Study On Slope Slip Measurement Technology Based On Binocular Vision And Differential Correction

The landslide caused by the instability of the slope caused huge losses,and it poses a serious threat to China's economic construction and people's lives and property.The occurrence of landslide disaster is a gradual accumulation process and there are obvious changes in the characteristics of this process.Monitoring the slope through scientific methods,grasping the law of its change and giving early warning,by this way,it can reduce or avoid the loss caused by landslide disaster to a certain extent.The traditional measurement method is used by professionals to measure with high-precision instruments,which has a long cycle and cannot be monitored in real time.Modern sensor measurement has a limited scope of application or high cost,which is not conducive to a wide range of applications.In order to reduce the loss caused by landslide disasters,this paper uses digital image processing and binocular vision related technologies to propose a low-cost slope slip measurement technology that is suitable for long-term monitoring of slopes where landslides may occur.The method mainly uses artificial markers in the area to be monitored in the slope,uses two industrial cameras to construct a binocular vision system to periodically capture the image of the area to be monitored on the slope surface,and identifies the positioning markers and calculates three-dimensional(3D)coordinates and 3D displacement through image processing algorithms.The main research content includes the following aspects:1.The digital model of the binocular vision system was studied,including the camera imaging model,pinhole model and nonlinear distortion.The commonly used camera calibration method was analyzed.The process of calibrating the camera by the improved Zhang's calibration method that introduces tangential distortion and removes the tilt factor was discussed in detail.According to the results of the calibration of the two cameras,the binocular camera calibration and 3D coordinates were jointly calculated.2.On the basis of analysis of various types of markers,a circular reflective marker suitable for the measurement technique proposed in this paper was designed.Image processing related technology was used to study the identification and positioning of markers,including image preprocessing,image conversion to HSI space for identification,morphological filtering to remove interference points,marker edge detection,ellipse fitting and positioning,and marker number.The influence of camera deflection on the 3D coordinates calculation of the marker was analyzed,and a differential correction method was proposed.The introduction of a reference marker reduces the pixel displacement of markers placed on the surface of the slope,which avoids the disadvantages of measuring errors caused by camera deflection and the time-consuming and labor-consuming need to recalibrate the camera.3.After studying the binocular vision system model and marker design,identification and positioning,and differential correction,a slide slope measurement technique based on binocular vision and differential correction was proposed.Without the need to recalibrate the camera,the 3D coordinate measurement error caused by camera deflection is reduced.The slope simulation environment was set up,and experiments such as camera calibration,marker identification and positioning,differential correction verification,relative distance measurement,and simulated slope monitoring were conducted.The experimental results show that the slope slip measurement technology based on binocular vision and differential correction can effectively measure the deformation of the simulated slope.The differential correction method proposed in this paper can reduce the measurement error caused by camera deflection without recalibrating the camera.It provides an effective exploration of the technology's practical application.