Terrain Change Detection Model Based On Mountain Feature Information

In mountainous areas in China,when natural disasters such as earthquakes,landslides,and debris flows occur,the local terrain often changes drastically.In this case,using the terrain change detection technology can timely,quickly and accurately obtain quantitative terrain change data,which is of great practical significance to accurately study and judge the disaster situations and scientifically formulate rescue and relief plans.Firstly,in order to eliminate the detection error of terrain change caused by multitemporal remote sensing image,a remote sensing image registration algorithm based on mountain feature information is proposed in this thesis.Traditional remote sensing image registration algorithms mostly use corner information and edge information as registration feature points,may causing problems such as uneven distribution of feature points,obtaining a large number of redundant feature points,and feature points not having actual geographical information significance in mountain remote sensing image registration with less ground object information.To solve these problems,this thesis takes the mountain peak of physical geographic information as the remote sensing image registration feature point,and takes mountain slope and slope aspect information as the matching basis of feature point pair to carry out remote sensing image registration.Experiments show that the remote sensing image registration algorithm based on mountain feature information designed in this thesis is more practical than the traditional remote sensing image registration algorithm for mountainous areas with less ground object information.Secondly,a terrain change detection model based on mountain feature information is designed.It takes geometrical features of the topographic contour information,topology information of terrain contour,mountain feature information as the influence factors of the terrain change detection model,obtains the overall terrain similarity by analytic hierarchy process,and weights the similarity value of each influencing factor to obtain the overall similarity of topographic change.By marking the area with a large degree of change,the obvious area and change degree of terrain can be intuitively obtained.It solves the offset problem between the traditional terrain change detection results and the actual geographical information changes.And,combining the change detection model with the actual geographical information,it greatly improves the accuracy of terrain change detection.The experiment proves that the terrain change detection algorithm based on mountain feature information designed in this thesis can achieve good combination of physical geographic information,and good adaptability,registration accuracy and operation efficiency for large areas with obvious terrain features.