التفاصيل البيبلوغرافية
العنوان: |
Multimodal Terrain Traversability Analysis for Unmanned Ground Vehicles in Forestry Environments |
المؤلفون: |
Vasques, Miguel Tiago Vieira |
المساهمون: |
Paulo, João Luís Ruivo Carvalho, Peixoto, Paulo José Monteiro |
سنة النشر: |
2023 |
المجموعة: |
Universidade de Coimbra: Estudo Geral |
مصطلحات موضوعية: |
Terrain Traversability Analysis, LIDAR, Multimodal Sensing, Forestry Robotics, Analise de Travessia, Deteção Multimodal, Robótica Florestal |
الوصف: |
Dissertação de Mestrado em Engenharia Eletrotécnica e de Computadores apresentada à Faculdade de Ciências e Tecnologia ; With the increasing intensity of extreme climate phenomena such as drought, the risk of forest fires has become even greater. A key tool for mitigating the risks posed by these increasingly common conditions is forest perimeter management, the removal of highly flammable material such as branches and dead shrubs – a task often neglected due to the lack of resources and the danger involved. Advances in autonomous navigation create the opportunity to automate this cleanup process with virtually no human resources by equipping the vehicle with sensors capable of providing an understanding of the environment for navigation and safe deployment of actuators that enable identification and removal of combustible material. The accuracy of the perception task is of paramount importance and often impossible to achieve with just a single sensor, so it is important to use auxiliary sensors such as IMU, GPS, etc. To perform autonomous navigation in forest environments, given the variety of observable structures in these areas, which often do not have fixed features as in urban environments (e.g., curbs, crosswalks) and surfaces constructed for navigation purposes, a unique analysis is required to determine whether or not navigation can be performed along the faced path. This analysis aims to obtain a measure of traversability and is performed taking into account various elements such as roughness, slope, elevation and the vehicle's own limitations.In this dissertation, a method of traversability analysis is presented that uses a 3D point cloud obtained from a LIDAR sensor and localization data from GPS and IMU to distinguish zones where traversability is possible/impossible, using parameters such as slope and elevation between surfaces. The method was subjected to various tests in forest environments for which there were no documented tests, and its performance was evaluated in different forest environments ... |
نوع الوثيقة: |
master thesis |
اللغة: |
English |
العلاقة: |
https://hdl.handle.net/10316/113094Test; 203511441 |
الإتاحة: |
https://hdl.handle.net/10316/113094Test |
حقوق: |
info:eu-repo/semantics/openAccess |
رقم الانضمام: |
edsbas.FBFAB8B8 |
قاعدة البيانات: |
BASE |