Charles G. O'Hara

Wavelet domain statistical hyperspectral soil texture classification

This communication presents an automatic soil texture classification system using hyperspectral soil signatures and wavelet-based statistical models. Previous soil texture classification systems are closely related to texture classification methods, where images are used for training and testing. In this study, we develop a novel system using hyperspectral soil textures, which provide rich information and intrinsic properties about soil textures, where two wavelet-domain statistical models, namely, the maximum-likelihood and hidden Markov models, are incorporated for the classification task. Experimental results show that these methods are both reliable and robust.

Multitemporal land use and land cover classification of urbanized areas within sensitive coastal environments

In the presented methodology, multitemporal Landsat images were used to develop enhanced information about complex assemblages of vegetation and patterns of seasonal land cover variability, thereby facilitating improved land use and land cover (LULC) classification of urbanized areas among sensitive environments along the Mississippi Gulf Coast. For Landsat-5 and Landsat-7 images acquired for leaf-off and leaf-on conditions for 1991 and 2000, exploratory spectral analyses and field studies were conducted to detect and analyze patterns of spectral variability in land cover observed in the multitemporal image data. Patterns were identified of seasonal spectral data changes associated with seasonal vegetation changes for known land cover and land use types, thus characterizing patterns of seasonal LULC thematic change for the area. Detected seasonal variability for known land use and land cover types were used to develop formal classification rules based upon a thematic-change logic table. An image subset area based on United States Geological Survey (USGS) 1:24000 quadrangles was used to develop a class-learning area within which unsupervised classification results were grouped into thematic classes. Signature files from the unsupervised classification results were applied to classify the balance of the study area. Individual images were classified for leaf-off and leaf-on conditions and thematic-change analyses were conducted. The formal class rules based on thematic-change logic were applied resulting in a classification that provided a level one accuracy exceeding 90% and a level two accuracy exceeding 85%.

Quality assessment of linear data

This paper presents methods to evaluate the geometric quality of spatial data. Firstly, a point‐based method is presented, adapting conventional assessment methods whereby common points between datasets are compared. In our approach, initial matches are established automatically and refined further through interactive editing. Second, a line‐based method which uses correspondences between line segments is proposed. Here, the geometry of line segments in vector is transformed into a set of rasterized values so that their combination at each pixel can restore their original vector geometry. Matching is performed on rasterized line segments and their matching lengths and displacements are measured. Experimental results show that the line‐based approach proposed is efficient to evaluate the geometric quality of spatial data without requirements of topological relationships among line features.

Detection of slump slides on earthen levees using polarimetric SAR imagery

Key results are presented of an extensive project studying the use of synthetic aperture radar (SAR) as an aid to the levee screening process. SAR sensors used are: (1) The NASA UAVSAR (Uninhabited Aerial Vehicle SAR), a fully polarimetric L-band SAR capable of sub-meter ground sample distance; and (2) The German TerraSAR-X radar satellite, also multi-polarized and featuring 1-meter GSD, but using an X-band carrier. The study area is a stretch of 230 km of levees along the lower Mississippi River. The L-band measurements can penetrate vegetation and soil somewhat, thus carrying some information on soil texture and moisture which are relevant features to identifying levee vulnerability to slump slides. While X-band does not penetrate as much, its ready availability via satellite makes multitemporal algorithms practical. Various feature types and classification algorithms were applied to the polarimetry data in the project; this paper reports the results of using the Support Vector Machine (SVM) and back-propagation Artificial Neural Network (ANN) classifiers with a combination of the polarimetric backscatter magnitudes and texture features based on the wavelet transform. Ground reference data used to assess classifier performance is based on soil moisture measurements, soil sample tests, and on site visual inspections.

Use of remote sensing to screen earthen levees

Multi-polarized L-band Synthetic Aperture Radar is investigated for its potential to screen earthen levees for weak points. Various feature detection and classification algorithms are tested for this application, including both radiometric and textural methods such as grey-level co-occurrence matrix and wavelet features.

Scale-Dependency and sensitivity of hydrological estimations to land use and topography for a coastal watershed in mississippi

This paper investigates the effect of land use and digital elevation models spatial resolution and scale on the simulation of stream flow in two coastal watersheds located in the Mississippi Gulf Coast (USA). Four elevation datasets were used: USGS DEM, NED, NASAs SRTM and IFSAR (300, 30, 30, and 5 meter resolution, respectively). Three land use datasets were included in this study: USGS GIRAS, NLCD, and NASA MODIS MOD12Q1 (400, 30, and 1000 m resolution, correspondingly). The Hydrological Program Fortran (HSPF) was used for estimating stream flow in the two watersheds. Results showed that swapping datasets in a factorial design experiment produce equivalent statistical fit of measured and simulated stream flow data. The results also showed that HSPF-estimated stream flows are not sensitive to scale and spatial resolution of the datasets included in the study.

Bridging decision‐making process and environmental needs in corridor planning

Purpose – The aim of this paper is to clarify the spatial multi‐criteria workflow for stakeholders and decision makers, for which feedback rankings are vital to the success of the transportation planning.Design/methodology/approach – The experimental approach was designed to integrate in a novel fashion both analytical hierarchy process (AHP) and multi‐criteria decision making (MCDM) within a geospatial information system (GIS) framework to deliver visual and objective tabular results useful to estimate environmental costs of the alignments generated. The method enables ranking, prioritization, selection, and refinement of preferred alternatives. The Interstate‐269, the newly planned bypass of Memphis‐TN, for which a recent environmental impact study (EIS) was completed, was selected as the experiment test‐bed.Findings – The results indicate that the approach can automate the delivery of feasible alignments that closely approximate those generated by traditional approaches. Furthermore, via integration of...

Use and analysis of temporal map algebra for vegetation index compositing

Temporal image cubes are created using co-registered temporal image data sets as ordered stacks of bands within a multi-band image. These may be manipulated and analyzed using new temporal map algebra (TMA) functions that extend normal raster map algebra from operating on a single raster band to operating on one, many, or all bands within the temporal image cube. Temporal image cubes can be constructed to encode attribute information such as image quality, scan angle, or other attribute per each pixel. Multiple cubes may be utilized to manipulate image data and generate model-specific results. Low resolution imagery such as NOAA-AVHRR and MODIS require the use maximum value compositing (MVC) that consider local pixel values in time series multi-temporal NDVI image cube. Using temporal map algebra multiple criteria may be imposed on attribute cubes to create masks cubes that can select from temporal image cubes only those specific pixels that meet scan angle, quality, or other user-defined criteria. After reducing the image data to only the desired pixels, local and focal functions may be employed to create custom composites for specific temporal intervals.

Earthen levee monitoring with Synthetic Aperture Radar

The latest results are presented from an ongoing study of the use of multi-polarized Synthetic Aperture Radar as an aid in screening earthen levees for weak points. Both L-band airborne and X-band spaceborne radars are studied, using the NASA UAVSAR and the German TerraSAR-X platforms. Feature detection and classification algorithms tested for this application include both radiometric and textural methods. Radiometric features include both the simple backscatter magnitudes of the HH, VV, and HV channels as well as decompositions such as Entropy, Anisotropy, and Alpha angle. Textural methods include grey-level co-occurrence matrix and wavelet features. Classifiers tested include Maximum Likelihood and Artificial Neural Networks. The study area includes 240 km of levees along the lower Mississippi River. Results to date are encouraging but still very preliminary and in need of further validation and testing.

Parametric Investigation of the Performance of Lidar Filters Using Different Surface Contexts

Lidar technology has provided an accurate and efficient way to obtain digital elevation models. While digital terrain models (DTMs) are essential products for three-dimensional spatial applications, extraction of ground points from a mixture of ground and non-ground points is not straightforward, and interactive classification of massive point data sets is prohibitive. To automate the filtering process, many algorithms have been proposed and demonstrated to produce satisfactory results when applied with suitably tuned parameters. For obtaining quality products using lidar filters, however, not only to figure out their optimal performance, but also to analyze the cause and effect relationships between filtering steps and their effects under variable conditions is important. Hence, this study examined the performance of three popular surface models for lidar data filtering: morphological operations, triangulation, and linear prediction. For the test, consistent setting of parameters was applied across considerably different landscape datasets. The strengths and weaknesses of the test filters were investigated by comparing the metrics of omission and commission errors and volumetric distortions, and by observing resulting DTMs and relevant surface profiles.