Key points are not available for this paper at this time.
Land Value Zone (LVZ) is essential for spatial planning, taxation, and land management in Indonesia. However, identifying these zones still faces significant challenges, especially in linking land use change with land value. Despite their object-capturing limitations, pixel-based identification methods have been the main approach in satellite image analysis. The Object-Based Image Analysis (OBIA) approach offers the potential to understand a more contextualized spatial relationship between land use change and land value. To our knowledge, no study has compared pixel- vs object-based performance in LVZ. We tested OBIA and pixel-based approach using Landsat-8 to link land use change with LVZ change in Gondangrejo District during 2015, 2019, and 2024. The pixel-based method used the Normalized Difference Built-up Index (NDBI), whereas the OBIA applied multi-resolution segmentation and rule-based classification. The results demonstrated that the OBIA method consistently outperformed the pixel-based across all observation years. The overall accuracy of OBIA in 2015, 2019, and 2024 were 96,15%, 92,30%, and 94,23%. Meanwhile, the pixel-based method achieved 80,76%, 80,76%, and 82,69%, respectively. In terms of land value, the Average Indication Value (AIR) increased from IDR 7,000 in 2015 to IDR 22,000 in 2024, with the highest recorded value reaching IDR 5,106,000 in 2019 before decreasing to IDR 4,740,000 in 2024. These changes were concentrated in residential development areas. The results show that the OBIA approach provides a higher classification accuracy and more clearly identify zones with significant land value dynamics. OBIA shows strong potential to improve remote sensing-based LVZ updates.
Adiningsih et al. (Tue,) studied this question.