Research On The Method Of Building Information Extraction Based On High Resolution Remote Sensing Images In Oil Field Areas

With the development of satellite remote sensing technology,the resolution of remote sensing image is increasing. At present a lot of remote sensing satellite on orbit can provide us with a lot of sub-meter remote sensing image data,which brings more details of the ground to the researchers. Nowadays, both urban and rural areas are changing rapidly. So the effective use of these remote sensing images is of great significance to the control of information and the future planning. For the oil field areas, it is also very important to effectively control the real time position and state information of the ground construction. But traditional building information extraction method based on low resolution remote sensing image cannot be applied to high resolution remote sensing image data. So it is necessary to study on the extraction method of buildings in the oil field areas of high resolution remote sensing image data. The research on this paper was based on the project of Jiangsu Oil field geographic information management and monitoring system.In this paper, based on the full analysis and investigation of the characteristics of the topography and climate and the distribution of the construction of the oil field, the following work has been completed:(1) The high resolution remote sensing images of the oil field areas were processed such as noise elimination, enhancement and so on.(2) On the basis of a detailed comparison of all kinds of image segmentation algorithm, the remote sensing images were divided into the objects to be processed with the application of image segmentation method based on mean shift filtering algorithm. And the space scale filtering was operated according to the situation of the small object segmentation.(3) A multi feature extraction algorithm based on spectral feature, shape feature and texture feature was proposed. The feature selection algorithm was used to obtain the feature subset which can reflect the characteristic of the sample. And a gray level co-occurrence matrix method based on neighborhood weighted average was proposed to calculate the texture feature extraction of irregular objects.(4) Based on the extracted feature subset, this paper used IOSDATA clustering method to classify the regional objects. And in view of the uneven shape of the building area, the morphological optimization is done to remove the burr and cavity, so that it is more in line with the original shape of the edge of the building.After the analysis of the experiments, the methods in this paper can extract accurate information and shape the rules of the building information from high-resolution remote sensing images of oil field areas within a relatively short period of time, which has brought great help to the geological information management of the oil field areas.