Abstract: The propeller blade is a key component in industrial applications. The shape of this surface often contains multiple complex surfaces, and the traditional measurement methods have defects of accuracy and timeliness. To address this problem, a quality assessment method of regional feature difference facing multi-faceted surfaces is proposed based on an automatic propeller blade measurement device. Firstly, the measurement device is constructed and the simulation attributes are designed based on mechatronic platform. Secondly, a regional energy clustering segmentation method driven by feature vector is proposed. The problem of multi-faceted concentration on the propeller blade surface is solved by segmenting the surface measurement region. Then, the local optimal measurement points are extracted based on segmentation regions and the layout of measurement point is also planned based on segmentation regions. Then, the non-linear regional accuracy difference hierarchical alignment method is proposed. The method achieves the positional transformation by hierarchical alignment through differences in accuracy between measurement regions, and the algorithm efficiency is optimized by combining non-linear calculation to achieve fast and efficient quality assessment. Finally, the feasibility of the method is verified by joint simulation and physical test.