Displacement Monitoring System Of Slope Dangerous Points Based On Close Range Photogrammetry With Large Intersection Angle

In order to solve the monitoring problem of rockfall and landslide in unattended area,and to master the slope structure in real time,a slope displacement monitoring system is designed based on close range photogrammetry.The slope features are enhanced by placing markers on the slope,and the markers are identified and roughly located by convolution neural network,which is obtained by image edge extraction and graph fitting Through the registration of images captured by binocular cameras and the correction and conversion of spatial coordinates,the spatial coordinates and displacement of monitoring points can be obtained,so as to obtain the displacement of the slope.This paper has carried out the following research(1)The original security equipment and its lines around the road and other facilities are used,and some cameras and other equipment are added when necessary to cross photograph the slope with binocular vision method.The front-end layout scheme of the binocular close range photogrammetry system is designed.The original transmission system is combined with the additional switching equipment to connect the camera.At the same time,the network bridge is added to realize the communication,so as to complete the functions of real-time transmission,video recording,storage and forwarding of images.On this basis,the software is used to complete the image acquisition,image processing and calculation,and finally obtain the displacement of the slope.(2)Using the artificial marker as the displacement monitoring object,the ball is selected as the marker and its structure is designed.In order to achieve better imaging effect,different marker layout schemes are designed according to the different positions of the camera relative to the ball.(3)In order to reduce the manpower consumption and improve the efficiency of identification,convolution neural network method is used to extract the rough region of markers.In order to reduce the amount of data processing and improve the speed of calculation,the obtained marker color image is binarized,and then the edge of the obtained binary image is detected,so as to obtain the marker edge of this study.The Hough transform is used to fit the edge in a circle,and then the center coordinates of the marker in the image are obtained.Since the coordinates of the center of the circle obtained are the pixel coordinates in the image clipped according to the rough extraction results of the marker,in order to obtain the coordinates of the center of the marker in the original image,the clipped image is matched with the original image,so as to obtain the coordinates of the center of the marker in the original image.So far,the pixel coordinates of the center of the marker are obtained,which can be used for three-dimensional identification Three dimensional space coordinate solution.(4)Through the simulation test of large intersection angle,the influence of intersection angle on measurement accuracy is studied.Since the coordinates of markers measured by total station are different from the coordinates of monitoring points,the coordinates measured by total station are corrected.The test results show that the error level is low when the intersection angle is 90 ° and it is feasible to use large intersection angle for displacement monitoring.(5)Based on the results of simulation test,the slope displacement monitoring test is carried out on a construction site.Through the operation of on-site monitoring point layout,camera erection,wireless bridge configuration and transmission image setting,the preparatory work of image transmission is completed.The displacement monitoring of several monitoring points fixed on the slope surface is carried out by using binocular close range photogrammetry method.The test results show that this method can meet the monitoring requirements of slope mutation.