A wavelet-based method of moment calculation with a set of basis functions is presented for centroid estimation in a Shack-Hartmann Wavefront Sensor. The method has been compared with other algorithms such as statistical averaging, FFT and least-squares method. A comparative analysis shows that wavelet method has a high accuracy and processing speed, and better suited for wavefront reconstruction applications. Further, the wavelet method presented here has a variable accuracy and resolution, and can be optimized for a particular application under consideration. © 2005 Elsevier GmbH. All rights reserved.