The Enhancement of Thermographic Data in the Exploitation of Urban Environment Development

The usage of thermographic data has been utilized more frequently with the advancement of infrared cameras. Thermographic data can be used in a variety of applications in the field of science and technology. The application of thermographic data depends on the type of acquisition itself. In the context of this thesis, thermal data were acquired with a Mid-Range infrared camera mounted on General Aviation airplanes. This thesis addresses some of the most important questions that arise from dealing with thermal data. The solutions provided by this thesis aim to be a practical way to test the accuracy and improve the usage of the acquired thermal data. On several occasions, the radiometric calibration of infrared cameras is done in a lab using expensive equipment and high-end blackbodies. This type of calibration cannot be replicated in most work environments; however, alternative methodologies may be exploited. In this thesis, a calibrated camera was utilized to calibrate an uncalibrated sensor by means of empirical methods being adopted. Using an aluminum checkerboard panel, two FLIR A8581 Infrared cameras were compared, where the calibrated one served as a control unit and benchmark for the analysis of the uncalibrated camera readings in a standardized environment. More in detail, the Python library OpenCV was used by determining the temperature value offsets generated when the calibrated camera was compared with the uncalibrated one. The usage of thermography was optimized in this thesis by comparing several image formats. These images were studied and compared to view the best methods to map urban temperature values. Three main types of image formats were used, TIFF32 images, R-Jpeg images, and Grayscale Jpeg images. Data acquired via airborne sensing suites were processed using two different software. Agisoft Metashape, a commercial closed-source software utilized to process all the image formats while Open Drone Map, an open-source application, was adopted to process the non-TIFF32 images. The objective was to compare the functionality and performance of commercial software with open-source applications in the field of photogrammetry. As the comparison led to interesting results, a new visual solution was generated mapping the location of temperature anomalies in the urban environments. Using these anomalies, six locations were inputted in Python to have their highest temperature value studied. The results generate a clear opportunity to develop more practical ways to map different contexts and landscapes, providing useful insights on the territory utilizing the same testing methodologies on different scales. These methods can be further adopted by companies that use thermographic data as these empirical methodologies can minimize computation losses and performance decrease depending on their application.