Real-Time Spot Detection and Ordering for a Shack–Hartmann Wavefront Sensor With a Low-Cost FPGA

摘要

A novel approach for the real-time image processing of a Shack–Hartmann wavefront sensor (SHWFS) is presented. A prevalent problem of these sensors is that during its operation highly distorted wavefronts may lead to spots that leave the confined area of the subapertures. This unnecessarily limits the dynamic range of the SHWFS. When using an SHWFS in a highly dynamic control loop, additionally, customary image processing approaches may result in large latencies, which decisively reduce the closed-loop performance. For increasing the dynamic range of an SHWFS, we employ a single-pass connected-component labeling algorithm that enables a simultaneous calculation of spots and centroids. To reduce the latency, the image processing is implemented on an FPGA in view of fully exploiting its parallelizing features. Thus, the image processing may be done in real-time with considerably lower latency than with a conventional processor. The algorithm is implemented in VHDL on a Xilinx Spartan-6. Obtained results are compared with the established approaches.