of satellite information. Considering that the multi-collinearity problem is present in remote sensing time-series, regression models are widespread tools to model multi-way data. This paper presents the results of the analysis of a high order data of Sentinel-2-time series, conducted in the framework of extreme weather event. A feature extraction method for multi-way data, N-CovSel was used to identify the most relevant features explaining the loss of yield in Mediterranean vineyards during the 2019 heatwave. Different regression models (uni-way and multi-way) from features extracted from the N-CovSel algorithm were calibrated based on available heat wave impact data for 107 vineyard blocks in the Languedoc-Roussillon region and multispectral time-series predictor data for the period May to August. The performance of the models was evaluated by the r2 and the root mean square of error (RMSE) as follows: for the temporal N-PLS model (r2 = 0.62—RMSE = 11%), for the spatial N-PLS model (r2 = 0.61—RMSE = 12%) and the temporal-spectral PLS model (r2 = 0.63—RMSE = 11%). The results validated the effectiveness of the proposed N-CovSel algorithm in order to reduce the number of total variables and restricting it to the most significant ones. The N-CovSel algorithm seems to be a suitable choice to interpret complex multispectral imagery by temporally discriminating the most appropriate spectral information.

EXPLOITATION OF SPECTRAL AND TEMPORAL INFORMATION FOR MAPPING PLANT SPECIES IN A FORMER INDUSTRIAL SITE

Abstract. Characterization and seasonal (periodic) monitoring of plant species distribution in the context of former industrial activity is crucial to assess long-term anthropogenic footprint on vegetated area. Species discrimination has shown promising results using both HyperSpectral (HS) and MultiSpectral (MS) images. Airborne HS instruments enable high spatial and spectral resolution imagery while time series of satellite MS images provide high temporal resolution and phenological information. This paper aims to compare supervised classification results obtained with non-parametric (Random Forest, RF, Support Vector Machine, SVM) and parametric methods (Regularized Logistic Regression, RLR) applied on both kinds of images acquired on an industrial brownfield. The studied site is a complex vegetated environment due to species diversity: 8 dominant species are retained. The performance obtained by preliminary feature selection based on principal component analysis and vegetation indices, to improve separability of spectral or temporal information according to species, is analysed. The best performance is obtained by RLR method applied on HS data without feature selection (global accuracy of 93 %). Feature selection is found to be a necessary step to perform classification with time series of MS images. Species that are difficult to distinguish from the HS image, namely Salix and Populus, are well separated using Sentinel-2 images (precision around 70%).

A Metric for Evaluating the Geometric Quality of Land Cover Maps Generated with Contextual Features from High-Dimensional Satellite Image Time Series without Dense Reference Data

Land cover maps are a key resource for many studies in Earth Observation, and thanks to the high temporal, spatial, and spectral resolutions of systems like Sentinel-2, maps with a wide variety of land cover classes can now be automatically produced over vast areas. However, certain context-dependent classes, such as urban areas, remain challenging to classify correctly with pixel-based methods. Including contextual information into the classification can either be done at the feature level with texture descriptors or object-based approaches, or in the classification model itself, as is done in Convolutional Neural Networks. This improves recognition rates of these classes, but sometimes deteriorates the fine-resolution geometry of the output map, particularly in sharp corners and in fine elements such as rivers and roads. However, the quality of the geometry is difficult to assess in the absence of dense training data, which is usually the case in land cover mapping, especially over wide areas. This work presents a framework for measuring the geometric precision of a classification map, in order to provide deeper insight into the consequences of the use of various contextual features, when dense validation data is not available. This quantitative metric, named the Pixel Based Corner Match (PBCM), is based on corner detection and corner matching between a pixel-based classification result, and a contextual classification result. The selected case study is the classification of Sentinel-2 multi-spectral image time series, with a rich nomenclature containing context-dependent classes. To demonstrate the added value of the proposed metric, three spatial support shapes (window, object, superpixel) are compared according to their ability to improve the classification performance on this challenging problem, while paying attention to the geometric precision of the result. The results show that superpixels are the best candidate for the local statistics features, as they modestly improve the classification accuracy, while preserving the geometric elements in the image. Furthermore, the density of edges in a sliding window provides a significant boost in accuracy, and maintains a high geometric precision.

A Novel Approach to the Unsupervised Update of Land-Cover Maps by Classification of Time Series of Multispectral Images

This paper presents an unsupervised approach that extracts reliable labeled units from outdated maps to update them using time series (TS) of recent multispectral (MS) images. The method assumes that: 1) the source of the map is unknown and may be different from remote sensing data; 2) no ground truth is available; 3) the map is provided at polygon level, where the polygon label represents the dominant class; and 4) the map legend can be converted into a set of classes discriminable with the TS of images (i.e., no land-use classes that require manual analysis are considered). First, the outdated map is adapted to the spatial and spectral properties of the MS images. Then, the method identifies reliable labeled units in an unsupervised way by a two-step procedure: 1) a clustering analysis performed at polygon level to detect samples correctly associated to their labels and 2) a consistency analysis to discard polygons far from the distribution of the related land-cover class (i.e., having high probability of being mislabeled). Finally, the map is updated by classifying the recent TS of MS image with an ensemble of classifiers trained using only the reference data derived from the map. The experimental results obtained updating the 2012 Corine Land Cover (CLC) and the GlobLand30 in Trentino Alto Adige (Italy) achieved 93.2% and 93.3% overall accuracy (OA) on the validation data set. The method increased the OA up to 18% and 11.5% with respect to the reference methods on the 2012 CLC and the GlobLand30, respectively.

Tropical Deforestation and Recolonization by Exotic and Native Trees: Spatial Patterns of Tropical Forest Biomass, Functional Groups, and Species Counts and Links to Stand Age, Geoclimate, and Sustainability Goals

We mapped native, endemic, and introduced (i.e., exotic) tree species counts, relative basal areas of functional groups, species basal areas, and forest biomass from forest inventory data, satellite imagery, and environmental data for Puerto Rico and the Virgin Islands. Imagery included time series of Landsat composites and Moderate Resolution Imaging Spectroradiometer (MODIS)-based phenology. Environmental data included climate, land-cover, geology, topography, and road distances. Large-scale deforestation and subsequent forest regrowth are clear in the resulting maps decades after large-scale transition back to forest. Stand age, climate, geology, topography, road/urban locations, and protection are clearly influential. Unprotected forests on more accessible or arable lands are younger and have more introduced species and deciduous and nitrogen-fixing basal areas, fewer endemic species, and less biomass. Exotic species are widespread—except in the oldest, most remote forests on the least arable lands, where shade-tolerant exotics may persist. Although the maps have large uncertainty, their patterns of biomass, tree species diversity, and functional traits suggest that for a given geoclimate, forest age is a core proxy for forest biomass, species counts, nitrogen-fixing status, and leaf longevity. Geoclimate indicates hard-leaved species commonness. Until global wall-to-wall remote sensing data from specialized sensors are available, maps from multispectral image time series and other predictor data should help with running ecosystem models and as sustainable development indicators. Forest attribute models trained with a tree species ordination and mapped with nearest neighbor substitution (Phenological Gradient Nearest Neighbor method, PGNN) yielded larger correlation coefficients for observed vs. mapped tree species basal areas than Cubist regression tree models trained separately on each species. In contrast, Cubist regression tree models of forest structural and functional attributes yielded larger such correlation coefficients than the ordination-trained PGNN models.

Remote sensing-based crop biomass with water or light-driven crop growth models in wheat commercial fields

This paper explores the ability of Remote Sensing data from space platforms combined with available meteorological parameters to monitor crop biomass accumulation at satellite scale in a direct, operational way, exploiting the temporal information from time series of multispectral images. We describe a methodology to estimate biomass growth by integrating VI-based biophysical parameter and meteorological input along the growing cycle into the physiologically-based crop growth models, employing water or light use efficiency.

Optimization of spectral indices and long-term separability analysis for classification of cereal crops using multi-spectral RapidEye imagery

Crop monitoring using remotely sensed image data provides valuable input for a large variety of applications in environmental and agricultural research. However, method development for discrimination between spectrally highly similar crop species remains a challenge in remote sensing. Calculation of vegetation indices is a frequently applied option to amplify the most distinctive parts of a spectrum. Since no vegetation index exist, that is universally best-performing, a method is presented that finds an index that is optimized for the classification of a specific satellite data set to separate two cereal crop types. The η2 (eta-squared) measure of association – presented as novel spectral separability indicator – was used for the evaluation of the numerous tested indices. The approach is first applied on a RapidEye satellite image for the separation of winter wheat and winter barley in a Central German test site. The determined optimized index allows a more accurate classification (97%) than several well-established vegetation indices like NDVI and EVI (<87%). Furthermore, the approach was applied on a RapidEye multi-spectral image time series covering the years 2010–2014. The optimized index for the spectral separation of winter barley and winter wheat for each acquisition date was calculated and its ability to distinct the two classes was assessed. The results indicate that the calculated optimized indices perform better than the standard indices for most seasonal parts of the time series. The red edge spectral region proved to be of high significance for crop classification. Additionally, a time frame of best spectral separability of wheat and barley could be detected in early to mid-summer.

Classification of small agricultural fields using combined Landsat-8 and RapidEye imagery: case study of northern Serbia

A pixel-based cropland classification study based on the fusion of data from satellite images with different resolutions is presented. It is based on a time series of multispectral images acquired at different resolutions by different imaging instruments, Landsat-8 and RapidEye. The proposed data fusion method capabilities are explored with the aim of overcoming the shortcomings of different instruments in the particular cropland classification scenario characterized by the very small size of crop fields over the chosen agricultural region situated in the plains of Vojvodina in northern Serbia. This paper proposes a data fusion method that is successfully utilized in combination with arobust random forest classifier in improving the overall classification performance, as well as in enabling application of satellite imagery with a coarser spatial resolution in the given specific cropland classification task. The developed method effectively exploits available data and provides an improvement over the existing pixel-based classification approaches through the combination of different data sources. Another contribution of this paper is the employment of crowdsourcing in the process of reference data collection via dedicated smartphone application.

A parametric model for classifying land cover and evaluating training data based on multi-temporal remote sensing data

Time series of multispectral images are widely used to monitor and map land cover. However, high dimensionality and missing data present significant challenges for classification algorithms that use multi-temporal remotely sensed data. Further, generation and assessment of high quality training data, including detection of outliers and changed pixels in training data, is difficult. In this paper we present a new statistical framework that is based on a parametric model that enables a targeted principal component analysis (PCA) to reduce the dimensionality of multi-temporal remote sensing data. In doing so, the model provides a novel basis for land cover classification and evaluating the nature and quality of training data used for supervised classifications. The methodology we describe uses a Kronecker operator to reduce the spectral dimensionality of multi-temporal images while preserving their temporal structure, thereby providing low-dimensional data that is well-suited for classification and outlier detection problems. As part of our framework, we use an expectation–maximization method to impute missing data, and propose new metrics that characterize the representativeness and pixel-to-pixel homogeneity of training sites used for supervised classification. To evaluate our approach, we use data from NASA’s Moderate Resolution Imaging Spectroradiometer (MODIS) and extracted more than 200 training sites where the land cover has been characterized from high spatial resolution imagery. The original input data was composed of 196 features (28 dates×7 bands), and the PCA-based approach we describe captured 91% of the variance, in these 7 bands, in 3 components. Results from maximum likelihood classification show that the retained principal components successfully distinguish land cover classes from one another, with classification results that were comparable to supervised machine learning methods applied to the original MODIS data. Analysis of our site composition metrics show that they successfully characterize the homogeneity (or lack thereof) and representativeness of individual pixels and entire sites relative to other training sites in the same class.

Classification of Time Series of Multispectral Images With Limited Training Data

Image classification usually requires the availability of reliable reference data collected for the considered image to train supervised classifiers. Unfortunately when time series of images are considered, this is seldom possible because of the costs associated with reference data collection. In most of the applications it is realistic to have reference data available for one or few images of a time series acquired on the area of interest. In this paper, we present a novel system for automatically classifying image time series that takes advantage of image(s) with an associated reference information (i.e., the source domain) to classify image(s) for which reference information is not available (i.e., the target domain). The proposed system exploits the already available knowledge on the source domain and, when possible, integrates it with a minimum amount of new labeled data for the target domain. In addition, it is able to handle possible significant differences between statistical distributions of the source and target domains. Here, the method is presented in the context of classification of remote sensing image time series, where ground reference data collection is a highly critical and demanding task. Experimental results show the effectiveness of the proposed technique. The method can work on multimodal (e.g., multispectral) images.