Wykorzystanie danych teledetekcyjnych w miejskich systemach informacyjnych

Management of urban space expects from public institutions to be active in three fields concerning spatial planning, land administration, and also urban infrastructure management. At each of these fields of activity it is indispensable to have current data about spatial distribution of objects and some additional descriptive information concerning these objects. Thus, there are created resources of information containing detailed maps (1:2000.1:10000), which are usually linked to descriptive database, as well as to some other sources of spatial data. One of the most useful sources of spatial information is photogrammetry. It supplies image data such as aerial photographs, analog or digital, and various kinds of satellite imageries. At present there is one more source of image data - commercial high resolution satellite images. They seem to be especially useful for urban space management and control. The paper presents an analysis of usefulness of QuickBird imagery as a source of information for solving decision problems of a city, and then there are pointed out some problems referring to its use for change detection changes on an urban maps. The QuickBird satellite circuits along sun-synchronous orbit at altitude of 450 km and inclination of97,2°. Imageries are made at panchromatic band of 450-900 nm, with lowest nadir resolution of 0,61 m, and also at four spectral bands with resolution of 2,44 m . blue (450-520 nm), green (520-600 nm), red (630-690 nm) and near infrared (760-900 nm). Images are delivered for public use at three variants (Basic, Standard and Ortho-rectified) in 11-bit format. The QuickBird imagery was tested in order to assess real ground resolution, geometric accuracy and a manner of presentation of objects, thus the attributes cause efficient use for urban purposes. An image of a part of Poznan was investigated, which was acquired 1-07-2003 and then transformed to a pan-sharpened standard. The valuation was done comparing to NIIRS . National Image Interpretability Rating Scales of USA, and also to the digital City Map of Poznan (at a resolution compared to a scale of 1:2000) and orthophotomap 1:2000. As a result of quality assessment it was stated, that QB imagery is located between 5 and 6 level of NIIRS, thus it can be successfully used for production of orthophotos at a scale of 1:5000, as well as for verification of the city map. It was also found, that the image is useful for evaluation of some traffic parameters. At the second part of the paper, diverse strategies of satellite image interpretation are discussed to evaluate their usefulness for updating the city information system. The simplest one is based on the visual detection of changes, and then their manual vectorization. More advanced techniques use unsupervised or supervised classification, or both of them made in sequential way. The detection process deals with common classification of images from different epochs or it is done in postclassification mode. In order to enhance the efficiency of information, change detection additional information is used, such as digital terrain or surface model (DTM or DSM), image or vector data about existing objects, LIDAR, and also knowledge and sophisticated theories (fuzzy logic, neurofuzzy) in the range of teaching the system. It is also considered theto use of a presegmentation of images (separately or together). The reported results are interesting but their practical value is limited. Summing up, we state, that monitoring of changes of urban spatial databases is commonly accepted as a basic need for purposes of spatial planning, decision-making and management. Investigation of changes can refer to such cases as: m evaluation of city growth in historical frame, with the use of old and current cartographic materials and remote sensing images, m judgment of correctness and the degree of relevance of city map, m assessment of illegal building activities (wild buildings, demolitions, over sizing recreation buildings), protection of green spaces (cut outs) and farming (land uses, parcel divisions). All of the mentioned cases can be reduced to one task: upgrading of basic cartographic data base resources. Despite of lack of satisfactory effects there is a necessity to continue investigations on the described subject.