Research On Automatic Extraction Method Of Oil Tank Information Based On High-resolution Remote Sensing Image

Oil tanks are used as industrial facilities to store petroleum or fossil fuels.The detection of oil tanks can be used to estimate local oil and gas reserves,so monitoring them is of great significance.With the continuous development of remote sensing Earth observation technology,the spatial resolution of remote sensing images is continuously improved.High-resolution remote sensing images can provide rich features of texture,texture and image details,and provide remote sensing technology for feature extraction technology.Convenience conditions.This paper takes Tianjin Dagang oil and gas field as the research area,high-resolution remote sensing image as the data source,researches the oil tank information extraction method based on high-resolution remote sensing image,improves the traditional object-oriented oil tank information extraction method,and creatively proposes a A method for extracting oil tank information provides a reference method for oil tank information extraction based on high resolution remote sensing images.The main research contents and results are as follows:(1)According to the object-oriented image classification method,when the feature set is artificially designed,the uncertainty is easily introduced,and the phenomenon that the feature description of the feature is low is generated.A universal application for oil tank information extraction is proposed.The characteristics of the tank information feature,the oil tank information extraction for the object-oriented method,can effectively improve the efficiency of the feature set construction in the tank information extraction application.(2)In the object-oriented information extraction method,the image segmentation result affects the accuracy of image classification.The image segmentation result of improved watershed segmentation is replaced by the segmentation result in objectoriented image segmentation,and the object-oriented image classification method is used for oil tank.Information recognition,compared with the object-oriented oil tank information extraction method based on multi-scale segmentation,the accuracy of tank information extraction is improved by 2%.(3)Exploring and constructing a multi-scale matching template that can be applied to different sizes of oil tanks and applying this template to the tank information positioning of high-resolution images,and generating oil tank boundary and tank information extraction results based on the tank positioning results.The oil tank information extraction method based on multi-scale matching template is compared with the oil tank information extraction method based on object-oriented method and the oil tank information extraction method based on improved watershed segmentation.The extraction accuracy of oil tank information is improved by 4.35% and 2.45%,and extraction is performed.Time is also significantly shortened.(4)The traditional oil tank information extraction method based on threshold segmentation and Hough transform is implemented,and compared with the three methods proposed in the paper.The three methods proposed in this paper are used to extract high-resolution remote sensing image oil tank information.It can achieve more than 90% extraction accuracy,especially the oil tank information extraction method based on multi-scale matching template can realize automatic extraction of oil tank information with high efficiency and high precision.