A method for detecting large-scale forest cover change using coarse spatial resolution imagery

Many large countries, including Canada, rely on earth observation as a practical and cost-effective means of monitoring their vast inland ecosystems. A potentially efficient approach is one that detects vegetation changes over a hierarchy of spatial scales ranging from coarse to fine. This paper presents a Change Screening Analysis Technique (Change-SAT) designed as a coarse filter to identify the location and timing of large (>5-10 km²) forest cover changes caused by anthropogenic and natural disturbances at an annual, continental scale. The method uses change metrics derived from 1-km multi-temporal SPOT VEGETATION and NOAA AVHRR imagery (reflectance, temperature, and texture information) and ancillary spatial variables (proximity to active fires, roads, and forest tenures) in combination with logistic regression and decision tree classifiers. Major forest changes of interest include wildfires, insect defoliation, forest harvesting, and flooding. Change-SAT was tested for 1998-2000 using an independent sample of change and no-change sites over Canada. Overall accuracy was 94% and commission error, especially critical for large-area change applications, was less than 1%. Regions identified as having major or widespread changes could be targeted for more detailed investigation and mapping using field visits, aerial survey, or fine resolution EO methods, such as those being applied under Canadian monitoring programs. This multi-resolution approach could be used as part of a forest monitoring system to report on carbon stocks and forest stewardship.     

Statistical models of fragmented land cover and the effect of coarse spatial resolution on the estimation of area with satellite sensor imagery

The feasibility of correcting for errors in apparent extent of land cover types on coarse spatial resolution satellite imagery was analysed using a modelling approach. The size distributions for small burn scars mapped with two Landsat Multi-spectral Scanner (MSS) images and ponds mapped with an ERS-1 synthetic aperture radar (SAR) image were measured using geographical information system (GIS) software. Regression analysis showed that these size distributions could be modelled with two types of statistical distributions a power distribution and an exponential distribution. A comparison of the size distributions of small burn scars as observed with the Landsat MSS imagery to the distribution observed with National Oceanic and Atmospheric Administration (NOAA) Advanced Very High Resolution Radiometer (AVHRR) imagery indicated that distortions due to the coarse spatial resolution of AVHRR caused overestimation of the burn area. This bias was primarily caused by detection in two or three AVHRR pixels of burns whose actual size was on the order of a single AVHRR pixel. Knowledge of the type of the actual size distribution of small fragments in a scene and the causes of distortion may lead to methods for correcting area estimates involving models of the size distribution observed with coarse imagery and requiring little or no recourse to fine scale data.     

Comparison and assessment of coarse resolution land cover maps for Northern Eurasia

Information on land cover at global and continental scales is critical for addressing a range of ecological, socioeconomic and policy questions. Global land cover maps have evolved rapidly in the last decade, but efforts to evaluate map uncertainties have been limited, especially in remote areas like Northern Eurasia. Northern Eurasia comprises a particularly diverse region covering a wide range of climate zones and ecosystems: from arctic deserts, tundra, boreal forest, and wetlands, to semi-arid steppes and the deserts of Central Asia. In this study, we assessed four of the most recent global land cover datasets: GLC-2000, GLOBCOVER, and the MODIS Collection 4 and Collection 5 Land Cover Product using cross-comparison analyses and Landsat-based reference maps distributed throughout the region. A consistent comparison of these maps was challenging because of disparities in class definitions, thematic detail, and spatial resolution. We found that the choice of sampling unit significantly influenced accuracy estimates, which indicates that comparisons of reported global map accuracies might be misleading. To minimize classification ambiguities, we devised a generalized legend based on dominant life form types (LFT) (tree, shrub, and herbaceous vegetation, barren land and water). LFT served as a necessary common denominator in the analyzed map legends, but significantly decreased the thematic detail. We found significant differences in the spatial representation of LFT's between global maps with high spatial agreement (above 0.8) concentrated in the forest belt of Northern Eurasia and low agreement (below 0.5) concentrated in the northern taiga-tundra zone, and the southern dry lands. Total pixel-level agreement between global maps and six test sites was moderate to fair (overall agreement: 0.67-0.74, Kappa: 0.41-0.52) and increased by 0.09-0.45 when only homogenous land cover types were analyzed. Low map accuracies at our tundra test site confirmed regional disagreements and difficulties of current global maps in accurately mapping shrub and herbaceous vegetation types at the biome borders of Northern Eurasia. In comparison, tree dominated vegetation classes in the forest belt of the region were accurately mapped, but were slightly overestimated (10%-20%), in all maps. Low agreement of global maps in the northern and southern vegetation transition zones of Northern Eurasia is likely to have important implications for global change research, as those areas are vulnerable to both climate and socio-economic changes.     

Super resolution land cover mapping of hyperspectral images using the deep image prior-based approach

ABSTRACT

Due to the instantaneous field-of-view (IFOV) of the sensor and diversity of land cover types, some pixels, usually named mixed pixels, contain more than one land cover type. Soft classification can predict the portion of each land cover type in mixed pixels in the absence of spatial distribution. The spatial distribution information in mixed pixels can be solved by super resolution mapping (SRM). Typically, SRM involves two steps: soft class value estimation, which is similar to the image super resolution of image restoration, and land cover allocation. A new SRM approach utilizes a deep image prior (DIP) strategy combined with a super resolution convolutional neural network (SRCNN) to estimate fine resolution fraction images for each land cover type; then, a simple and efficient classifier is used to allocate subpixel land cover types under the constraint of the generated fine fraction images. The proposed approach can use prior information of input images to update network parameters and no longer require training data. Experiments on three different cases demonstrate that the subpixel classification accuracy of the proposed DIP-based SRM approach is significantly better than the three conventional SRM approaches and a transfer learning-based neural network SRM approach. In addition, the DIP-SRM approach performs very robustly about small-area objects within multiple land cover types and significantly reduces soft classification uncertainty. The results of this paper provide an extension for utilizing SRCNN to address SRM issues in hyperspectral images.     

Global land cover classification at 1 km spatial resolution using a classification tree approach

This paper on reports the production of a 1 km spatial resolution land cover classification using data for 1992-1993 from the Advanced Very High Resolution Radiometer (AVHRR). This map will be included as an at-launch product of the Moderate Resolution Imaging Spectroradiometer (MODIS) to serve as an input for several algorithms requiring knowledge of land cover type. The methodology was derived from a similar effort to create a product at 8 km spatial resolution, where high resolution data sets were interpreted in order to derive a coarse-resolution training data set. A set of 37 294 x 1 km pixels was used within a hierarchical tree structure to classify the AVHRR data into 12 classes. The approach taken involved a hierarchy of pair-wise class trees where a logic based on vegetation form was applied until all classes were depicted. Multitemporal AVHRR metrics were used to predict class memberships. Minimum annual red reflectance, peak annual Normalized Difference Vegetation Index (NDVI), and minimum channel three brightness temperature were among the most used metrics. Depictions of forests and woodlands, and areas of mechanized agriculture are in general agreement with other sources of information, while classes such as low biomass agriculture and high-latitude broadleaf forest are not. Comparisons of the final product with regional digital land cover maps derived from high-resolution remotely sensed data reveal general agreement, except for apparently poor depictions of temperate pastures within areas of agriculture. Distinguishing between forest and non-forest was achieved with agreements ranging from 81 to 92% for these regional subsets. The agreements for all classes varied from an average of 65% when viewing all pixels to an average of 82% when viewing only those 1 km pixels consisting of greater than 90% one class within the high-resolution data sets.     

Per-pixel vs. object-based classification of urban land cover extraction using high spatial resolution imagery

In using traditional digital classification algorithms, a researcher typically encounters serious issues in identifying urban land cover classes employing high resolution data. A normal approach is to use spectral information alone and ignore spatial information and a group of pixels that need to be considered together as an object. We used QuickBird image data over a central region in the city of Phoenix, Arizona to examine if an object-based classifier can accurately identify urban classes. To demonstrate if spectral information alone is practical in urban classification, we used spectra of the selected classes from randomly selected points to examine if they can be effectively discriminated. The overall accuracy based on spectral information alone reached only about 63.33%. We employed five different classification procedures with the object-based paradigm that separates spatially and spectrally similar pixels at different scales. The classifiers to assign land covers to segmented objects used in the study include membership functions and the nearest neighbor classifier. The object-based classifier achieved a high overall accuracy (90.40%), whereas the most commonly used decision rule, namely maximum likelihood classifier, produced a lower overall accuracy (67.60%). This study demonstrates that the object-based classifier is a significantly better approach than the classical per-pixel classifiers. Further, this study reviews application of different parameters for segmentation and classification, combined use of composite and original bands, selection of different scale levels, and choice of classifiers. Strengths and weaknesses of the object-based prototype are presented and we provide suggestions to avoid or minimize uncertainties and limitations associated with the approach.     

A linearised pixel-swapping method for mapping rural linear land cover features from fine spatial resolution remotely sensed imagery

Accurate maps of rural linear land cover features, such as paths and hedgerows, would be useful to ecologists, conservation managers and land planning agencies. Such information might be used in a variety of applications (e.g., ecological, conservation and land management applications). Based on the phenomenon of spatial dependence, sub-pixel mapping techniques can be used to increase the spatial resolution of land cover maps produced from satellite sensor imagery and map such features with increased accuracy. Aerial photography with a spatial resolution of 0.25 m was acquired of the Christchurch area of Dorset, UK. The imagery was hard classified using a simple Mahalanobis distance classifier and the classification degraded to simulate land cover proportion images with spatial resolutions of 2.5 and 5 m. A simple pixel-swapping algorithm was then applied to each of the proportion images. Sub-pixels within pixels were swapped iteratively until the spatial correlation between neighbouring sub-pixels for the entire image was maximised. Visual inspection of the super-resolved output showed that prediction of the position and dimensions of hedgerows was comparable with the original imagery. The maps displayed an accuracy of 87%. To enhance the prediction of linear features within the super-resolved output, an anisotropic modelling component was added. The direction of the largest sums of proportions was calculated within a moving window at the pixel level. The orthogonal sum of proportions was used in estimating the anisotropy ratio. The direction and anisotropy ratio were then used to modify the pixel-swapping algorithm so as to increase the likelihood of creating linear features in the output map. The new linear pixel-swapping method led to an increase in the accuracy of mapping fine linear features of approximately 5% compared with the conventional pixel-swapping method.     

The use of temporal metrics for land cover change detection at coarse spatial scales

Successful land cover change analysis requires selection of an appropriate set of variables for measuring and characterizing change. Coarse spatial resolution satellite sensors offer the advantage of frequent coverage of large areas and this facilitates the monitoring of surface processes. Fine spatial resolution satellite sensors provide reliable land cover information on a local basis. This work examines the ability of several temporal change metrics to detect land cover change in sub-Saharan Africa using remote sensing data collected at a coarse spatial resolution over 16 test sites for which fine spatial resolution data are available. We model change in the fine-resolution data as a function of the coarse spatial resolution metrics without regard to the type of change. Results indicate that coarse spatial resolution temporal metrics (i) relate in a statistically significant way to aggregate changes in land cover, (ii) relate more strongly to fine spatial resolution change metrics when including a measure of surface temperature instead of a vegetation index alone, and (iii) are most effective as land cover change indicators when various metrics are combined in multivariate models.     

Object-based urban detailed land cover classification with high spatial resolution IKONOS imagery

Improvement in remote sensing techniques in spatial/spectral resolution strengthens their applicability for urban environmental study. Unfortunately, high spatial resolution imagery also increases internal variability in land cover units and can cause a ‘salt-and-pepper’ effect, resulting in decreased accuracy using pixel-based classification results. Region-based classification techniques, using an image object (IO) rather than a pixel as a classification unit, appear to hold promise as a method for overcoming this problem. Using IKONOS high spatial resolution imagery, we examined whether the IO technique could significantly improve classification accuracy compared to the pixel-based method when applied to urban land cover mapping in Tampa Bay, FL, USA. We further compared the performance of an artificial neural network (ANN) and a minimum distance classifier (MDC) in urban detailed land cover classification and evaluated whether the classification accuracy was affected by the number of extracted IO features. Our analysis methods included IKONOS image data calibration, data fusion with the pansharpening (PS) process, Hue–Intensity–Saturation (HIS) transferred indices and textural feature extraction, and feature selection using a stepwise discriminant analysis (SDA). The classification results were evaluated with visually interpreted data from high-resolution (0.3 m) digital aerial photographs. Our results indicate a statistically significant difference in classification accuracy between pixel- and object-based techniques; ANN outperforms MDC as an object-based classifier; and the use of more features (27 vs. 9 features) increases the IO classification accuracy, although the increase is statistically significant for the MDC but not for the ANN.     

Unsupervised linear contact distributions segmentation algorithm for land cover high resolution panchromatic images

Multimedia Tools and Applications - Automatic segmentation of land use and land cover from high resolution remote sensing imagery has been an essential research area in image processing for the...     

Spatial variation in land cover and choice of spatial resolution for remote sensing

Prior to acquiring remotely sensed imagery with which to map land cover investigators may wish to select an appropriate spatial resolution. Previously, statistics such as the local variance and scale variance have been used to facilitate this goal. However, where such statistics vary locally over the region of interest, their use in selecting a single spatial resolution may be undermined. The variogram and scale variance (plotted as a function of spatial resolution) were predicted for airborne multispectral imagery with a spatial resolution of 4?m of St Albans, Hertfordshire, UK and of Arundel, Sussex, UK. The remotely sensed response in the red and near-infrared wavelengths was found to vary appreciably both within and between broad land categories (such as urban, agricultural and semi-natural areas). These differences mean that where the subject of interest is a general region rather than a specific feature or object the mean local variance or scale variance over that region may be unhelpful in selecting a single spatial resolution. Further, differences observed between the red and near-infrared wavelengths have implications for users who wish to select a single spatial resolution for multispectral imagery.     

A super resolution approach for spectral unmixing of remote sensing images

It is well known that coarse spatial resolution is an important factor for the occurrence of mixed pixels in remote sensing images, and conventional approaches for spectral unmixing adopt various techniques on spectral dimension only in a fixed spatial resolution. In this article, a super resolution (SR) approach for spectral unmixing is proposed, based on the assumption that increasing the spatial resolution helps to retrieve the composition of a pixel. Firstly, a remote sensing image is downscaled into an SR image using example-based kernel ridge regression (EBKRR). Secondly, the SR image is classified using supervised hard classification, and then the class map is decomposed into thematic class layers. Thirdly, the thematic class layers are upscaled into the original spatial resolution with an averaging operation, and the abundance maps are finally derived. In two simulated data-based experiments and one ground data-based experiment, this approach was compared with linear spectral mixture analysis (LSMA) and artificial neural network (ANN)-based spectral unmixing methods. The accuracy assessment indicated that the SR approach outperformed LSMA and ANN under measurements of mean absolute error and absolute bias in the three experiments.     

Moment-based feature extraction from high spatial resolution hyperspectral images

In this article, we propose a method for extracting spatio-spectral features from high spatial resolution hyperspectral (HS) images. The method is based on extracting two-dimensional moments from neighbourhoods of pixels. Three different types of moments are considered: geometric, complex Zernike and Legendre. Moments of a given type are extracted from a few principal components (PC) of HS data, and are stacked on the original HS data to form a joint spatio-spectral feature space. These features are classified using a support vector machine (SVM) classifier. The influence of the moments orders and the size of the neighbourhood window on the quality of the extracted features are analysed. A few experiments are conducted on two widely used HS data sets, Pavia University and Salinas. The results demonstrate high capabilities of the proposed method in comparison with some state-of-the-art spatio-spectral HS classification methods.     

Land cover classification with coarse spatial resolution data to derive continuous and discrete maps for complex regions

Regularly updated land cover information at continental or national scales is a requirement for various land management applications as well as biogeochemical and climate modeling exercises. However, monitoring or updating of map products with sufficient spatial detail is currently not widely practiced due to inadequate time-series coverage for most regions of the Earth. Classifications of coarser spatial resolution data can be automatically generated on an annual or finer time scale. However, discrete land cover classifications of such data cannot sufficiently quantify land surface heterogeneity or change. This study presents a methodology for continuous and discrete land cover mapping using moderate spatial resolution time series data sets. The method automatically selects sample data from higher spatial resolution maps and generates multiple decision trees. The leaves of decision trees are interpreted considering the sample distribution of all classes yielding class membership maps, which can be used as estimates for the diversity of classes in a coarse resolution cell. Results are demonstrated for the heterogeneous, small-patch landscape of Germany and the bio-climatically varying landscape of South Africa. Results have overall classification accuracies of 80%. A sensitivity analysis of individual modules of the classification process indicates the importance of appropriately chosen features, sample data balanced among classes, and an appropriate method to combine individual classifications. The comparison of classification results over several years not only indicates the method's consistency, but also its potential to detect land cover changes.     

Accuracy assessment using sub-pixel fractional error matrices of global land cover products derived from satellite data

Information on land cover distribution at regional and global scales has become fundamental for studying global changes affecting ecological and climatic systems. The remote sensing community has responded to this increased interest by improving data quality and methodologies for extracting land cover information. However, in addition to the advantages provided by satellite products, certain limitations exist that need to be objectively quantified and clearly communicated to users so that they can make informed decisions on whether and how land cover products should be used. Accuracy assessment is the procedure used to quantify product quality. Some aspects of accuracy assessment for evaluating four global land cover maps over Canada are discussed in this paper. Attempts are made to quantify limiting factors resulting from the coarse spatial resolution of data used for generating land cover information at regional and global levels. Sub-pixel fractional error matrices are introduced as a more appropriate way for assessing the accuracy of mixed pixels. For classification with coarse spatial resolution data, limitations of the classification method produce a maximum achievable accuracy defined as the average percent fraction of dominant land cover of all pixels in the mapped area. Relationships among spatial resolution, landscape heterogeneity and thematic resolution were studied and reported. Other factors that can affect accuracy, such as misregistration and legend conversion, are also discussed.     

Effects of satellite image spatial aggregation and resolution on estimates of forest land area

Satellite imagery is being used increasingly in association with national forest inventories (NFIs) to produce maps and enhance estimates of forest attributes. We simulated several image spatial resolutions within sparsely and heavily forested study areas to assess resolution effects on estimates of forest land area, independent of other sensor characteristics. We spatially aggregated 30 m datasets to coarser spatial resolutions (90, 150, 210, 270, 510 and 990 m) and produced estimates of forest proportion for each spatial resolution using both model‐ and design‐based approaches. Average‐based aggregation had no effect on per‐image estimates of forest proportion; image variability decreased with increasing spatial resolution and local variability peaked between 210 and 270 m. Majority‐based aggregation resulted in overestimation of forest land in a heavily forested landscape and underestimation of forest land in a sparsely forested landscape, with both trends following a natural log distribution. Of the spatial resolutions tested, 30 m was superior for obtaining estimates using model‐based approaches. However, standard errors of design‐based inventory estimates of forest proportion were smallest when accompanying stratification maps which were aggregated to between 90 and 150 m spatial resolutions and strata thresholds were optimized by study area. These results suggest that spatially aggregating existing 30 m land cover datasets can provide NFIs with gains in precision of their estimates of forest land area, while reducing image storage size and processing times; land cover datasets derived from coarser spatial resolution sensors may provide similar benefits.     

Deep learning architectures for land cover classification using red and near-infrared satellite images

Multimedia Tools and Applications - Classification of remotely sensed data requires the modelling of suitable image processing algorithms. The rise of machine learning systems upgraded the...     

Validation of a large area land cover product using purpose-acquired airborne video

Large area land cover products generated from remotely sensed data are difficult to validate in a timely and cost effective manner. As a result, pre-existing data are often used for validation. Temporal, spatial, and attribute differences between the land cover product and pre-existing validation data can result in inconclusive depictions of map accuracy. This approach may therefore misrepresent the true accuracy of the land cover product, as well as the accuracy of the validation data, which is not assumed to be without error. Hence, purpose-acquired validation data is preferred; however, logistical constraints often preclude its use — especially for large area land cover products. Airborne digital video provides a cost-effective tool for collecting purpose-acquired validation data over large areas. An operational trial was conducted, involving the collection of airborne video for the validation of a 31,000 km² sub-sample of the Canadian large area Earth Observation for Sustainable Development of Forests (EOSD) land cover map (Vancouver Island, British Columbia, Canada). In this trial, one form of agreement between the EOSD product and the airborne video data was defined as a match between the mode land cover class of a 3 by 3 pixel neighbourhood surrounding the sample pixel and the primary or secondary choice of land cover for the interpreted video. This scenario produced the highest level of overall accuracy at 77% for level 4 of classification hierarchy (13 classes). The coniferous treed class, which represented 71% of Vancouver Island, had an estimated user's accuracy of 86%. Purpose acquired video was found to be a useful and cost-effective data source for validation of the EOSD land cover product. The impact of using multiple interpreters was also tested and documented. Improvements to the sampling and response designs that emerged from this trial will benefit a full-scale accuracy assessment of the EOSD product and also provides insights for other regional and global land cover mapping programs.     

An accuracy assessment experiment of the BRDF measured at coarse spatial resolution from space

The accuracy assessment of the Bidirectional Reflectance Distribution Function (BRDF) acquired from satellite sensors at coarse spatial resolution (1 or more km) is not straightforward. One possible strategy is to compare BRDF measurements from spaceborne and airborne sensors on a test area with appropriate characteristics of spatial homogeneity. This Letter describes an experiment based on this principle, aiming at the control of the BRDF measured by POLDER/ADEOS. The BRDF derived from the satellite data is compared with that derived from spatially averaged airborne POLDER data over a forested area of the Landes region. The magnitude and shape of the BRDF measured by both sensors are found to be consistent within an accuracy of 0.01 in reflectance units, except for the principal plane in the blue band where some inconsistency is found. Similar experiments could be used for the verification of other sensors data, such as those of MISR and MODIS onboard Terra, and POLDER 2 onboard ADEOS 2 in the near future.     

Study of land cover classification based on knowledge rules using high-resolution remote sensing images

This paper deals with the limitations of visual interpretation of high-resolution remote sensing images and of automatic computer classification completely dependent on spectral data. A knowledge-rule method is proposed, based on spectral features, texture features obtained from the gray-level co-occurrence matrix, and shape features. QuickBird remote sensing data were used for an experimental study of land-use classification in the combination zone between urban and suburban areas in Beijing. The results show that the deficiencies of methods where only spectral data are used for classification can be eliminated, the problem of similar spectra in multispectral images can be effectively solved for the classification of ground objects, and relatively high classification accuracy can be reached.