Drone Technology to Map Mine SitesWed, 10/19/2016 - 12:47
Detector Exploraciones’ drone project was designed to create georeferenced aerial photography of any linear or surface project. In order to conduct an aerial photographic survey, the processing and post processing of the information was divided into two activities, a field phase and an interpretation phase. The study area comprised two areas, one a powerline with a length of 40km, and the second with a length of 70km. The 110 control points were defined using sub-meter GPS accuracy, in order to make an adjustment in the post-processing for generating the photomosaic of the land. Each and every one of the points was marked with square shaped references with a dimension of 2m2, and in the center a 0.5in rod was placed. Checkpoints were erected prior to the flight day. During the preparation and calibration stage, air vehicle assembly, calibration, remote control, telemetry link, review, and breakeven load was performed. Based on the reference coordinates (AOI) and points previously recorded, control software defined the flight pattern in the flight plan to cover the area of interest. In this case, it defined a flight path from recorded coordinates. 10 photography sessions were held with an approximate duration of 45 minutes each. In total, more than 202 pictures were taken with a geographic reference (geotag).
Subsequently, during the interpretation stage, the processing and initial calibration of photos was carried out to determine flight altitude. Regarding the flying height and positioning data of each of the photos, the initial processing and calculation with an optimal resolution of 7.96cm/pixel was generated. The data processing was performed using the WGS 84 UTM zone 13 coordinates system. From geolocation data, relative terrain heights were calculated, with the goal of using the point cloud to generate the mooring or tie between images. Once the point cloud generated the mooring, both the filtered point cloud and the densified format were created. Based on the structure of the point cloud, the digital surface model (DSM), which corresponds to a continuous numerical model representing the terrain heights and surface objects like buildings and vegetation, is generated. From the set of pictures calibrated, the process of image orthorectification with the help of ground control points was performed. Pairing or matching of images was established from the control points and the number of pictures present on the scenes. Finally, an investigation of the area was carried out in the National Agrarian Registry and the Land Registry in order to obtain all the legal background and to define the land surface layer.