A new wavefront sensor data processing algorithm is described and analyzed. The wavefront sensing concept is similar to a Shack-Hartmann type wavefront sensor, but uses an image projection correlation algorithm as opposed to a centroiding approach to estimate optical tilt. This allows the wavefront sensor to estimate tilt parameters while guiding off of point sources and extended objects such as the surface granulation of the sun. The projection-based cross-correlating scheme differs from a 2-D correlation-based tilt estimation approach in that the images are vectorized on the focal plane array itself prior to readout. This on-chip preprocessing approach allows the wavefront sensor data to be compressed, which results in a large reduction in the amount of data read out of the focal plane array while maintaining the desired bandwidth of the adaptive optical system. An implementation of the projection-based wavefront sensor algorithm is described in detail, showing important signal processing steps on and off of the focal plane array of the sensor. The algorithm design is tested in simulation for speed and accuracy by processing simulated solar and astronomical datasets. Timing analysis is presented, which shows how the collection and processing of image projections is computationally efficient and lends itself to a wavefront sensor design that can produce both competitive speed and tilt estimation accuracy.