The Study Of Slope Displacement Monitoring Method Based On Monocular Vision

Due to the instability of geological structure, China is a high incidence zone oflandslide and other natural disasters. Landslides and collapse often occur on the Slopeand other traffic facilities because of geological changes. As a result, it seriouslyaffects the traffic safety. Therefore, In order to detect dangerous in time and avoid theoccurrence of security incidents, it is necessary to strengthen the monitoring of slope inwhich safety hazards may exist. Slope deformation is the most significant parametercharacterization in the detection of surface conditions on the slope. It also can reflectthe change of slope condition and development trend. If we can monitor slopedisplacement accurately and forecast the slope state of deformation trend, we will beable to step back to take relevant measures. In this way, this kind of disaster losses canbe reduced to a minimum.Slope displacement detection is based on a fact that insecurity will inevitably leadto certain change (Structure, size, spatial position, etc.) on monitored surface, includingArtificial field observation method, GPS detection method, three-dimensional laserimaging method and image detection method. Because of the limitations of thetechnology and conditions, some kind of traditional methods such as artificial field testmethod, GPS and so on are ineffective, expensive and dangerous. As the computerscience and image recognition technology has become more sophisticated, Imagemeasurement method is paid more and more attention by the displacement detectionfield. Image processing and recognition technology are used to obtain parameters oftargets, to extract interested targets in images and to meet the demand for displacementmeasurement. This method is low cost, ease of implementation and withoutpost-maintenance. These characteristics make it applicable to the displacement matrixmonitoring.Based on the above conclusions, this article has put forward a slope displacementmonitoring based on monocular vision. There are three aspects of work in this paper:First, Slope detection system design: an integrated study is developed on slopedisplacement detection process. Then we design the slope displacement detectionsystem based on it. We also analyze the system structure and logic processes. Second, The detection and recognition of target bodies: use image processingtechniques to analyses-process slope image.To achieve the target identificationaccurately and extract the center coordinate values by use the recognition model. In thetargets detection, compared with the advantages and disadvantages of the commonlyused method, we launched a new algorithm for the research. This algorithm has highefficiency and reliability.it fit the quality of real time and accuracy which intelligentdetection system requires.Upon completion of identity recognition and coordinate Calculation, a positionmeasurement algorithm is improved based on the characteristics of space round. Ithelps to complete the displacement array positioning. A computation model isestablished according to this algorithm. We also adopt the numerical simulation toensure that it can be applied to the slope detection.