Urban built-up area extraction using triangle threshold algorithm and Naive Bayes classification model with multidata fusion

Accurate identification of urban built-up areas is crucial for monitoring urbanization and promoting sustainable development. To overcome the limitations of single-data-source methods in capturing human activities, this study proposes a novel approach for urban built-up area extraction, based on machine learning techniques, and multi-source data fusion (POI, Luojia-1 nighttime light, and Landsat 8 imagery). First, we construct the POI-Adjusted Luojia-1 Urban Index (PALUI) by combining POI and Luojia-1nighttime light data quantify population distribution patterns. Next, we segment remote sensing indices (NDVI, MNDWI, NDBI) using the triangle threshold algorithm to differentiate vegetation, water, and built-up areas. Finally, a Naive Bayes classifier is used to fuse these features for built-up area extraction. Experiments in Nanchang City show that the proposed method achieves a precision of 0.89 and recall of 0.76, outperforming SVM, Random Forest, U-net, and YOLO11. The PALUI index effectively reduces light overflow issues in nighttime data, while the multi-source fusion strategy improves edge accuracy in complex urban environments. This method offers a reliable solution for high-precision urban mapping.