Abstract:

Using photo data of unmanned aerial vehicle (UAV) for building 3D models has been widely used in recent years. However, building a 3D model for deep open - pit coal mines with the mean height difference between surface and bottom of mines to over 500 m, there has not been researched mentioned. The paper deals with the assessment possibility of developing 3D models for deep open - pit mines from UAV image data. To accomplish this goal, DJI's Inspire 2 flying device is used to take the photo at Coc Sau coal mine. The flying area is 4 km2, the flight altitude compared to the takeoff point on the mine surface is 250 m, the overlaying coverage is both horizontal and vertical is 70%. The average errors of the horizontal and height elements of the reference points photo correlates are 0.011 m, 0.017 m, 0.016 m, 0.049 m, and 0.051 m. The maximum error on the X-axis is - 0,025 m, and the Y-axis is 0.028 m, the maximum horizontal error is 0.034 m, the maximum error on the Z-axis is 0.095 m, and the position error is 0.095 m. These results show that the 3D model established from photographic data by Inspire 2 device has satisfied the requirements of the accuracy of establishing the mining terrain map 1: 1000 scale.