Detection Of Rolling Stones On Mountainous Region Using Aerial Image And Airborne LiDAR

In recent years,Rolling Stones have occurred.It has a wide range of distribution,the occurrence of sudden,high frequency characteristics and becomes a serious threat to the surrounding human and various structures,with serious harm.In general,Rolling Stone protection needs to be obtained through field measurements,and then take engineering protective measures.This method is time-consuming and labor-intensive,and the mountainous terrain is complex,the field mapping has a certain risk.At present,domestic and foreign use of image of remote sensing and other data for rolling stone extraction,there is a low extraction accuracy,false alarm rate,low degree of automation and other issues.Aiming at these problems,this paper proposes a new method to extract the information of mountain rolling stones by using aerial image and LiDAR data which includes the multi-scale segmentation of the image,the green vegetation information and the shadow information extraction,the elimination of the green vegetation and the shadow of the final rolling information extraction interference,the removal of vegetation information and shadow information segmentation object based on the use of this article based on Shadow and elevation of the rolling stone extraction algorithm.The main work is as follows:1.This paper combines green vegetation index and object-oriented thinking to extract green vegetation information.The green vegetation information extraction based on pixel is easy to have the problem of extracting more noise points and incomplete extraction results.In view of these problems,this paper proposes a combination of green vegetation index and object-oriented thinking for green vegetation extraction to obtain a better extraction effect.Extraction of green vegetation information,excluding the formation of green vegetation due to the extraction of rolling information on the interference.2.This paper improve a shadow detection algorithm based on color model.For the aerial image shadow detection,with the shadow area has a similar spectral nature of the green space,the brightness of the dark non-shadow,high brightness of the shadow area prone to false detection and other problems.Based on the RGB and HSI color space,this paper proposes a shadow index to extract the shadow information respectively.According to the relationship between the blue and the green band in the RGB color space,the preliminary extraction result is further filtered and then get the public space of the two color spaces as the final result.Finally,the algorithm of shadow extraction is verified by experiments.The extracted shadow area will be used as an important contextual feature for the Rolling Stone information extraction.3.In this paper,an algorithm for rolling information extraction based on shadows and elevation is proposed.This paper analyzes the characteristics of the rolling stone in the image and LiDAR data,synthesizes the spectral characteristics of the rolling stone,the context characteristics,the characteristics of the LiDAR data,and so on,and puts forward the rolling information extraction algorithm based on the shadow and elevation information and then combines with elevation information filtering to get the final result of rolling information extraction.Finally,the accuracy of assessment and analysis,the basic to meet the political and other departments of the Rolling Stones protection needs.4.The object-oriented C ++ language,based on the Qt development framework,uses the open source QGIS,GDAL,Laslib and other libraries for secondary development.In the Microsoft Visual Studio 2015 development environment,this paper proposes the rolling information extraction algorithm proposed in this paper,including image data processing,LiDAR data processing,rolling stone information extraction and other modules.