Using unmanned aerial vehicle （UAV） oblique photogrammetry technology， multi-view image sequences of engineering construction sites are obtained. Ultra-high-density point clouds are formed by dense matching of structure from motion （SfM） and multi-view stereo （MVS） algorithms. The cloth simulation filtering （CSF） algorithm is used to remove the point cloud of vegetation and abandoned buildings to obtain the real ground point cloud. The earthwork volume between the two-phase point clouds based on the base design model is calculated to realize the intelligent measurement of complex earthwork engineering. Using the above-mentioned UAV method， the rapid earthwork measurement and accuracy assessment of the two construction lands at the foot of Changsha Yuelu Mountain are carried out， and a point cloud density of 538.46 million points per hectare is obtained. The error of the earthwork calculation result relative to the RTK-GNSS measurement result is 2.8%， which satisfies the requirement that the earthwork measurement error is less than 5%. This method has a good application prospect and engineering value.