| Digitalization and informationalization are dominant tide and trend for high-tech development at home and abroad nowadays. With technical development of geological observation, e-map, Internet, and virtual reality and construction of IT infrastructure of state, concepts of digital earth, digital city and digital oil field have gradually come out and been put into implementation. In 2003, Tianjin City Survey Institute and Jilin Provincial No. 2 Survey Institute took QuickBird satellite data as the data source to produce the 1:5000 scale e-map for No. 8 Oil Extraction Plant of Daqing Oil Field. We set up basic geological information data and surface facilities of information and image database such as building structure, etc., using 3D technology as the main channel to obtain the information, taking 2002's high-precision full color US QuickBird as the main source of information, and accompanying with existing vector data; we adopted ERDAS8.6,Photoshop7.0 and other image processing software to process image, and for information collection, based on comprehensive survey and mapping, we adopted digital photographing and survey method as well as AutoCAD and Microstation to carry out digitalized survey and mapping, extract the information from image in full color range, and carry out actual survey with coordination of local surface features, omni-directional station instrument and high precision GPS; and for other data collections, they made full use of their existing maps, charts and data to realize digitalization and vectorialization by means of scanner and keyboard, etc. The 1:5000 data production and collection for Daqing Oil Field cover all data about surface and underground infrastructure such as underground water, electricity, air and pipeline systems, surface building structure and other facilities, etc., including image e-maps, vector e-maps and some files. 13 views of QuickBird Satellite Data are provided, and this project employed 287 pieces of 1:5000 image e-maps, and 270 pieces of vector e-maps, and the later was divided into two parts of the new graphic and modification to existing graphic. We selected control point for collection of full range image for each view, obtained the geodetic coordinates through field GPS and The Routine Total stations surveying instrument, and then, made correction, immerge, subset image and collection. The Routine Total stations instrument and high-precision GPS were used to obtain surface control point required for orientation and positioning for digitalized survey and mapping, where, all GPS points were corrected and positioned through known points. Since the image e-maps is to be used for data collection of the 1:5000 vector e-maps, the precision must be very high, and digital elevation model (DEM) will be also required during the process of correction, we used the existing 1:10000 digital line graphics to draw contour line, and then, used the contour line and equal height point collected on the field to generate DEM. Firstly, we corrected the QuickBird full color range data, and rectified multispectral color image. The method of correction is to use special-purpose QuickBird correction module, use RPC parameters and DEM to carry out correction of the full range image. Before immerge of full range color and multispectral images, the multispectral images must be well rectified, however, multispectral rectifying is based on well-collected full range image. Therefore, high precision rectifying can be carried out to multispectral images by means of RPS parameters and full-range images. There are many methods for immerge of full range color and multispectral images. In view of nature of this project and the requirements (production of 1:5000 vector maps), higher precision will be expected. Therefore, we used the main components change method and weighted factor method to carry out immerge. After immerge, 4,3,2,1 and the full range 5 bands were remained, and information flow and resolution of images were further increased. Because Daqing required real color image maps, we eliminated band 4, and had bands 3,2,1 left, while we also maintained images of thefour bands. The processed typical 1:5000 maps, plus boundary, became the 8bit typical 1:5000 TIF image. Digital products of this project are: Digital linear graphs, digital orthophoto models, field control point pictures, field control points achievements, and field survey and mapping pictures. This thesis describes system structure, general design concept and system itself of the digital oil field project, and main work is to use remote sensing technology to process Quick bird images. Key point of this project is to guarantee high modification precision, high precision extraction of information about surface features, and processing of image tune. For reasons of surface features and photographing time of images, quality of some pictures are not so good, which may affect modification precision, correct extraction of information about surface features and processing of tune. Quality defects of original pictures: (1) Rural villages densely distributed look pale or yellow in the multispectral images. (2) Saline-alkali soil from oil production in big area looks pale in the images. (3) Due to influence of black earth and water deposit, the farm field looks black. The quality of original image is not so good to some extent, therefore, the resolution and tune processing of the graphics are affected. For this project, we have provided 6 pieces of maps, including 2 pieces of modified maps, and 4 pieces of new maps. After completion of 1:5000 maps, we printed the maps, respectively, and submitted them for check in the field, where 380 pieces of information have been checked in total Errors of statistic data from the field shows that error of our e-maps is within 1:5000 map range, and therefore, the precision conforms to the aquirements of the 1:5000 Scale Mapping Specification stipulated .
|
PDF Full Text Request