As urban development recently, the need for detailed 3D city models grows exponentially. Central to these models are the 3D depictions of buildings, indispensable for hazard assessments such as flood and earthquake risks, and post-disaster evaluations. Central to these models are the 3D depictions of buildings, indispensable for hazard assessments such as flood and earthquake risks, and post-disaster evaluations. These models are primarily derived from 3D point cloud data, sourced from LiDAR scans or photogrammetry. While aerial mapping is a prevalent method to collect this point cloud data, the task of generating thorough point cloud maps for buildings becomes challenging. These involve maintaining a balance between coverage and resolution, as well as occlusions like trees or neighboring buildings. This study presents a sophisticated method that enhances building point cloud maps by seamlessly merging aerial data with 3D ground point cloud data. By harnessing building footprint data, point clouds from both aerial and ground sources are extracted, focusing on shared static points. Using these points, a transformation is computed to precisely align the ground map to its aerial counterpart, enriching the overall building point cloud data. To validate our approach, we conducted an experiment in a residential region to achieve a dense and accurate point cloud of individual residential buildings. We also introduced a novel bike cargo scanner, designed for rapid, close-range ground data collection. Our results conclusively demonstrated the successful integration of rich and accurate building point cloud data into aerial data.