Digital Twins (DTs) is an emerging technology which revolutionized geospatial system by providing real-time virtual representation of the physical system. The synchronization of digital twins with a physical world is an open challenge, mainly in the detection and tracking of real-time objects. In terms of geospatial system, this technology provides infrastructure management, urban monitoring and environmental analysis with improved spatial intelligence. To maintain the accuracy of DT, dynamic detection of objects from remote sensing data (aerial and satellite images) is very crucial. This chapter presents an artificial intelligence-based object detection using UAV, LiDAR and multispectral data. The proposed work shows high consistency across all object classes such as trees, roads, construction, vehicles and buildings. The performance analysis shows that building and vehicles have higher F1-measure as compared to others. The IoU metric shows that bounding box alignment is accurate and the average of mAP of all 5 objects is 87.3% which validates that the model performs well in urban images while the model performs average in case of construction areas may due to complex background. Further, the advantages, disadvantages, and future work are discussed.