The geometric dimensions and tolerances of blades, which are critical parts of turbomachinery with complex features, must be strictly controlled to ensure the efficiency and safety of the engine. Optical-based inspection systems for blades are increasingly receiving attention because of their high efficiency and flexibility. However, as a key issue in blade inspection, the matching of the part coordinate system and machine coordinate system directly determines the measurement accuracy and automation. The blade surface is complex and has no obvious features, and accurate and rapid matching thus remains a challenging problem to solve. To overcome these problems and realize the accurate inspection of blade profiles, an automatic and high-accuracy matching method for a blade measurement system integrating fringe projection profilometry and conoscopic holography is proposed in this paper. First, automatic rough matching is realized making use of the ability of fringe projection profilometry to quickly obtain high-resolution cloud of points and improving the four-point congruent sets algorithm. The path of the conoscopic holography measurement based on the calibration and rough matching result is then planned, to sample high-precision and uniform cloud-of-points data on the blade surface. Finally, a fine matching optimization algorithm is implemented with the signal-to-noise ratio as the weight. The results of simulation experiments and inspection case studies demonstrate that the proposed matching method is efficient and accurate.