Wietske Bijker

Context-sensitive extraction of tree crown objects in urban areas using VHR satellite images

Municipalities need accurate and updated inventories of urban vegetation in order to manage green resources and estimate their return on investment in urban forestry activities. Earlier studies have shown that semi-automatic tree detection using remote sensing is a challenging task. This study aims to develop a reproducible geographic object-based image analysis (GEOBIA) methodology to locate and delineate tree crowns in urban areas using high resolution imagery. We propose a GEOBIA approach that considers the spectral, spatial and contextual characteristics of tree objects in the urban space. The study presents classification rules that exploit object features at multiple segmentation scales modifying the labeling and shape of image-objects. The GEOBIA methodology was implemented on QuickBird images acquired over the cities of Enschede and Delft (The Netherlands), resulting in an identification rate of 70% and 82% respectively. False negative errors concentrated on small trees and false positive errors in private gardens. The quality of crown boundaries was acceptable, with an overall delineation error <0.24 outside of gardens and backyards.

Image Mining for Modeling of Forest Fires From Meteosat Images

Meteosat satellites with the Spinning Enhanced Visible and Infrared Imagery (SEVIRI) sensor onboard provide remote-sensing images nowadays every 15 min. This paper investigates and applies image-mining methods to make an optimal use of images. It develops a simple, time-efficient, and generic model to facilitate pattern discovery and analysis. The focus of this paper is to develop a model for monitoring and analyzing forest fires in space and time. As an illustration, a diurnal cycle of fire in Portugal on July 28, 2004 was analyzed. Kernel convolution characterized the hearth of the fire as an object in space. Objects were extracted and tracked over time automatically. The results thus obtained were used to make a linear model for fire behavior with respect to vegetation and wind characteristics as explanatory variables. This model may be useful for predicting hazards at an almost real-time basis. The research illustrates how image mining improves information extraction from the Meteosat SEVIRI images

Quality Aspects in Spatial Data Mining

Describes the State-of-the-Art in Spatial Data Mining, Focuses on Data Quality Substantial progress has been made toward developing effective techniques for spatial information processing in recent years. This science deals with models of reality in a GIS, however, and not with reality itself. Therefore, spatial information processes are often imprecise, allowing for much interpretation of abstract figures and data. Quality Aspects in Spatial Data Mining introduces practical and theoretical solutions for making sense of the often chaotic and overwhelming amount of concrete data available to researchers. In this cohesive collection of peer-reviewed chapters, field authorities present the latest field advancements and cover such essential areas as data acquisition, geoinformation theory, spatial statistics, and dissemination. Each chapter debuts with an editorial preview of each topic from a conceptual, applied, and methodological point of view, making it easier for researchers to judge which information is most beneficial to their work. Chapters Evolve From Error Propagation and Spatial Statistics to Address Relevant Applications The book advises the use of granular computing as a means of circumventing spatial complexities. This counter-application to traditional computing allows for the calculation of imprecise probabilities the kind of information that the spatial information systems community wrestles with much of the time.Under the editorial guidance of internationally respected geoinformatics experts, this indispensable volume addresses quality aspects in the entire spatial data mining process, from data acquisition to end user. It also alleviates what is often field researchers most daunting task by organizing the wealth of concrete spatial data available into one convenient source, thereby advancing the frontiers of spatial information systems.

Aboveground biomass estimation of tropical forest from Envisat advanced synthetic aperture radar data using modeling approach

Abstract. Retrieval of the aboveground forest biomass (AGB), especially in high biomass forests ( > 100     t / ha ), remains a challenging task for the researchers worldwide. The retrieval of AGB over a tropical forest area in India using Envisat advanced synthetic aperture radar C-band backscatter, interferometric synthetic aperture radar (InSAR) coherence and semi-empirical models viz., water cloud model (WCM) and interferometric water cloud model (IWCM), is studied. In process, the model parameters, i.e., backscatter from vegetation and ground, two-way tree transmissivity, and coherence from vegetation and ground were retrieved. The model training procedure to retrieve the model parameters consisted of an iterative regression of WCM and IWCM. High AGB accuracy ( R 2 = 0.73 ) with low root mean square error ( RMSE = 53.76     t / ha ) was achieved through multidate weighted averaging using RMSE-based weighting coefficients and WCM. Multidate data and InSAR coherence images showed better results ( R 2 = 0.90 , RMSE = 35.92     t / ha ) compared to individual coherence images. The InSAR coherence was found to be better for AGB retrieval than SAR backscatter as the former did not saturate for high AGB values.

Region-based urban road extraction from VHR satellite images using Binary Partition Tree

Abstract This paper provides a hierarchical method for urban road extraction. It consists of (1) obtaining the road region of interest from a VHR image, (2) hierarchically representing this road region of interest in a Binary Partition Tree (BPT), and extracting the roads based on this BPT at hierarchical levels. Besides using two existing geometrical features (i.e. compactness and elongation), we define two other structural features based on orientation histograms and morphological profiles to guide the region merging of BPT. The morphological profiles are constructed using a series of path openings, which facilitate modeling linear or curved structures. The proposed method was applied to two types of VHR images with different urban settings, corresponding to a Pleiades-B image of Wuhan, China, and a Quickbird image of Enschede, the Netherlands. Experimental results show that the proposed method was able to group adjacent small segments that have high spectral heterogeneity and low road-like geometrical properties to form more meaningful roads sections, and performed superior to the existing methods. Furthermore, we compared the proposed method with two other existing methods in the literature. We conclude that the proposed method can provide an effective means for extracting roads over densely populated urban areas from VHR satellite images.

Image mining for drought monitoring in eastern Africa using Meteosat SEVIRI data

We propose an image mining approach to monitor drought using Meteosat Spinning Enhanced Visible and InfraRed Imager (SEVIRI) image data. SEVIRI image data provide frequent Normalized Difference Vegetation Index (NDVI) time series which are important to assess the evolution of drought conditions. Vegetation condition is characterized in space by the deviation of the current NDVI observations at locations from their temporal mean values. In this paper we assume a gradual evolution of vegetation stress caused by drought and hence address this aspect with the use of a membership function applied to vegetation stress values to model drought. Our approach is implemented on subset image data of eastern Africa. Vegetated sites in a drought prone area of the region serve as an illustration using the drought spell at the end of 2005. This study shows that the use of a membership function allows capturing the gradual evolution of drought and can be used to model drought from observed vegetation conditions.

Fuzzy Classification for Shoreline Change Monitoring in a Part of the Northern Coastal Area of Java, Indonesia

This study presents an unsupervised fuzzy c-means classification (FCM) to observe the shoreline positions. We combined crisp and fuzzy methods for change detection. We addressed two perspectives of uncertainty: (1) uncertainty that is inherent to shoreline positions as observed from remote sensing images due to its continuous variation over time; and (2) the uncertainty of the change results propagating from object extraction and implementation of shoreline change detection method. Unsupervised FCM achieved the highest kappa (κ) value when threshold (t) was set at 0.5. The highest κ values were 0.96 for the 1994 image. For images in 2013, 2014 and 2015, the κ values were 0.95. Further, images in 2003, 2002 and 2000 obtained 0.93, 0.90 and 0.86, respectively. Gradual and abrupt changes were observed, as well as a measure of change uncertainty for the observed objects at the pixel level. These could be associated with inundations from 1994 to 2015 at the northern coastal area of Java, Indonesia. The largest coastal inundations in terms of area occurred between 1994 and 2000, when 739 ha changed from non-water and shoreline to water and in 2003–2013 for 200 ha. Changes from water and shoreline to non-water occurred between 2000 and 2002 (186 ha) and in 2013–2014 (65 ha). Urban development in flood-prone areas resulted in an increase of flood hazards including inundation and erosion leading to the changes of shoreline position. The proposed methods provided an effective way to present shoreline as a line and as a margin with fuzzy boundary and its associated change uncertainty. Shoreline mapping and monitoring is crucial to understand the spatial distribution of coastal inundation including its trend.

Angular Backscatter Variation in L-Band ALOS ScanSAR Images of Tropical Forest Areas

Scanning synthetic aperture radar (ScanSAR) systems provide continuous information over large areas, but for effective use of such products in tropical forest, the decrease of radar backscatter with large variation of incidence angles requires attention. This letter analyzes the dependence of radar backscatter on incidence angle for L-band ScanSAR images of tropical forest. We investigated and modeled the angular backscatter effect per land-cover class in three ScanSAR images of the Colombian Orinoco. We found that there is an evident effect of incidence angle on radar backscatter, depending on land-cover class, moisture content, and physical structure of the reflecting targets. To normalize the angular backscatter variation, we proposed two methods. The first one applies a cosine correction estimated through linear regression. The second one models the radar backscatter of flooded forest considering second-order signal interactions. The model explains the observed backscatter of flooded forest areas in the rainy season (R2 that is larger than 0.77).

Pattern validation for MODIS image mining of burned area objects

An image mining method was applied to Moderate Resolution Imaging Spectroradiometer (MODIS) satellite data to estimate the area burned by forest fires occurring in Galicia (Spain) between 4 August and 15 August 2006. Five different inputs were considered: post-fire near-infrared reflectance (NIR) band, post-fire Normalized Difference Vegetation Index (NDVI) image, pre-fire and post-fire NDVI difference image and 4-μm and 11-μm thermal bands. The proposed image mining method consists of three steps: a pre-classification step, applying kernel smoothing, based on the fast Fourier transform (FFT), a modelling step applying Gaussian distributions on individual grid cells with deviating values, and a thresholding step classifying the model into burned and unburned classes. Polygons collected in the field with Global Positioning System (GPS) measurements from a helicopter permitted validation of the burned area estimation. A Z-test based on the κ statistic compared the accuracy of this estimation with the accuracies achieved by traditional methods based both on spectral changes in reflectance after the fire and active fire detection. The results showed a significant improvement (7.5%) in the accuracy of the burned area estimation after kernel smoothing. Burned area estimation based on the smoothed difference image between pre-fire and post-fire NDVI image had the highest accuracy (κ = 0.72). We conclude that the image mining algorithm successfully extracted burned area objects and that extraction was best when smoothing was applied prior to classification. Image mining methods based on using the κ statistic thus provide a valuable validation procedure when selecting the optimal MODIS input image for estimating burned area objects.

Change vector analysis to monitor the changes in fuzzy shorelines

Mapping of shorelines and monitoring of their changes is challenging due to the large variation in shoreline position related to seasonal and tidal patterns. This study focused on a flood-prone area in the north of Java. We show the possibility of using fuzzy-crisp objects to derive shoreline positions as the transition zone between the classes water and non-water. Fuzzy c-means classification (FCM) was used to estimate the membership of pixels to these classes. A transition zone between the classes represents the shoreline, and its spatial extent was estimated using fuzzy-crisp objects. In change vector analysis (CVA) applied to water membership of successive shorelines, a change category was defined if the change magnitude between two years, T1 and T2, differed from zero, while zero magnitude corresponded to no-change category. Over several years, overall change magnitude and change directions of the shoreline allowed us to identify the trend of the fluctuating shoreline and the uncertainty distribution. The fuzzy error matrix (FERM) showed overall accuracies between 0.84 and 0.91. Multi-year patterns of water membership changes could indicate coastal processes such as: (a) high change direction and high change magnitude with a consistent positive direction probably corresponding to land subsidence and coastal inundation, while a consistent negative direction probably indicates a success in a shoreline protection scheme; (b) low change direction and high change magnitude indicating an abrupt change which may result from spring tides, extreme waves and winds; (c) high change direction and low change magnitude which could be due to cyclical tides and coastal processes; and (d) low change direction and low change magnitude probably indicating an undisturbed environment, such as changes in water turbidity or changes in soil moisture. The proposed method provided a way to analyze changes of shorelines as fuzzy objects and could be well-suited to apply to coastal areas around the globe.