scholarly journals Improving 3-m Resolution Land Cover Mapping through Efficient Learning from an Imperfect 10-m Resolution Map

2020 ◽  
Vol 12 (9) ◽  
pp. 1418
Author(s):  
Runmin Dong ◽  
Cong Li ◽  
Haohuan Fu ◽  
Jie Wang ◽  
Weijia Li ◽  
...  
Keyword(s):  
Land Cover ◽  
Training Data ◽  
Training Dataset ◽  
Land Cover Mapping ◽  
Remotely Sensed Data ◽  
Large Area ◽  
National Scale ◽  
Substantial Progress ◽  
Efficient Learning ◽  
Land Cover Maps

Substantial progress has been made in the field of large-area land cover mapping as the spatial resolution of remotely sensed data increases. However, a significant amount of human power is still required to label images for training and testing purposes, especially in high-resolution (e.g., 3-m) land cover mapping. In this research, we propose a solution that can produce 3-m resolution land cover maps on a national scale without human efforts being involved. First, using the public 10-m resolution land cover maps as an imperfect training dataset, we propose a deep learning based approach that can effectively transfer the existing knowledge. Then, we improve the efficiency of our method through a network pruning process for national-scale land cover mapping. Our proposed method can take the state-of-the-art 10-m resolution land cover maps (with an accuracy of 81.24% for China) as the training data, enable a transferred learning process that can produce 3-m resolution land cover maps, and further improve the overall accuracy (OA) to 86.34% for China. We present detailed results obtained over three mega cities in China, to demonstrate the effectiveness of our proposed approach for 3-m resolution large-area land cover mapping.

scholarly journals Improved land cover mapping using aerial photographs and satellite images

2014 ◽  
Vol 7 (1) ◽  
Author(s):  
Katalin Varga ◽  
Szilárd Szabó ◽  
Gergely Szabó ◽  
György Dévai ◽  
Béla Tóthmérész
Keyword(s):  
Land Cover ◽  
Vegetation Cover ◽  
Satellite Images ◽  
Automated Analysis ◽  
Reliable Information ◽  
Aerial Photographs ◽  
Surface Model ◽  
Land Cover Mapping ◽  
Remotely Sensed Data ◽  
Large Area

AbstractManual Land Cover Mapping using aerial photographs provides sufficient level of resolution for detailed vegetation or land cover maps. However, in some cases it is not possible to achieve the desired information over large areas, for example from historical data where the quality and amount of available images is definitely lower than from modern data. The use of automated and semiautomated methods offers the means to identify the vegetation cover using remotely sensed data. In this paper automated methods were tested on aerial photographs and satellite images to extract better and more reliable information about vegetation cover. These testswere performed by using automated analysis of LANDSAT7 images (with and without the surface model of the Shuttle Radar Topography Mission (SRTM)) and two temporally similar aerial photographs. The spectral bands were analyzed with supervised (maximum likelihood) methods. In conclusion, the SRTM and the combination of two temporally similar aerial photographs from earlier years were useful in separating the vegetation cover on a floodplain area. In addition the different date of the vegetation season also gave reliable information about the land cover. High quality information about old and present vegetation on a large area is an essential prerequisites ensuring the conservation of ecosystems

scholarly journals Unseen Land Cover Classification from High-Resolution Orthophotos Using Integration of Zero-Shot Learning and Convolutional Neural Networks

2020 ◽  
Vol 12 (10) ◽  
pp. 1676 ◽  
Author(s):  
Biswajeet Pradhan ◽  
Husam A. H. Al-Najjar ◽  
Maher Ibrahim Sameen ◽  
Ivor Tsang ◽  
Abdullah M. Alamri
Keyword(s):  
Feature Extraction ◽  
High Resolution ◽  
Land Cover ◽  
Training Data ◽  
Training Dataset ◽  
Land Cover Mapping ◽  
Second Phase ◽  
Test Dataset ◽  
Promising Tool ◽  
The Impact

Zero-shot learning (ZSL) is an approach to classify objects unseen during the training phase and shown to be useful for real-world applications, especially when there is a lack of sufficient training data. Only a limited amount of works has been carried out on ZSL, especially in the field of remote sensing. This research investigates the use of a convolutional neural network (CNN) as a feature extraction and classification method for land cover mapping using high-resolution orthophotos. In the feature extraction phase, we used a CNN model with a single convolutional layer to extract discriminative features. In the second phase, we used class attributes learned from the Word2Vec model (pre-trained by Google News) to train a second CNN model that performed class signature prediction by using both the features extracted by the first CNN and class attributes during training and only the features during prediction. We trained and tested our models on datasets collected over two subareas in the Cameron Highlands (training dataset, first test dataset) and Ipoh (second test dataset) in Malaysia. Several experiments have been conducted on the feature extraction and classification models regarding the main parameters, such as the network’s layers and depth, number of filters, and the impact of Gaussian noise. As a result, the best models were selected using various accuracy metrics such as top-k categorical accuracy for k = [1,2,3], Recall, Precision, and F1-score. The best model for feature extraction achieved 0.953 F1-score, 0.941 precision, 0.882 recall for the training dataset and 0.904 F1-score, 0.869 precision, 0.949 recall for the first test dataset, and 0.898 F1-score, 0.870 precision, 0.838 recall for the second test dataset. The best model for classification achieved an average of 0.778 top-one, 0.890 top-two and 0.942 top-three accuracy, 0.798 F1-score, 0.766 recall and 0.838 precision for the first test dataset and 0.737 top-one, 0.906 top-two, 0.924 top-three, 0.729 F1-score, 0.676 recall and 0.790 precision for the second test dataset. The results demonstrated that the proposed ZSL is a promising tool for land cover mapping based on high-resolution photos.

scholarly journals Continental-Scale Land Cover Mapping at 10 m Resolution Over Europe (ELC10)

2021 ◽  
Vol 13 (12) ◽  
pp. 2301
Author(s):  
Zander Venter ◽  
Markus Sydenham
Keyword(s):  
Land Cover ◽  
Atmospheric Correction ◽  
Google Earth ◽  
Tree Planting ◽  
Training Data ◽  
Classification Model ◽  
Night Time ◽  
Continental Scale ◽  
Land Cover Maps ◽  
Sentinel 2

Land cover maps are important tools for quantifying the human footprint on the environment and facilitate reporting and accounting to international agreements addressing the Sustainable Development Goals. Widely used European land cover maps such as CORINE (Coordination of Information on the Environment) are produced at medium spatial resolutions (100 m) and rely on diverse data with complex workflows requiring significant institutional capacity. We present a 10 m resolution land cover map (ELC10) of Europe based on a satellite-driven machine learning workflow that is annually updatable. A random forest classification model was trained on 70K ground-truth points from the LUCAS (Land Use/Cover Area Frame Survey) dataset. Within the Google Earth Engine cloud computing environment, the ELC10 map can be generated from approx. 700 TB of Sentinel imagery within approx. 4 days from a single research user account. The map achieved an overall accuracy of 90% across eight land cover classes and could account for statistical unit land cover proportions within 3.9% (R2 = 0.83) of the actual value. These accuracies are higher than that of CORINE (100 m) and other 10 m land cover maps including S2GLC and FROM-GLC10. Spectro-temporal metrics that capture the phenology of land cover classes were most important in producing high mapping accuracies. We found that the atmospheric correction of Sentinel-2 and the speckle filtering of Sentinel-1 imagery had a minimal effect on enhancing the classification accuracy (< 1%). However, combining optical and radar imagery increased accuracy by 3% compared to Sentinel-2 alone and by 10% compared to Sentinel-1 alone. The addition of auxiliary data (terrain, climate and night-time lights) increased accuracy by an additional 2%. By using the centroid pixels from the LUCAS Copernicus module polygons we increased accuracy by <1%, revealing that random forests are robust against contaminated training data. Furthermore, the model requires very little training data to achieve moderate accuracies—the difference between 5K and 50K LUCAS points is only 3% (86 vs. 89%). This implies that significantly less resources are necessary for making in situ survey data (such as LUCAS) suitable for satellite-based land cover classification. At 10 m resolution, the ELC10 map can distinguish detailed landscape features like hedgerows and gardens, and therefore holds potential for aerial statistics at the city borough level and monitoring property-level environmental interventions (e.g., tree planting). Due to the reliance on purely satellite-based input data, the ELC10 map can be continuously updated independent of any country-specific geographic datasets.

scholarly journals Effects of Training Set Size on Supervised Machine-Learning Land-Cover Classification of Large-Area High-Resolution Remotely Sensed Data

2021 ◽  
Vol 13 (3) ◽  
pp. 368
Author(s):  
Christopher A. Ramezan ◽  
Timothy A. Warner ◽  
Aaron E. Maxwell ◽  
Bradley S. Price
Keyword(s):  
Machine Learning ◽  
Sample Size ◽  
Remotely Sensed ◽  
Training Data ◽  
Supervised Machine Learning ◽  
Sample Sizes ◽  
Remotely Sensed Data ◽  
Large Area ◽  
Training Set ◽  
Set Size

The size of the training data set is a major determinant of classification accuracy. Nevertheless, the collection of a large training data set for supervised classifiers can be a challenge, especially for studies covering a large area, which may be typical of many real-world applied projects. This work investigates how variations in training set size, ranging from a large sample size (n = 10,000) to a very small sample size (n = 40), affect the performance of six supervised machine-learning algorithms applied to classify large-area high-spatial-resolution (HR) (1–5 m) remotely sensed data within the context of a geographic object-based image analysis (GEOBIA) approach. GEOBIA, in which adjacent similar pixels are grouped into image-objects that form the unit of the classification, offers the potential benefit of allowing multiple additional variables, such as measures of object geometry and texture, thus increasing the dimensionality of the classification input data. The six supervised machine-learning algorithms are support vector machines (SVM), random forests (RF), k-nearest neighbors (k-NN), single-layer perceptron neural networks (NEU), learning vector quantization (LVQ), and gradient-boosted trees (GBM). RF, the algorithm with the highest overall accuracy, was notable for its negligible decrease in overall accuracy, 1.0%, when training sample size decreased from 10,000 to 315 samples. GBM provided similar overall accuracy to RF; however, the algorithm was very expensive in terms of training time and computational resources, especially with large training sets. In contrast to RF and GBM, NEU, and SVM were particularly sensitive to decreasing sample size, with NEU classifications generally producing overall accuracies that were on average slightly higher than SVM classifications for larger sample sizes, but lower than SVM for the smallest sample sizes. NEU however required a longer processing time. The k-NN classifier saw less of a drop in overall accuracy than NEU and SVM as training set size decreased; however, the overall accuracies of k-NN were typically less than RF, NEU, and SVM classifiers. LVQ generally had the lowest overall accuracy of all six methods, but was relatively insensitive to sample size, down to the smallest sample sizes. Overall, due to its relatively high accuracy with small training sample sets, and minimal variations in overall accuracy between very large and small sample sets, as well as relatively short processing time, RF was a good classifier for large-area land-cover classifications of HR remotely sensed data, especially when training data are scarce. However, as performance of different supervised classifiers varies in response to training set size, investigating multiple classification algorithms is recommended to achieve optimal accuracy for a project.

scholarly journals Spatio-Temporal Super-Resolution Land Cover Mapping Based on Fuzzy C-Means Clustering

2018 ◽  
Vol 10 (8) ◽  
pp. 1212 ◽  
Author(s):  
Xiaohong Yang ◽  
Zhong Xie ◽  
Feng Ling ◽  
Xiaodong Li ◽  
Yihang Zhang ◽  
...  
Keyword(s):  
Land Cover ◽  
Spatial Pattern ◽  
Spatial Resolution ◽  
Super Resolution ◽  
Land Cover Mapping ◽  
Fuzzy C Means Clustering ◽  
Land Cover Map ◽  
Spatio Temporal ◽  
Fine Resolution ◽  
Land Cover Maps

Super-resolution land cover mapping (SRM) is a method that aims to generate land cover maps with fine spatial resolutions from the original coarse spatial resolution remotely sensed image. The accuracy of the resultant land cover map produced by existing SRM methods is often limited by the errors of fraction images and the uncertainty of spatial pattern models. To address these limitations in this study, we proposed a fuzzy c-means clustering (FCM)-based spatio-temporal SRM (FCM_STSRM) model that combines the spectral, spatial, and temporal information into a single objective function. The spectral term is constructed with the FCM criterion, the spatial term is constructed with the maximal spatial dependence principle, and the temporal term is characterized by the land cover transition probabilities in the bitemporal land cover maps. The performance of the proposed FCM_STSRM method is assessed using data simulated from the National Land Cover Database dataset and real Landsat images. Results of the two experiments show that the proposed FCM_STSRM method can decrease the influence of fraction errors by directly using the original images as the input and the spatial pattern uncertainty by inheriting land cover information from the existing fine resolution land cover map. Compared with the hard classification and FCM_SRM method applied to mono-temporal images, the proposed FCM_STSRM method produced fine resolution land cover maps with high accuracy, thus showing the efficiency and potential of the novel approach for producing fine spatial resolution maps from coarse resolution remotely sensed images.

Measuring and monitoring land cover, land use, and vegetation characteristics

2010 ◽  
Author(s):  
Ned Horning ◽  
Julie A. Robinson ◽  
Eleanor J. Sterling ◽  
Woody Turner ◽  
Sacha Spector
Keyword(s):  
Land Use ◽  
Land Cover ◽  
Deciduous Forest ◽  
Land Cover Classification ◽  
Remotely Sensed ◽  
Land Cover Mapping ◽  
Remotely Sensed Data ◽  
Soil P ◽  
Vegetation Characteristics ◽  
Classification Project

In terrestrial biomes, ecologists and conservation biologists commonly need to understand vegetation characteristics such as structure, primary productivity, and spatial distribution and extent. Fortunately, there are a number of airborne and satellite sensors capable of providing data from which you can derive this information. We will begin this chapter with a discussion on mapping land cover and land use. This is followed by text on monitoring changes in land cover and concludes with a section on vegetation characteristics and how we can measure these using remotely sensed data. We provide a detailed example to illustrate the process of creating a land cover map from remotely sensed data to make management decisions for a protected area. This section provides an overview of land cover classification using remotely sensed data. We will describe different options for conducting land cover classification, including types of imagery, methods and algorithms, and classification schemes. Land cover mapping is not as difficult as it may appear, but you will need to make several decisions, choices, and compromises regarding image selection and analysis methods. Although it is beyond the scope of this chapter to provide details for all situations, after reading it you will be able to better assess your own needs and requirements. You will also learn the steps to carry out a land cover classification project while gaining an appreciation for the image classification process. That said, if you lack experience with land cover mapping, it always wise to seek appropriate training and, if possible, collaborate with someone who has land cover mapping experience (Section 2.3). Although the terms “land cover” and “land use” are sometimes used interchangeably they are different in important ways. Simply put, land cover is what covers the surface of the Earth and land use describes how people use the land (or water). Examples of land cover classes are: water, snow, grassland, deciduous forest, or bare soil.

scholarly journals Operational Large-Area Land-Cover Mapping: An Ethiopia Case Study

2020 ◽  
Vol 12 (6) ◽  
pp. 954
Author(s):  
Reza Khatami ◽  
Jane Southworth ◽  
Carly Muir ◽  
Trevor Caughlin ◽  
Alemayehu N. Ayana ◽  
...  
Keyword(s):  
Land Cover ◽  
Classification Accuracy ◽  
Landsat 8 ◽  
Large Area ◽  
National Scale ◽  
Night Time ◽  
Explanatory Variables ◽  
Image Compositing ◽  
Elevation Data ◽  
Multi Temporal

Knowledge of land cover and land use nationally is a prerequisite of many studies on drivers of land change, impacts on climate, carbon storage and other ecosystem services, and allows for sufficient planning and management. Despite this, many regions globally do not have accurate and consistent coverage at the national scale. This is certainly true for Ethiopia. Large-area land-cover characterization (LALCC), at a national scale is thus an essential first step in many studies of land-cover change, and yet is itself problematic. Such LALCC based on remote-sensing image classification is associated with a spectrum of technical challenges such as data availability, radiometric inconsistencies within/between images, and big data processing. Radiometric inconsistencies could be exacerbated for areas, such as Ethiopia, with a high frequency of cloud cover, diverse ecosystem and climate patterns, and large variations in elevation and topography. Obtaining explanatory variables that are more robust can improve classification accuracy. To create a base map for the future study of large-scale agricultural land transactions, we produced a recent land-cover map of Ethiopia. Of key importance was the creation of a methodology that was accurate and repeatable and, as such, could be used to create earlier, comparable land-cover classifications in the future for the same region. We examined the effects of band normalization and different time-series image compositing methods on classification accuracy. Both top of atmosphere and surface reflectance products from the Landsat 8 Operational Land Imager (OLI) were tested for single-time classification independently, where the latter resulted in 1.1% greater classification overall accuracy. Substitution of the original spectral bands with normalized difference spectral indices resulted in an additional improvement of 1.0% in overall accuracy. Three approaches for multi-temporal image compositing, using Landsat 8 OLI and Moderate Resolution Imaging Spectroradiometer (MODIS) data, were tested including sequential compositing, i.e., per-pixel summary measures based on predefined periods, probability density function compositing, i.e., per-pixel characterization of distribution of spectral values, and per-pixel sinusoidal models. Multi-temporal composites improved classification overall accuracy up to 4.1%, with respect to single-time classification with an advantage of the Landsat OLI-driven composites over MODIS-driven composites. Additionally, night-time light and elevation data were used to improve the classification. The elevation data and its derivatives improved classification accuracy by 1.7%. The night-time light data improve producer’s accuracy of the Urban/Built class with the cost of decreasing its user’s accuracy. Results from this research can aid map producers with decisions related to operational large-area land-cover mapping, especially with selecting input explanatory variables and multi-temporal image compositing, to allow for the creation of accurate and repeatable national-level land-cover products in a timely fashion.

Sign in / Sign up

Export Citation Format

Share Document