Optimization of grid-based flight path planning and processing time for the application of photogrammetry in mining with low-cost UAVs

The digitization and modelling of mining installations has been revolutionized by the introduction of UAV photogrammetry. However, due to the large size of mining operations and facilities, the high volume of images to be processed implies hardware requirements, computational power and time consumption, which complicates the rapid obtaining of results. Coverage Path Planning allows to define how the images are captured. Conventional approaches define automatic flights with cross-flight trajectories using high coverage values (over 80% frontlap and 70% sidelap) and consequently a high number of images, which increases processing time. To optimize the time of UAV photogrammetry, this research proposes the use of grid-based flight with reduced overlap values (less than 70% frontlap and 60% sidelap) at different heights. The case study proposed for an observation area of average size and moderate elevation difference using low-cost equipment employs a single commercial software package to generate 3D representations. The results show that it is possible to obtain around 48.5% and even 74.2% fewer images while maintaining the accuracy of the final model, with an increase in RMSE of 9 mm in the worst-case scenario. In terms of model information or point density, the optimized models show reductions in the amount of information of less than 5%, reaching 21.0% in the most unfavorable case considered. In addition, reductions in processing time of around 50% are achieved depending on the conditions, while reducing the computational demands to obtain results quickly through a rapid mapping mode for application in the mining field.