Monitoring of slope activity was a hot issue in slope monitoring research.Through monitoring the slope activity could provide technical basis for predicting slope instability and was the key to reduce landslide disaster.The existing monitoring methods were expensive and most of them adopt point-based measurement,which could not reflect the overall movement state of the slope,and it was difficult to carry out long-term real-time monitoring.In view of the shortcomings of the existing monitoring methods,this paper applied digital image correlation method to slope monitoring,analyzed the difficulties existing in the application of digital image correlation method in the site monitoring of slope,and studied the influence of external factors such as climate,environment and physics on the calculation accuracy of digital image correlation method through experiments.In this paper,a slope activity monitoring method suitable for the site was proposed.The program of slope activity monitoring method was written based on MATLAB language and the corresponding GUI interface was designed.The slope of Chengqin Expressway in Qinhuangdao city was taken as the observation site,the above method was applied to the slope site monitoring,which verified the feasibility of the method and expanded the application scope of digital image correlation method.In the aspect of climate factors,the indoor simulation experiments were carried out to analyze the influence of light changes and fog conditions on slope image features.According to different climate characteristics,the local sensitive histogram was used as the illumination invariant feature of the image to reduce the influence of illumination change on the monitoring accuracy;The dark channel defogging algorithm was used to defog the image,which improved the image clarity and contrast.In the aspect of environmental factors,this paper analyzed the distribution characteristics of different landforms on the slope image,and a method for the division of slope optical measurement area was proposed to divide the vegetation area and the geotechnical area.The SIFT feature point method was used as a supplementary method,and the correlation coefficient method was proposed to replace the Euclidean distance method for matching,and according to the applicable conditions of digital image correlation method and SIFT feature point method to calculate different regions.In the aspect of physical factors,the influence of different deformation stages of landslide on the camera sampling frequency were analyzed.According to the evolution relationship between deformation rate and time in different deformation stages of landslide,the camera acquisition program was optimized,and a camera speed control program was designed to adjust the sampling frequency according to the current deformation rate.The monitoring camera was used to build the observation system,and the slope activity monitoring method proposed in this paper was applied to the site monitoring of Chengqin Expressway Slope in Qinhuangdao city.Based on the characteristics of the slope image,the slope images at different times were matched.Through the change of coordinates of corresponding points,the related parameters of slope activity are obtained,and the evolution characteristics of slope activity are analyzed.Based on the principle of binocular stereo vision and the site camera calibration results,the twodimensional measurement was extended to three-dimensional measurement. |