Sistema de Visão Termográfico para Veículos Autónomos Aéreos

• Autor(a) principal: Ferrás, Carlos Manuel de Sousa
• Data de Publicação: 2021
• Tipo de documento: Dissertação
• Idioma: por
• Título da fonte: Repositório Científico de Acesso Aberto de Portugal (Repositórios Cientìficos)
• Texto Completo: http://hdl.handle.net/10400.22/18711
• topic: IR Camera; Thermography; UAV; Câmara IR; Termografia

Visual perception systems are one of the main sources of sensory information used by autonomous systems, mainly for location, navigation and special information about the means of operation where it is located. These sensors allow the acquisition of a large amount of information about the environment in which the system finds itself, which can then be processed to acquire information relevant to the task in question. One of the information that can be obtained using vision systems is the temperature of a certain object/body. This information can be acquired using an IR camera. There are many objects that emit infrared radiation as a function of temperature and the higher the temperature the greater the intensity of radiation emitted by that object/body. This information can be useful in several applications, such as: gas detection, fire detection and surveillance, among others. However, despite providing important complementary information to visible spectrum images, it has some limitations such as capturing images in the dark, and also the high price of these systems. Observing that there are more and more low-cost vehicles capable of taking advantage of these systems, such as UAVs, for various operations (for example, observation or search and rescue missions), it is interesting to develop a small and efficient solution. This will allow its application in more and more vehicles, reducing the cost of the various operations that benefit from the use of these sensors. Using a low cost CMOS IR sensor, Lepton 3.5, the necessary hardware and software for the acquisition of data from this sensor was developed, through Ethernet, which can be integrated with the ROS framework. Increasingly, ROS is an excellent solution in the scope of autonomous systems, allowing the sensor to be integrated in any autonomous system that has a ROS version compatible with the developed software. The developed system has been subjected to laboratory tests that allow comparing its results with those of a system which was known. In addition, the low-cost IR system was integrated into the autonomous aerial vehicle, which confirms its good functioning and possibility of application in various application scenarios. This dissertation aimed to develop a low-cost thermographic system that can be integrated with the existing ROS architecture in robots. This allows that, since the autonomous system has the ROS framework, its integration is simple, fast and mainly low cost and with good results when compared to high cost systems.