Novel railway goaf boundary extraction method based on D-InSAR

Abstract

The invention discloses a novel railway goaf boundary extraction method based on D-InSAR. The method comprises the steps of obtaining a time sequence SAR image and a precision track file in a researcharea; preprocessing the SAR image data to establish a connection graph; carrying out interference processing on interference pairs in the connection graph to generate a differential interference graph, and carrying out geocoding on the differential interference graph; performing goaf identification on the geocoded differential interference graph by using a U-net neural network to generate a goaf-non-goaf binary image; converting the identified binary image grids of the goaf into vectors; carrying out spatial fusion on goaf results after multi-stage vectorization; and counting and analyzing the goaf area, the goaf development direction and the nearest distance between the goaf boundary and the railway. According to the method, the underground goaf threatening railway safety is automatically recognized, the distribution characteristics and development trend of the goaf are systematically analyzed, and technical support is provided for optimizing railway line selection and checking hidden danger points of the railway goaf.

Description

technical field [0001] The invention relates to a new method for extracting the boundary of a railway goaf, in particular to a new method for extracting a boundary of a railway goaf based on D-InSAR (Differential Interferometric Synthetic Aperture Radar). Background technique [0002] The goaf is a prominent ecological environment problem, which will lead to different degrees of ground subsidence around it, directly threatening the safety of houses, roads and infrastructure on the surface. When optimizing the railway direction plan, it is necessary to fully consider safety and later maintenance costs, and actively avoid these hidden mined-out areas. Therefore, fast and accurate identification of gobs is of great significance for optimizing the railway line selection scheme. At present, traditional techniques such as leveling and GPS are mainly used to monitor and identify gobs. Although real-time and high-precision point monitoring results can be obtained, the results can o...

AI Technical Summary

Problems solved by technology

However, goafs are often located in mining areas with complex geological environments, large subsidence levels, and fast-changing speeds. D-INSAR technology cannot directly obtain large-gradient terrain deformation information, and its application is limited.

Images