Development of techniques for identification and mapping of vegetation under high voltage lines using LIDAR / Drone

Abstract

The use of drones / vants has become popular worldwide in civil applications and in the academic community, where research has been developed to improve both the performance of the platform and the association of sensors to drones, substantially increasing the range of use possibilities These platforms. Applications using optical sensors (multi and hyperspectral) and laser technology for the mapping of features and occurrences on small bands of extension of the terrestrial surface have been widely developed. The objective of the project is to establish technical viability parameters for the development of applications using optical sensors (multi and hyperspectral) and laser technology for the mapping and monitoring of vegetation under transmission lines. The specific application will be the generation of classified images showing the type and height of existing vegetation under the high voltage line and systematization of the pruning as a function of the height of the vegetation. In order to achieve the proposed objective, in Phase 1, a LiDAR device, along with its inertial system and software, will be acquired compatible with the company's drones. The sensor / platform integration will be developed and tests will be carried out with the new composition. Experiment, to be conducted in Phase 2, will be planned, in Phase 1, according to DoE (Design of Experiments) techniques to evaluate the LiDAR + Camera combination due to the large number of factors that relate to the density Of the cloud of points collected on the surface such as: scanning frequency; Speed of flight; Scanning angle; Flight height; Terrain topography and reflectance properties of the object. Also in Phase 1, a review of the literature will be done in order to select the best classification algorithms for the types of generated images and the intended objectives. The final results (Phase 2) should be compatible (formats and metadata) with the national spatial data infrastructures (INDE) and state (DataGeo-SP) in order to make public the results obtained. In Phase 1, there will be identification of needs (data transformation, metadata content, etc.) and a negotiation with those involved (company, energy transmission company, FAPESP, SMA / SP, IBGE) for the achievement of this product. The results are expected to have a direct impact on two segments: from LiDAR / Drones applications to cheaper, time-consuming and reliable monitoring of vegetation under high voltage lines; The use of LiDAR sensors by small / micro / medium-sized drone services. Another possible impact is the use of LiDAR technology and the methodology proposed in other environmental applications by governments, state-owned enterprises, NGOs and other companies involved in environmental services. (AU)