Annular subaperture interferometry (ASI) has been developed for low cost and flexible test of rotationally symmetric aspheric surfaces, in which accurately combining the subaperture measurement data corrupted by misalignments and noise into a complete surface figure is the key problem. By introducing the Zernike annular polynomials which are orthogonal over annulus, a method that eliminates the coupling problem in the earlier algorithm based on Zernike circle polynomials is proposed. Vector-matrix notation is used to simplify the description and calculations. The performance of this reduction method is evaluated by numerical simulation. The results prove this method with high precision and good anti-noise capability.