Use of UAV and comparative analysis of PE3D for real estate cadastre update

Abstract

Cadastral updating in urban management, despite its importance for efficient municipal management, has not been continuous in Brazilian municipalities. Research and extension projects aimed at dynamic solutions in urban areas are challenging, especially due to the lack of systematic mapping at scales larger than 1:2000. Recognizing the reduced cost of mapping processes when using Unmanned Aerial Vehicles (UAVs), this article makes it possible to develop an application and evaluate a methodology for the use of UAVs in cadastral updating. The proposed study was carried out in the Luiz Gonzaga neighborhood in the city of Caruaru, where a flight was made to update the cadastre in December 2019. The data obtained from this flight was compared with information from Pernambuco Tridimensional (PE3D), a program created after the floods in the state at 2014. This program resulted in a large-scale map of the territory of Pernambuco at scales of 1:5000 and 1:1000 scales. The proposed methodology was to test automation in the acquisition of terrain features, especially built-up areas, and to analyze the quality of the results by comparing the results of automatically collected data with manually acquired data. Some software and application development solutions hosted on GitHub were researched and tested. Remote sensing techniques, machine learning and the semi-automated image classification algorithm Support Vector Machine (SVM), present in the Dzetsaka plugin optimized for QGIS software were also used. The results obtained from the SVM classification, when compared with the data acquired through manual interpretation of the images, differed by around 8% of the number of built-up areas. As such, the process was considered to have potential for application in a previous computer simulation for municipal governance in relation to cadastral updating or the need for new cadastral mapping. Following on from the study, another comparison was made using the 2014 PE3D mapping base with the 2021 UAV mapping, which showed a 35% increase in built-up areas over a 5-year period. This result helped in the decision as to the need to update the cadastre and ensure a more assertive investment in this type of service contract.