Spatialization of Territorial Parcels for Efficient Land Management

Abstract

Georeferencing urban land parcels is essential for municipal management, supporting urban planning, land regularization, tax collection, integrated public services, and legal security in real estate transactions. It also promotes transparency and citizen participation while aligning administration with geographic information systems and federal regulations, including the City Statute and National System for Territorial Information Management (SINTER). Despite these benefits, georeferencing urban land parcel in Brazilian municipalities is limited. In 2019, only 21% of municipalities had a georeferenced cadastral database, while 59% collected only descriptive data, often without spatialization, and some lacked automation. This highlights the need for methods enabling large-scale spatialization of urban parcels and updated cadastral mapping. This work proposes a Fit-For-Purpose Land Administration methodology using National Address Database for Statistical Purposes (CNEFE) and municipal cadastre data combined with OpenStreetMap. Results show that georeferencing can achieve a maximum error limited to half the block dimensions when CNEFE points are adequately distributed, while discrepancies reveal potential cadastre integrity issues. Collected images and operation logs may support future deep learning applications, enhancing accuracy and automation.