Context and objectives
The A.L.E. company (Association Liégeoise d'Électricité - user partner) requires large scale maps for its daily activities relating to the power supply and the management of the power network in the region of Liège.
Digital maps cover one square kilometre on the field and are georeferenced to the Belgian Lambert 72 system. Paper maps can be drawn from digital data at different scales, from 1/500 to1/2 000. The most important objects in the maps concern various kinds of buildings, and the quality assessment relies upon several measures and indices such as: the rate of correctly identified objects, planimetric accuracy thresholds and a completeness index. It is obvious that field surveys and photogrammetric restitution are necessary to fit the accuracy requirements of the digital map. However the updating process only resorts to these costly techniques after that significant changes have been identified by aerial photo interpretation.
The aim of this feasibility study is to assess the potentials of very high resolution (VHR) remotely sensed imagery (~1 m) to correctly identify and locate the different kinds of buildings . The project will make use of CAPI -like techniques applied upon geometrically corrected satellite images. The geometric correction process is of first importance for planimetric accuracy, so this topic will be carefully examined in the first stage of the project. It is expected to fusion panchromatic and multispectral data (from the same sensor/satellite) to generate colour composites if they could improve interpretation. Genuine fusion algorithm developed by SURFACES would be used for this task. The three major quality indices defined in the terms of reference, concerning identification, accuracy and completeness, will be used too to check the quality of the satellite image analysis. Reference data constituted by existing digital maps covering selected regions of interest will be provided by the user partner. The capability of satellite image to be an aerial photo surrogate to detect and to locate changes of the built-up area would be considered as the main positive result. Moreover, according to the obtained values of the quality indices, satellite image analysis could be integrated in the assessment process of the field / photogrammetric works, for all or some categories of buildings. The consequences of the introduction of satellite remote sensing techniques in the elaboration (updating / assessment) of the digital maps will be considered, notably in terms of time and costs.
|Project leader(s):||ULg - Unité de Géomatique|