The objective of the present research effort was the investigation of expert system classification techniques for land use mapping from very high resolution images for a typical Greek landscape. Data used included an IKONOS image of the Arkadi area in Crete acquired on September 2000, and a digital terrain model. Photointerpretation was carried out using color composites, band ratios and maps of scale 1:5.000 and 1:50.000. Maximum likelihood was used for per pixel supervised classification and its accuracy was 72%. A knowledge base containing 51 rules, 44 hypotheses and 12 variables was developed in the Expert Classifier module of ERDAS Imagine. A hierarchical organization of thematic classes was developed in four levels through photointerpretation and study of the spectral reflectance diagrams and thematic class histograms. The image was first classified into three general categories: water-like, vegetation-like and soil-like materials. These were then separated into sub-classes. Classification rules were enriched with ancillary data such as the slopes, the road network, the NDVI vegetation index, the results of a spatial model computing texture, and indices reflecting the polygon shape and perimeter. Overall accuracy of the classification with the expert system was 82%.
Urban green is recognized as an important functional element of the city, which affects directly the standard of living. The present paper is concerned with the study of urban green by means of object-oriented image analysis of high-resolution IKONOS data. More specifically, the potential for detecting urban green and quantitatively assessing it was explored. The analysis included two levels of segmentation and classification. On the first level, objects to which the image was segmented were subsequently classified according to a vegetation index (Scaled MSAVI) to areas with dense, thin or no vegetation. On the second level the image was classified in larger areas that simulated building blocks according to the relative area of vegetation, in order to create a thematic map of urban green density. The evaluation of the results indicated that detection and quantitative assessment of urban green was achieved with satisfactory accuracy. The use of additional data (DEM, hyperspectral, GIS) will allow a more detail study of the urban green from high resolution data by means of object-oriented image analysis
The aim of the present work was to implement and evaluate spatial filtering and automatic edge extraction techniques for assisting the geological lineament detection process. The selected study area was Alevrada in Central Greece, an area of sedimentary terrain with many faults and folds. A Landsat-7 ETM+ image of the study area was geometrically registered on the geological map, and then radiometrically corrected, to subtract the path radiance of the optical bands. Various linear and nonlinear spatial high pass operators (Laplacian, Ford, directional filters, Sobel, Kirsch) were applied and an interpretation of the lineaments was made. Certain edge detection algorithms introduced in medical imaging and scene analysis were applied and assessed, including the Canny multi-scale edge detector, the Rothwell edge detector based on edge topology and the Black's anisotropic diffusion edge detector, followed by morphological cleaning and pruning processes. The interpreted lineaments were qualitatively compared to the edge maps derived from the edge extraction algorithms, and a satisfactory matching was observed. This work provides a preliminary step towards lineament mapping automation.
The objective of this research was the investigation of advanced image analysis methods for geomorphological mapping. Methods employed included multiresolution segmentation of the Digital Elevation Model (DEM) GTOPO30 and fuzzy knowledge based classification of the segmented DEM into three geomorphological classes: mountain ranges, piedmonts and basins. The study area was a segment of the Basin and Range Physiographic Province in Nevada, USA. The implementation was made in eCognition. In particular, the segmentation of GTOPO30 resulted into primitive objects. The knowledge-based classification of the primitive objects based on their elevation and shape parameters, resulted in the extraction of the geomorphological features. The resulted boundaries in comparison to those by previous studies were found satisfactory. It is concluded that geomorphological feature extraction can be carried out through fuzzy knowledge based classification as implemented in eCognition.
This paper describes the experience gained from the evaluation of selected automatic edge detection techniques applied to LANDSAT TM, SPOT HRV, IRS 1C and IKONOS images. Emphasis was given to the detection of man-made objects and linear features such as coastlines, roads and parcel boundaries in combination with selected preprocessing and postprocessing operations. As preprocessing Gaussian, adaptive and morphological operators were implemented and tested for image enhancement and smoothing. Edge extraction processing followed. First the Canny edge detector was applied. Then a morphological nonlinear Laplacian operator was applied and its zero-crossings yielded edge locations. Finally an edge detector resulting by overlaying two thresholded images from the Prewitt gradient, preserving edges appearing in both images, was applied. Postprocessing followed to eliminate noisy edges and restore edge connectivity through morphological operators. An analysis of the relative performance of the processing scheme indicated each detector's relation to noise (features at certain undesired scales, shadows along roads boundaries, irrelevant edges within parcel boundaries) and the set of specific parameters needed for proper enhancement and smoothing before edge extraction.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
INSTITUTIONAL Select your institution to access the SPIE Digital Library.
PERSONAL Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.