François Cavayas

Estimating forest stand structure from high resolution imagery using the directional variogram

Abstract The height and stocking of forest stands can be estimated with relatively high precision using an empirical model relating parameters extracted from the directional variogram of high resolution images and forest structure parameters. A geometrical-optical model of the forest was first used to generate images of artificial forest stands in order to establish the relation between tree size. tree density and image texture. The resulting equations were then applied on the computer generated images as well as on high resolution MEIS II images to predict the forest structure values. The results show a good concordance between actual and predicted values, even when spatial resolution was degraded from 0·36m to 2·16m.

A comparative analysis of scanned maps and imagery for mapping applications

Abstract In mapping organizations, the implementation of more automation coupled with the availability of heterogeneous data requires the investigation, adaptation and evaluation of new approaches and techniques. The demand for rapid mapping operations such as database generation and updating is continuously increasing. Due to the rising use of raster data, image analysis techniques have been investigated and tested in this study to introduce automation in the assessment of scanned topographic monochrome maps and Landsat 7 ETM+ imagery for feature separation and extraction in northern Canada. The work focuses on the detection and extraction of lakes—predominant features in the North—as well as on to their spatio-temporal comparison. Various approaches using digital image processing techniques were implemented and evaluated. Thresholding and texture measures were used to evaluate the potential of rapid extraction of certain topographic elements from scanned monochrome maps of northern Canada. A raster to vector approach (R→V) followed for the vectorization of these extracted features. The extraction of features from Landsat 7 ETM+ imagery involved image and theme enhancement by applying various image fusion and spectral transformations (e.g., Brovey, PCI-IMGFUSE, intensity–hue–saturation (IHS), principal component analysis (PCA), Tasseled Cap, Normalized Difference Vegetation Index (NDVI)), followed by image classification and thresholding. Tests showed that the approaches were more or less feature-dependent, while, at the same time, they can augment and significantly enhance the conventional topographic mapping methods. Following the analysis of the map and image data, change detection between two lake datasets was performed both interactively and in an automated mode based on the non-intersection of old and new features. The various approaches and methodology developed and implemented within a GIS environment along with examples, results and limitations are presented and discussed.

Map-image matching using a multi-layer perceptron: the case of the road network

Abstract To help automatize map revision at a scale of 1 : 50,000, a map-guided method is described to update the road network of a map database. This paper describes the essential first step of the procedure, which consists of matching the roads present on both the image and the map database. This matching has to be performed precisely in order to generate meaningful hypotheses on the location of new roads. The matching is conducted by using a multi-layer perceptron (MLP) trained to recognize road segments on the SPOT-HRV panchromatic image corresponding to the cartographic database being treated. Two template matching methods using the trained MLP weight matrix are developed. The first method locates all the road intersections on the image, while the second method locates the segments only. Both methods are not accurate enough to be used alone. However, combining both approaches gives results that are reliable enough to be used in the generation of the hypotheses needed to extract new roads.

Polarimetric Radarsat-2 wetland classification using the Touzi decomposition: case of the Lac Saint-Pierre Ramsar wetland

Wetlands play a key role in regional and global environments and are linked to climate change, water quality, and hydrological and carbon cycles. They also contribute to wildlife habitat and biodiversity and can act as indicators of overall environmental health. Unfortunately, wetlands continue to be under threat. There is an immediate need for improved mapping and monitoring of wetlands to better manage and protect these sensitive areas. Recently, the Touzi decomposition was introduced and proved very promising for wetland characterization using polarimetric airborne (Convair-580) SAR data. The purpose of this study is to assess the Touzi incoherent target-scattering decomposition (ICTD) for wetland classification using polarimetric Radarsat-2 (RS2) data collected over the RAMSAR wetland site in Lac Saint-Pierre, Canada. In particular, the sensitivity of the ICTD parameters to seasonal evolution of marsh and swamp scattering is discussed and demonstrated. The intent is to show that the dominant scattering type magnitude (αs1) and phase (Φs1), and the dominant (η1) and lowest scattering eigenvalues (η3), lead to an effective characterization of the various backscattering mechanisms of the wetland plant species. The ICTD parameters form the basis of a new hierarchical classification that is efficient for wetland classification. The use of multitemporal information obtained from multidate RS2 acquisitions between April and September allows an accurate wetland classification. The RS2 polarimetric classification is then compared with a supervised maximum-likelihood classification using a pair of Landsat-5 images.

Automatic road extraction using fuzzy mask concepts

In this paper, fuzzy concepts are used in order to realize road pixels extraction. The aim of the proposed algorithms is to attribute a road membership value to each pixel. A set of twelve fuzzy sets representing basic 2D road structures is first defined. The road structure, defined as a logical concatenation of these elementary predefined 2D structures, is then used in order to enhance the road membership values. The proposed approach is applied using a georeferenced SPOT panchromatic image acquired on August 1989 and representing the municipality of Charlesbourg, Quebec, Canada. Obtained results are encouraging in terms of detected road validity as well as in terms of noise resistance.

Mapping urban vegetation cover using WorldView-2 imagery

There are clear indications that densification of built-up areas within cities and new developments in their outskirts, in conjunction with urban population activities, are at the origin of climate changes at the local level and have a direct impact on air and water quality. Densification of the vegetation cover is often mentioned as one of the most important means to mitigate the impacts of climate changes and to improve the quality of the urban environment. Decision making on vegetation cover densification presupposes that urban planners and managers know exactly the actual situation in terms of vegetation location, types and biomass. However, in many cities, inventories of vegetation cover are usually absent. This study examines the feasibility of an automatic system for vegetation cover inventory and mapping in urban areas based on WorldView-2 imagery. The city of Laval, Canada, was chosen as the experimental site. The principal conclusions are as follows: a) conversion of digital counts to ground reflectances is a crucial step in order to fully exploit the potential of WV-2 multispectral images for mapping vegetation cover and recognizing vegetation classes; b) the combined use of NDVIs computed using the three infrared available bands and the red band provides an accurate means of differentiating vegetation cover from other land covers; and c) it is possible to separate trees from other vegetation types and to identify tree species even in dense urban areas using spectral signature characteristics and segmentation algorithms.

An automatic road extraction method using a map-guided approach combined with neural networks for cartographic database validation purposes

A method is proposed to extract road intersections from a SPOT panchromatic image, using a map-guided approach combined with the application of a neural network. The results show an average increase of 36% of planimetric accuracy after applying the method instead of simply superimposing the roads on the geocoded image. Also, only 8 out 42 samples were previously correctly traced, compared to 27 after application of the algorithm.

Texture-based segmentation of temperate-zone woodland in panchromatic IKONOS imagery

We have performed a study to identify optimal texture parameters for woodland segmentation in a highly non-homogeneous urban area from a temperate-zone panchromatic IKONOS image. Texture images are produced with the sum- and difference-histograms depend on two parameters: window size f and displacement step p. The four texture features yielding the best discrimination between classes are the mean, contrast, correlation and standard deviation. The f-p combinations 17-1, 17-2, 35-1 and 35-2 are those which give the best performance, with an average classification rate of 90%.

VERS UNE NOUVELLE MÉTHODE D'INVENTAIRE ET DE MISE À JOUR DE L'OCCUPATION/UTILISATION DU SOL EN MILIEU URBAIN

SUMMARYA new inventory method for land use in an urban environment is suggested. It is based on an analysis of satellite images divided according to functional islands extracted from the urban fabric. These islands are determined from a digitized map of the road network that has been registered with the images. A series of radiometric and geometric parameters of the islands is extracted and analysed through an image recognition system. The latter is based on interpretation and a priori knowledge rules of the environment under study in order to assign a label to the island. Accuracies resulting from applying this methodology for the identification of urban functions using panchromatic HRV-SPOT images collected in 1987 and 1989 over the Island of Montreal were 81% and 86%, respectively.