Real-time people detection from thermal images by using an Unmanned Aerial System
| Autor: | Sergio Salazar Cruz, Eduardo S. Espinoza Quesada, Rogelio Lozano Leal, Filiberto Munoz Palacios, Ivan Gonzalez-Hernandez, Alejandro Morfin-Santana |
|---|---|
| Rok vydání: | 2019 |
| Předmět: |
010504 meteorology & atmospheric sciences
business.industry Computer science 05 social sciences Thermal signature 01 natural sciences Vehicle dynamics Set (abstract data type) Region of interest Control theory 0502 economics and business Thermal Computer vision Artificial intelligence 050207 economics business Representation (mathematics) 0105 earth and related environmental sciences |
| Zdroj: | CCE |
| DOI: | 10.1109/iceee.2019.8884561 |
| Popis: | In this paper a real-time approach for human bodies detection at outdoors environments by using thermal images taken from an Unmanned Aerial System (UAS) is presented. The first step of the proposed algorithm consist on the extraction of the thermal signature of human bodies. After that, we used the Moments Representation technique as a shape descriptor to identify human presence in a scene. Finally, a robust controller is developed to enable the multi-rotor vehicle to track a desired reference in order to scan a region of interest. To show the effectiveness of the developed algorithm when it is implemented on an octorotor vehicle, a set of real-time experimental test were conducted and we present the obtained results. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|