APPARATUS FOR EVALUATING SAFETY OF CUT-SLOPES

摘要

The present invention relates to a joint safety evaluation device, and the joint safety evaluation device according to an aspect of the present invention receives stereo image data and GPS information generated by photographing an evaluation target surface from the outside, and the depth estimated from the stereo image data And a data processing unit for generating point cloud data consisting of a plurality of coordinates having coordinate values corresponding to the evaluation target surface based on the GPS information, generating mesh data consisting of a combination of a plurality of polygonal mesh surfaces from the point cloud data and stereo A data generating unit that generates modeling data by superimposing image data and mesh data, a plurality of images including a dark slope, and learning in advance using supervised learning values for a region corresponding to a dark slope in each of the plurality of images By applying modeling data to the model, a plurality of mesh surfaces are grouped by using the normal vector for each of the plurality of mesh surfaces extracted from the dark slope extraction unit and the dark slope extraction unit to extract the mesh surface corresponding to the dark slope. Calculate the inclination angle and the direction angle, respectively, and apply the inclination angle and the direction angle for each group to the pre-learned learning model so that the evaluation value for the inclination angle and the direction angle can be output. evaluation score calculation module that receives stereo image data, detects a boundary, and extracts a plurality of boundary coordinates that are coordinates corresponding to the detected boundary among a plurality of coordinates of the point cloud data; or a crack determination unit that selects and outputs either a first classification value corresponding to a crack or a second classification value corresponding to a non-crack as the stereo image data is applied to the learning model trained in advance so as to be discriminated as non-crack; When the first classification value is output, a length width calculator that calculates the length and width of the crack using a plurality of boundary coordinates, receives the length and width of the crack and outputs the crack evaluation value for the length and width of the crack in advance A crack that outputs a crack evaluation value as the length and width of the crack are applied to the learned learning model It includes an evaluation unit and a safety evaluation unit that selects and outputs any one of a plurality of preset safety grades according to the sum of the crack evaluation value and the joint evaluation score.