This study examines the conversion strategy of 2D to 3D cadastral maps by standardizing the height limits of land rights based on LU/LC. To achieve 3D cadastral maps, the research proposes a conversion strategy considering height factors. The height dimension of cadastral maps faces challenges in determining maximum heights for features like buildings, given varying regional regulations. As a solution, the concept of surface feature height (SFH) is applied along with LU/LC classification. Economic considerations, such as state revenue from taxes, are also factored into the proposed height limits. The results indicate that building/property heights in Bekasi Regency show significant development potential. In the residential sector, the maximum height reaches 24 m, lower than Bekasi City (48 m) and Bandung City (30 m). In the industrial sector, while heights can reach 25 m, the regulatory limit is only 9 m, posing challenges for investment. In the commercial sector, the maximum height can reach 45 m, but the low regulatory limit of 10 m restricts further development. This research provides a foundation for policy development and an effective 3D cadastral system, emphasizing the need for Bekasi Regency to re-evaluate its building height regulations to maximize its development potential.

Land registration is an important program in asset legalization with the vast land resources owned by Indonesia. The reality is that there are 48 million certificated lands out of an estimated 126 million certificates throughout Indonesia, so the Ministry of Agrarian Affairs and Spatial Planning/National Land Agency (ATR/BPN) made a breakthrough through the Complete Systematic Land Registration (PTSL) program. The grouping mechanism of the PTSL program is divided into four clusters, namely K-1, K-2, K-3, and K-4. Land parcels included in K-4 have problems, namely that they have been registered but not mapped/mapped incorrectly, so the idea of modeling the K-4 typology is needed for accelerating the improvement in land data quality (KW). The research location is in Bandung Regency, which is included in the top five land offices with the highest number of K-4 in West Java Province. This research method uses a mixed method, namely quantitative with a Slovin approach for the identification and clustering of K-4 typology and descriptive qualitative for justification of typology modeling in multiple aspects. The results of K-4 typology modeling were 128 clusters based on seven types of data criteria and obtained 4 clusters that matched the data sample, namely T-1, T-43, T-63, and T-128. The four clusters were justified against the K-4 concept in the Ministry of ATR/BPN, the FFP-LA concept, and the PESTLE framework. Dissemination of K-4 typology modeling is a breakthrough that can be implemented by the Ministry of ATR/BPN and its staff in various regions and the role of multi-concepts in this research can be an input for improving the K-4 concept that has been in effect so far.

Airborne laser technology produces point clouds that can be used to build 3D models of buildings. However, the work is a laborious process that could benefit from automation. Artificial intelligence (AI) has been widely used in automating building segmentation as one of the initial stages in the 3D modeling process. The algorithms with a high success rate using point clouds for automatic semantic segmentation are random forest (RF) and PointNet++, with each algorithm having its own advantages and disadvantages. However, the training and testing data to develop and test the model usually share similar characteristics. Moreover, producing a good automation model requires a lot of training data, which may become an issue for users with a small amount of training data (limited data). The aim of this research is to test the performance of the RF and PointNet++ models in different regions with limited training and testing data. We found that the RF model developed from a small amount data, in different regions between the training and testing data, performs well compared to PointNet++, yielding an OA score of 73.01% for the RF model. Furthermore, several scenarios have been used in this research to explore the capabilities of RF in several cases.

Before the development of 3D cadastre, cadastral systems were based on 2D representations, which now require transformation or updating. In this context, the first issue is that existing 2D rights are not aligned with recent 3D data acquired using advanced technologies such as Unmanned Aerial Vehicle–Light Detection and Ranging (UAV-LiDAR). The second issue is that point clouds of objects captured by UAV-LiDAR, such as fences and exterior building walls—are often neglected. However, these point cloud objects can be utilized to adjust 2D rights to correspond with recent 3D data and to update 3D building models with a higher level of detail. This research leverages such point cloud objects to automatically generate 3D rights and building models. By combining several algorithms, such as Iterative Closest Point (ICP), Random Forest (RF), Gaussian Mixture Model (GMM), Region Growing, the Polyfit method, and the orthogonality concept—an automatic workflow for generating 3D cadastral models is developed. The proposed workflow improves the horizontal accuracy of the updated 2D parcels from 1.19 m to 0.612 m. The floor area of the 3D models improves by approximately ±3 m². Furthermore, the resulting 3D building models provide approximately 43% to 57% of the elements required for 3D property valuation. The case study of this research is in Indonesia.

Land taxation, mainly the Land and Building Tax (LBT), is a critical revenue source for local governments in Indonesia, yet its contribution remains suboptimal due to inefficiencies in assessment and administration. This study focuses on the Lebak Regency to optimize the Sales Value of Taxable Objects (SVTO) assessment, a key determinant of LBT revenue. Using spatial analysis, Land Value Zones (LVZ), and socio-economic data, the research evaluates assessment ratios to identify discrepancies between SVTO and market values. Adjustments to SVTO are proposed based on spatial patterns, such as land use and infrastructure, and the Index of Developing Villages (IDV). Findings reveal significant disparities between assessed and market values, with assessment ratios ranging from 1% to 267%. Simulations indicate potential LBT revenue of IDR 224.7 billion under full compliance, compared to IDR 44.22 billion, the 2024 target. Adjustments based on spatial patterns from land use planning and village development indices enhance equity and accuracy in tax assessments, optimizing local revenues. Despite these improvements, the study’s limitations lie in the lack of community validation for the proposed methodology, which is essential to confirm its practicality and acceptance. Future research should address this gap and explore household-level dynamics, including tax affordability and spending patterns, to enhance policy inclusivity and align taxation systems with local socio-economic realities.