Classification Of Spatial Data Research Articles

LCM2021 – the UK Land Cover Map 2021

Abstract. Land cover is a key environmental variable, underpinning widespread environmental research and decision making. The UK Centre for Ecology and Hydrology (UKCEH) has provided reliable land cover information since the early 1990s; this supports multiple scientific, government and commercial objectives. Recent advances in computation and satellite data availability have enabled annual UKCEH land cover maps since 2017. Here, we introduce the latest, annual UK Land Cover Map representing 2021 (LCM2021), and we describe its production and validation. LCM2021 methods replicate those of LCM2017 to LCM2020 with minor deviations in cloud-masking processes and training data sourcing to enhance accuracy. LCM2021 is based on the classification of satellite and spatial context data into 21 land cover or habitat classes, from which a product suite is derived. The production of LCM2021 involved three highly automated key stages: pre-processing of input data, image classification and production of the final data products. Google Earth Engine scripts were used to create an input data stack of satellite and context data. A set of training areas was created based on data harvested from historic UKCEH land cover maps. The training data were used to construct a random forest classifier, which yielded classified images. Compiled results were validated against 35 182 reference samples, with correspondence tables indicating variable class accuracy and an overall accuracy of 82.6 % for the 21-class data and 86.5 % at a 10-aggregated-classes level. The UK Land Cover Map product suite includes a set of raster products in various projections, thematic and spatial resolutions (10 m, 25 m and 1 km), and land–parcel or vector products. The data are provided in 21-class (all configurations) and aggregated 10-class (1 km raster products only) versions. All raster products are freely available for academic and non-commercial research. The data for Great Britain (GB) are provided in the British National Grid projection (EPSG: 27700) and the Northern Ireland (NI) data are in the TM75 Irish Grid (EPSG: 29903). Information on how to access the data is given in the “Data availability” section of the paper.

A Behaviorally Evidence-based Method for Computing Spatial Comparisons of Image Scenarios

<p>Large amounts of noise and a lack of contextual domain knowledge lead to slow and inefficient cross-domain image learning. This paper proposes an image scenario spatial data classification model based on evidence-based behavioral logic, intervenes in image annotation through evidence-based dynamic knowledge graphs, and uses spatial similarity measurement to evaluate the effectiveness and robustness of the method. The results show that: 1) Organizing the dynamic knowledge graphs of contextual domain knowledge by behavioral logic can significantly improve the association efficiency of each model. 2) The calculation method of image scenario space comparison based on behavior evidence can decrypt the implicit knowledge of images and significantly improve the effectiveness of image scenario space interpretation. The research results are helpful to guide the design and implementation of cross-domain image interpretation systems and improve the efficiency of information sharing.</p> <p> </p>

Pattern recognition describing spatio-temporal drivers of catchment classification for water quality

Classification using spatial data is foundational for hydrological modelling, particularly for ungauged areas. However, models developed from classified land use drivers deliver inconsistent water quality results for the same land uses and hinder decision-making guided by those models. This paper explores whether the temporal variation of water quality drivers, such as season and flow, influence inconsistency in the classification, and whether variability is captured in spatial datasets that include original vegetation to represent the variability of biotic responses in areas mapped with the same land use. An Artificial Neural Network Pattern Recognition (ANN-PR) method is used to match catchments by Dissolved Inorganic Nitrogen (DIN) patterns in water quality datasets partitioned into Wet vs Dry Seasons and Increasing vs Retreating flows. Explainable artificial intelligence approaches are then used to classify catchments via spatial feature datasets for each catchment. Catchments matched for sharing patterns in both spatial data and DIN datasets were corroborated and the benefit of partitioning the observed DIN dataset evaluated using Kruskal Wallis method. The highest corroboration rates for spatial data classification with DIN classification were achieved with seasonal partitioning of water quality datasets and significant independence (p < 0.001 to 0.026) from non-partitioned datasets was achieved. This study demonstrated that DIN patterns fall into three categories suited to classification under differing temporal scales with corresponding vegetation types as the indicators. Categories 1 and 3 included dominance of woodlands in their datasets and catchments suited to classify together change depending on temporal scale of the data. Category 2 catchments were dominated by vineforest and classified catchments did not change under different temporal scales. This demonstrates that including original vegetation as a proxy for differences in DIN patterns will help guide future classification where only spatially mapped data is available for ungauged catchments and will better inform data needs for water modelling.

Development of a neural network model for spatial data analysis

Objectives. The paper aimed to develop and validate a neural network model for spatial data analysis. The advantage of the proposed model is the presence of a large number of degrees of freedom allowing its flexible configuration depending on the specific problem. This development is part of the knowledge base of a deep machine learning model repository including a dynamic visualization subsystem based on adaptive web interfaces allowing interactive direct editing of the architecture and topology of neural network models.Methods. The presented solution to the problem of improving the accuracy of spatial data analysis and classification is based on a geosystem approach for analyzing the genetic homogeneity of territorial-adjacent entities of different scales and hierarchies. The publicly available EuroSAT dataset used for initial validation of the proposed methodology is based on Sentinel-2 satellite imagery for training and testing machine learning models aimed at classifying land use/land cover systems. The ontological model of the repository including the developed model is decomposed into domains of deep machine learning models, project tasks and data, thus providing a comprehensive definition of the formalizing area of knowledge. Each stored neural network model is mapped to a set of specific tasks and datasets. Results. Model validation for the EuroSAT dataset algorithmically extended in terms of the geosystem approach allows classification accuracy to be improved under training data shortage within 9% while maintaining the accuracy of ResNet50 and GoogleNet deep learning models.Conclusions. The implemention of the developed model into the repository enhances the knowledge base of models for spatial data analysis as well as allowing the selection of efficient models for solving problems in the digital economy.

Development of the regional water balance regulation concept based on the geosystem approach

The article is devoted to solving the scientific problem of the regional water balance regulation concept development. The authors have proposed a methodology for solving the problem of the scarcity of labeled spatial data, based on the geosystem approach. A deep neural network model for the analysis of the geosystem model of the territory is proposed. An approach to the formation of ensembles of classifiers to improve the accuracy of the classification of spatial data is described. It is proved that the hierarchical structuring of geosystems optimizes diagnostics of the leading agents of physical and geographical factors interaction, patterns of spatio-temporal changes in their states, direction of metabolic processes and transformation of matter and energy development. Studies have shown that the combination of factors variety in the water balance formation forms many types of water resources conditions formation.

A technique for Spatial Data Classification Method Using Random Forest based Correlation

&#x0D; &#x0D; &#x0D; Spatial data mining is a technique mainly used for predictive analytics. Data mining is the method of determining necessary samples from spatial datasets using machine learning methods. But, the existing prediction methods failed to forecast accurately with more accuracy and reduced error. A Random Forest Correlation based Fuzzy classification technique is introduced to improve the spatial data classification and error reduction. The algorithm constructs a random forest and the correlation is measured between the training and testing features. A fuzzy rule based classification is performed for classification into subsets. The proposed method takes forest fire dataset as input and evaluates the parameters such as classification accuracy, space complexity and classification time. The output is a subset of classes of fired and non fired region with enhanced classification accuracy, minimal false-positive rate and reduced time complexity.&#x0D; &#x0D; &#x0D; &#x0D; &#x0D; &#x0D; &#x0D; &#x0D; &#x0D;

On the classification error rates in terms of semivariograms for Gaussian universal kriging models

Bayes multiclass classification of spatial Gaussian data following the universal kriging model is considered. The closed-form expressions for the maximum likelihood (ML) estimator of regression parameters and the actual error rate (AER) in terms of semivariograms are derived.

High Spatial Resolution Remote Sensing Data Classification Method Based on Spectrum Sharing

Existing remote sensing data classification methods cannot achieve the sharing of remote sensing image spectrum, leading to poor fusion and classification of remote sensing data. Therefore, a high spatial resolution remote sensing data classification method based on spectrum sharing is proposed. A page frame recovery algorithm (PFRA) is introduced to allocate the wireless spectrum resources in low-frequency band, and a dynamic spectrum sharing mechanism is designed between the primary and secondary users of remote sensing images. Based on this, D-S evidence theory is used to fuse high spatial resolution remote sensing data and correct the pixel brightness of the fused multispectral image. The initial data are normalized, the feature of spectral image is extracted, the convolution neural network classification model is constructed, and the remote sensing image is segmented. Experimental results show that the proposed method takes shorter time and has higher accuracy for high spatial resolution image segmentation. High spatial resolution remote sensing data classification is more efficient, and the accuracy of data classification and remote sensing image fusion are more ideal.

Machine Learning Algorithms for Urban Land Use Planning: A Review

Urbanization is persistent globally and has increasingly significant spatial and environmental consequences. It is especially challenging in developing countries due to the increasing pressure on the limited resources, and damage to the bio-physical environment. Traditional analytical methods of studying the urban land use dynamics associated with urbanization are static and tend to rely on top-down approaches, such as linear and mathematical modeling. These traditional approaches do not capture the nonlinear properties of land use change. New technologies, such as artificial intelligence (AI) and machine learning (ML) have made it possible to model and predict the nonlinear aspects of urban land dynamics. AI and ML are programmed to recognize patterns and carry out predictions, decision making and perform operations with speed and accuracy. Classification, analysis and modeling using earth observation-based data forms the basis for the geospatial support for land use planning. In the process of achieving higher accuracies in the classification of spatial data, ML algorithms are being developed and being improved to enhance the decision-making process. The purpose of the research is to bring out the various ML algorithms and statistical models that have been applied to study aspects of land use planning using earth observation-based data (EO). It intends to review their performance, functional requirements, interoperability requirements and for which research problems can they be applied best. The literature review revealed that random forest (RF), deep learning like convolutional neural network (CNN) and support vector machine (SVM) algorithms are best suited for classification and pattern analysis of earth observation-based data. GANs (generative adversarial networks) have been used to simulate urban patterns. Algorithms like cellular automata, spatial logistic regression and agent-based modeling have been used for studying urban growth, land use change and settlement pattern analysis. Most of the papers reviewed applied ML algorithms for classification of EO data and to study urban growth and land use change. It is observed that hybrid approaches have better performance in terms of accuracies, efficiency and computational cost.

Effect of anisotropy coeficient on error rates of linear discriminant functions

Paper deals with statistical classification of spatial data as a part of widely applicable statistical approach to pattern recognition. Error rates in supervised classification of Gaussian random field observation into one of two populations specified by different constant means and common stationary geometric anisotropic covariance are considered. Formula for the exact Bayesian error rate is derived. The influence of the ratio of anisotropy to the error rates is evaluated numerically for the case of complete parametric certainty.

Regression Tree Based Correlation Technique in Spatial Data Classification

Data mining is the process of discovering useful patterns from large geo-spatial datasets with the help of machine learning methods. . The machine learning methods plays an important role for data analytics modeling and visualization. Geo-spatial data is a significant task in many application domains, such as environmental science, geographic information science, and social networks. However, the existing spatial pattern discovery and prediction techniques failed to predict the event accurately with minimum error and time consumption. . In this paper, a novel Pearson Correlated Regression Tree-based Affine Projective spatial data Classification (PCRT-APSDC) technique is proposed to improve the spatial data classification and minimize error based on the Affine Projective classification technique. The proposed algorithm employs a fuzzy rule procedure that constructs the regression tree. The fuzzy rule is applied for linking the inputs (i.e. spatial data) with the outputs (i.e. classification results). Our goal is to classify the data into two subsets such as fired region and non-fired region. Experimental evaluation is carried out using a forest fire dataset with different factors such as classification accuracy, false-positive rate, and classification time. The results confirm that the proposed technique predicts the fired region with increased spatial data classification accuracy and minimized time as well as false-positive rate than the state-of-the-art methods.

Geospatial Approach for the Analysis of Forest Cover Change Detection using Machine Learning

Spatial data classification is famous over recent years in order to extract knowledge and insights into the data. It occurs because vast experimentation was used with various classifiers, and significant improvement was examined in accuracy and performance. This study aimed to analyze forest cover change detection using machine learning. Supervised and unsupervised learning methods were used to analyze spatial data. A Vector machine was used to support the supervised learning, and a neural network method was used to support unsupervised learning. The Normalized Difference Vegetation Index (NDVI) was used to identify the bands and extract pixel information relevant to the vegetation. The supervised method shows better results because of its robust performance and better analysis of spatial data classification using vegetation index. The proposed system experimentation was implemented by analyzing the results obtained from Support Vector Machine (SVM) and NN (Neural Network) methods. It is demonstrated in the results that the use of NDVI mainly enhances the performance and increases the classifier's accuracy to a greater extent.&#x0D; Keywords: Spatial data; Normalized Difference Vegetation Index; NDVI;Vegetation index, Support Vector Machine; Neural Network; Forest Cover Change&#x0D; Copyright (c) 2020 Geosfera Indonesia Journal and Department of Geography Education, University of Jember&#x0D; This work is licensed under a Creative Commons Attribution-Share A like 4.0 International License

Deep learning-based urban big data fusion in smart cities: Towards traffic monitoring and flow-preserving fusion

ObjectiveIn the last few years, several techniques and models are used for retrieving significant information from urban big data of smart cities. This research work aims at developing a data fusion-based traffic congestion control system in smart cities using a deep learning model. MethodologyA hybrid model based on the convolution neural network (CNN) and long short term memory (LSTM) architectures are used for region-based traffic flow predictions in smart cities. CNN is used for the classification of spatial data while LSTM for temporal data. ConclusionThe experiments used the CityPulse Traffic and CityPulse Pollution datasets, and measured root mean square error (RMSE), time consumption and accuracy. A small RMSE value of 49 and highest accuracy of 92.3% compared to other baseline models depicts the applicability of the proposed model in the region-based traffic flow prediction problems in the smart cities.

A novel approach for scene classification from remote sensing images using deep learning methods

ABSTRACT Deep learning plays a major task in classification of unsupervised data, which utilises network enhanced learning. This technique proves to be powerful in remote sensing (RS) field for spatial data classification. In the existing environment, huge amounts of data from the earth observation satellites known as satellite images time series (SITS) are gathered, which can be utilised for observing the areas related to geography over through time. In this proposed model the time series model utilised is based on geography. There exists a challenge on how these types of information can be analysed in the field of remote sensing. Notable, techniques related to deep learning substantiated in dealing with remote sensing usually for classification of scene . In this paper, we propose an enhanced classification method involving Recurrent Neural Network (RNN) along with Random forest (RF) for land classification using satellite images, which are publicly available for various research purposes. We utilised spatial data gathered from the satellite images (i.e. time series). Our experimental classification is based on pixel and object-based classification. The attained analysis illustrates that the proposed model outperforms the other present day remote sensing classification techniques by producing 87% target accuracy of classification scene from satellite images.

Optimalisasi model prediksi kesesuaian lahan kelapa sawit menggunakan algoritme pohon keputusan spasial

Land suitability evaluation has a vital role in land use planning aimed to increase food production effectiveness. Palm oil is a leading and strategic commodity for Indonesian people, which is predicted consumption will exceed production in the future. This study aims to evaluate palm oil land suitability using a spatial decision tree algorithm that is conventional decision tree modification for spatial data classification with adding spatial join relation. The spatial dataset consists of eight explanatory layers (soil nature and characteristics), and a target layer (palm oil land suitability) in Bogor District, Indonesia. This study produced three models, where the best model was obtained based on optimizing accuracy (98.18 %) and modeling time (1.291 seconds). The best model has 23 rules, soil texture as the root node, two variables (drainage and cation exchange capacity) are uninvolved, with land suitability visualization obtains percentage S2 (29.94 %), S3 (53.16 %), N (16.57 %), and water body (0.33 %).

Performance Evaluations of Gaussian Spatial Data Classifiers Based on Hybrid Actual Error Rate Estimators

Discrimination and classification of spatial data has been widely mentioned in the scientific literature, but lacks full mathematical treatment and easily available algorithms and software. This paper fills this gap by introducing the method of statistical classification based on Bayes discriminant function (BDF) and by providing original approach for the classifier performance evaluation. Supervised classification of spatial data with response variable modelled by Gaussian random fields (GRF) with continuous or discrete spatial index is studied. Populations are assumed to be with different regression parameters vectors. Classification rule based on BDF with inserted ML estimators of regression and scale parameters is studied. We focus on the derived actual error rate (AER) and the approximation of the expected error rate (AEER) for both types of models. These are used in the construction of hybrid actual error rate estimators that are spatial modifications of widely applicable D and O estimators applied in cases of independent observations. Simulation experiments are used for comparison of proposed AER estimators by the minimum of unconditional mean squared error criterion for both types of GRF models.

Comparing the Capabilities of SVMC and MLC Using Contingency Matrix and a Novel Template

Since extraction of useful information from remote sensing data is important, scientists manage to propose efficient algorithms for automatic extraction of constructive information from the satellite imageries. To date, image classification has benefitted from advancements in improved computational power and algorithm development. Therefore, Satellite image classification using GeoEye-1, High Resolution Satellite Image (HRSI) of 2016, Support Vector Machine Classifier (SVMC) and Maximum Likelihood Classifier (MLC) were performed with a view to comparing the capabilities of SVMC and MLC using Post-processing Accuracy Assessment (PAA) and a Novel Template in producing urban land use and land cover map of the area. The objectives include performing supervised classification using SVM and MLC in ENVI Software, analysing the performance of SVM and MLC in mapping geometric features using error matrix and a new template. The methodology used comprise Image acquisition, Image enhancement, Image Sub-setting, Extraction of Regions of Interests (ROIs) and its separability index analysis, supervised classification using SVMC and MLC, Post-Processing Accuracy Assessment, Statistical Analyses, and Preparation of maps. ENVI 5.1 software was used for image processing, masking, spatial data analysis and classification. Microsoft Excel, GraphPad Prism ver.7.0 and IBM SPSS ver.21 were used for statistical analysis. The result of image classification indicates that Nnewi-North L.G.A is having 13.52% of Built-up Areas, 24.23% of Vegetation, 22.05% of Water bodies, Farm lands is equal to 39.40% and open/bare surface is 0.81% using SVMC while MLC result shows that Built-up Areas is14.99%, Vegetation is 13.01%, Water bodies is 34.08%, Farm lands is 36.00% and open/bare surface is 1.32%. Error Matrix and Kappa Coefficient results revealed that SVMC is better than MLC as follows (SVMC overall Accuracy is 98.07% and Kappa Coefficient is 0.97 while MLC overall Accuracy is 82.50% and Kappa Coefficient is 0.76. Additional statistical testing with aggregate mean from SVM and MLC was used to determine the significance of the mean difference using the researcher’s developed template called “Post Confusion Matrix” (PoCoMa). The result showed that the t-statistics is 0.670 with probability value of -0.476 which is greater than 0.05, thus, the null hypothesis was accepted with a deduction that using any of the algorithms (SVM and MLC) yields no significant difference in performance and efficiency of result of the map produced. The overall study revealed that both classifiers are efficient and accurate statistically, without any significant difference but using error matrix analysis, the research revealed that ‘Support Vector Machine Classifier’ is robust in extracting urban landscape from HRSI, especially Built-up areas and open/bare surfaces. The research recommends there is need for periodic urban LULC analysis to guide stakeholders in Planning, Monitoring, and Management of ‘Urban Areas’ among others. Keywords: Support Vector Machine Classifier (SVMC), Maximum Likelihood Classifier (MLC), Post Confusion Matrix (PoCoMa), High Resolution Satellite Image (HRSI), ENVI 5.1 software, and GraphPad Prism ver.7.0. DOI : 10.7176/JIEA/9-4-06 Publication date :June 30 th 2019

Particle Swarm Optimization Algorithm for Neuro-Fuzzy Prospectivity Analysis Using Continuously Weighted Spatial Exploration Data

Classification of spatial exploration data for exploration targeting using neuro-fuzzy models means that the many spatial values have to be simplified and assigned to a few classes. The simplification of complex geological information, which illustrates a high degree of variability, results in overly simplistic models based on the presumption of homogeneous earth. However, such an assumption is not valid. In this paper, we illustrate the superiority of using continuously weighted spatial evidence values compared to discretely weighted evidence data, and how continuously weighted spatial evidence values can increase the efficiency of neuro-fuzzy exploration targeting models. The results of this study demonstrate that neuro-fuzzy targeting model generated with continuously weighted spatial evidence values is superior to that of the neuro-fuzzy model generated with discretely weighted exploration evidence data.

Mapping riparian habitat using a combination ofremote-sensing techniques

ABSTRACTLight detection and ranging (lidar) and object-oriented classification (OOC) can be used to overcome the shortcomings of the traditional pixel-based classification (PBC) of coarse spatial resolution data, such as Landsat data, for habitat mapping in riparian zones. The purposes of this study were to investigate methods to classify multispectral data and lidar for riparian habitat mapping, and to identify major habitat components for two target species. The mapping of riparian habitat based on OOC and Decision Tree Classification (DTC) was carried out by merging vertical data from lidar and spectral data of high-resolution imagery. Our results showed an overall classification accuracy of 88.2%. In particular, small and continuous habitat types, such as short and tall grasses, rock outcrop and gravel, and riffles, improved the classification accuracy compared with the pixel-based methods. The habitat patches and paths for each target species were identified by incorporating the point data from the field survey and the outcomes of image classification. Our study demonstrated that the proposed methodology can be successfully used for the identification and restoration of fragmented riparian habitats, and can offer an opportunity to obtain high classification accuracies for microhabitat components in dynamic riverine areas.

Deformation information system for facilitating studies of mining-ground deformations, development, and applications

Abstract. The paper presents the concept of the deformation information system (DIS) to support and facilitate studies of mining-ground deformations. The proposed modular structure of the system includes data collection and data visualisation components, as well as spatial data mining, modelling and classification modules. In addition, the system integrates interactive three-dimensional models of the mines and local geology. The system is used to calculate various parameters characterising ground deformation in space and time, i.e. vertical and horizontal displacement fields, deformation parameters (tilt, curvature, and horizontal strain) and input spatial variables for spatial data classifications. The core of the system in the form of an integrated spatial and attributive database has been described. The development stages and the functionality of the particular components have been presented and example analyses utilising the spatial data mining and modelling functions have been shown. These include, among other things, continuous vertical and horizontal displacement field interpolations, calculation of parameters characterising mining-ground deformations, mining-ground category classifications, data extraction procedures and data preparation preprocessing procedures for analyses in external applications. The DIS has been developed for the Walbrzych coal mines area in SW Poland where long-time mining activity ended at the end of the 20th century and surface monitoring is necessary to study the present-day condition of the former mining grounds.