Isabelle Couloigner

A new and efficient k-medoid algorithm for spatial clustering

A new k-medoids algorithm is presented for spatial clustering in large applications. The new algorithm utilizes the TIN of medoids to facilitate local computation when searching for the optimal medoids. It is more efficient than most existing k-medoids methods while retaining the exact the same clustering quality of the basic k-medoids algorithm. The application of the new algorithm to road network extraction from classified imagery is also discussed and the preliminary results are encouraging.

Benefit of the angular texture signature for the separation of parking lots and roads on high resolution multi-spectral imagery

The misclassification of roads and parking lots is one of the major difficulties in automating road network extraction from high resolution remotely-sensed imagery, especially in urban areas. This paper proposes a new integrated approach to road identification on high resolution multi-spectral imagery. The input images are first segmented using a traditional k-means clustering on normalized digital numbers. The road cluster is then automatically identified using a fuzzy logic classifier. A number of shape descriptors of angular texture signature are introduced for a road class refinement, i.e. to separate the roads from the parking lots that have been misclassified as roads. Intensive experiments have shown that the proposed methodology is effective in automating the separation of roads from parking lots on high resolution multi-spectral imagery.

Automatic road change detection and GIS updating from high spatial remotely-sensed imagery

This paper presents a framework for road network change detection in order to update the Canadian National Topographic DataBase (NTDB). The methodology has been developed on the basis of road extraction from IRS-pan images (with a 5.8 m spatial resolution) by using a wavelet approach. The feature matching and conflation techniques are used to road change detection and updating. Elementary experiments have showed that the proposed framework could be used for developing an operational road database updating system.

Geographic Object-Based Mosaicing (OBM) of High-Resolution Thermal Airborne Imagery (TABI-1800) to Improve the Interpretation of Urban Image Objects

As part of the Heat Energy Assessment Technologies (HEAT) project, we describe a novel geographic object-based mosaicing algorithm referred to as Object-Based Mosaicing (OBM) that joins thermal airborne flight lines around urban roof objects rather than bisecting them with arbitrary mosaic join lines. An OBM mosaic is compared with a “traditional” mosaic product (created in ENVI 4.8) consisting of 44 TABI-1800 flight lines of the City of Calgary, Alberta, Canada (825 km2). Compared with the traditional mosaic, OBM results in the following: 1) visually improved roof shapes within the scene; 2) reduced processing time (up to 50 %); 3) more accurate hot-spot detection; and 4) a better data set for more accurate home energy models-as the thermal imagery for each roof are from a single acquisition time. Conversely, without applying OBM to the full scene, 14 209 homes are bisected within the traditional mosaic product.

Comparing Different Localization Approaches of the Radon Transform for Road Centerline Extraction from Classified Satellite Imagery

Using a local Radon transform helps improve the performance of the Radon transform-based linear feature detection. In this paper, three different approaches to localize the Radon transform are implemented and compared in the context of road centerline extraction from classified satellite imagery

Iterative and localized radon transform for road centerline detection from classified imagery

An iterative and localized Radon transform is proposed in this paper for the specific application of road network extraction from high resolution satellite imagery. Based on an accurate estimation of the line width and line parameters in the radon space, the localized Radon transform makes it possible to detect the small road segments and the long curvilinear lines, which is a difficult task in road detection. Experiments on both synthetic and real-world imagery have shown that the proposed methodology is effective in detecting road centerlines from classified imagery.

An Assessment of Polynomial Regression Techniques for the Relative Radiometric Normalization (RRN) of High-Resolution Multi-Temporal Airborne Thermal Infrared (TIR) Imagery

Thermal Infrared (TIR) remote sensing images of urban environments are increasingly available from airborne and satellite platforms. However, limited access to high-spatial resolution (H-res: ~1 m) TIR satellite images requires the use of TIR airborne sensors for mapping large complex urban surfaces, especially at micro-scales. A critical limitation of such H-res mapping is the need to acquire a large scene composed of multiple flight lines and mosaic them together. This results in the same scene components (e.g., roads, buildings, green space and water) exhibiting different temperatures in different flight lines. To mitigate these effects, linear relative radiometric normalization (RRN) techniques are often applied. However, the Earth’s surface is composed of features whose thermal behaviour is characterized by complexity and non-linearity. Therefore, we hypothesize that non-linear RRN techniques should demonstrate increased radiometric agreement over similar linear techniques. To test this hypothesis, this paper evaluates four (linear and non-linear) RRN techniques, including: (i) histogram matching (HM); (ii) pseudo-invariant feature-based polynomial regression (PIF_Poly); (iii) no-change stratified random sample-based linear regression (NCSRS_Lin); and (iv) no-change stratified random sample-based polynomial regression (NCSRS_Poly); two of which (ii and iv) are newly proposed non-linear techniques. When applied over two adjacent flight lines (~70 km2) of TABI-1800 airborne data, visual and statistical results show that both new non-linear techniques improved radiometric agreement over the previously evaluated linear techniques, with the new fully-automated method, NCSRS-based polynomial regression, providing the highest improvement in radiometric agreement between the master and the slave images, at ~56%. This is ~5% higher than the best previously evaluated linear technique (NCSRS-based linear regression).

Transforming Image-Objects into Multiscale Fields: A GEOBIA Approach to Mitigate Urban Microclimatic Variability within H-Res Thermal Infrared Airborne Flight-Lines

In an effort to minimize complex urban microclimatic variability within high-resolution (H-Res) airborne thermal infrared (TIR) flight-lines, we describe the Thermal Urban Road Normalization (TURN) algorithm, which is based on the idea of pseudo invariant features. By assuming a homogeneous road temperature within a TIR scene, we hypothesize that any variation observed in road temperature is the effect of local microclimatic variability. To model microclimatic variability, we define a road-object class (Road), compute the within-Road temperature variability, sample it at different spatial intervals (i.e., 10, 20, 50, and 100 m) then interpolate samples over each flight-line to create an object-weighted variable temperature field (a TURN-surface). The optimal TURN-surface is then subtracted from the original TIR image, essentially creating a microclimate-free scene. Results at different sampling intervals are assessed based on their: (i) ability to visually and statistically reduce overall scene variability and (ii) computation speed. TURN is evaluated on three non-adjacent TABI-1800 flight-lines (~182 km2) that were acquired in 2012 at night over The City of Calgary, Alberta, Canada. TURN also meets a recent GEOBIA (Geospatial Object Based Image Analysis) challenge by incorporating existing GIS vector objects within the GEOBIA workflow, rather than relying exclusively on segmentation methods.