A stitching algorithm based on orthonormal polynomials in a sparse subsperture area was proposed. In this algorithm, Gram-Schimdt orthogonalization of circular Zernike polynomials is performed by using Mathematica9.0, and the standard orthonormal polynomials, Z-sparse polynomials, which show orthogonality in sparse subaperture area were established. Wavefront data in sparse subaperture area can be fitting with the new orthogonal polynomials. The experimental results show that the wavefront residuals of peak to valley value and root mean square are 0.071 9λ and 0.007 4λ respectively compared with direct testing result. Therefore the algorithm can effectively stitch the seven subapertureswavefront data of interferometry.