Research On Filtering Method Of GB-SAR Slope Monitoring Image Data

The stability monitoring of high and steep slopes is of great significance to the safety production of mines.With the development of surveying and mapping science and technology,monitoring technology has also entered a new era,especially the emergence of Ground-based Synthetic Aperture Radar(GB-SAR)monitoring technology.GB-SAR monitoring is one of the emerging monitoring technologies in recent years.Compared with traditional point-to-point monitoring technology,it can reflect the deformation information of the whole slope surface and each characteristic points.It also has the characteristics of all-weather,high precision,non-contact,and high efficiency.In this paper,the Dagushan open pit mine in Anshan City is used as the experimental site to obtain the image data of the slope by using the ground-based SAR monitoring technology.In order to improve the image data quality,the image data filtering method is systematically studied.Based on the research and analysis of the noise data source,influence and the mathematical model transformation of different noise data in GB-SAR image data,the experimental simulation of the ground-based SAR image data is carried out by analogy with the classical filtering method of spaceborne In SAR image data filtering,and the experimental simulation effect is analyzed.It is found that some effective information is contained in the noise data eliminated in the filtering process.Therefore,this paper proposes a new effective filtering method based on NLM filtering combined with the block wavelet threshold filtering method.The purpose is to filter the noise image data twice,extract and reconstruct the GB-SAR image data.This filtering method uses MATLAB software to write the filtering algorithm program,takes the simulation data as the prior carrier of research,carries on the filtering method experiment simulation,carries on the contrast analysis to the experiment simulation effect.Finally,combined with the actual project of Dagushan open pit mine in Anshan City,the application of each filtering method is verified.The experimental research is based on GB-SAR slope monitoring real image data as the data source,using the classical and the new filtering method proposed in this paper to filter the GB-SAR image data,demonstrating the practicality and operability of the filtering method of the image data.The experimental results show that the block wavelet thresholding composite filtering method based on NLM filter can effectively eliminate redundant information and noise data in the data while ensuring the quality of the image data,and protect the gray level information and edge details of the image.Compared with NLM and the total variation filtering method,it improves about 25% and 20% respectively,which verifies the slope monitoring image of GB-SAR.The data can be used as the data base for the early warning of the high and steep slope in the open pit mine.At the same time,the GB-SAR slope monitoring technology also provides an effective guarantee for the safe production of the mine.