Improving holographic search algorithms using sorted pixel selection

Abstract

Traditional search algorithms for computer hologram generation such as Direct Search and Simulated Annealing offer some of the best hologram qualities at convergence when compared to rival approaches. Their slow generation times and high processing power requirements mean, however, that they see little use in performance critical applications. This paper presents the novel Sorted Pixel Selection (SPS) modification for Holographic Search Algorithms (HSAs) that offers Mean Square Error (MSE) reductions in the range of 14.7 − 19.2% for the test images used. SPS operates by substituting a weighted search selection procedure for traditional random pixel selection processes. While small, the improvements seen are observed consistently across a wide range of test cases and require limited overhead for implementation.

Efficient Excitation of High-Purity Modes in Arbitrary Waveguide Geometries

“…values to these is sub-optimal, and that careful selection of the next pixel to be altered [53] and careful selection of the value it is set to [54], [55] can significantly improve both execution times and convergent error.…”

Relative limitations of increasing the number of modulation levels in computer generated holography

“…This is similar to a direct search algorithm, except that the pixel under consideration is set to the best possible quantised value, rather than merely considering the current value and a single randomly-selected alternative. [16] It is felt that this offers a better comparison between binaryand multi-level holograms as the randomness inherent in direct search algorithms is eliminated at the expense of higher computational requirements.…”

Benchmarking the Gerchberg-Saxton Algorithm

“…Traditionally FFT performance is significantly better when the resolution is a power of 2. A test of 500,000 FFTs with random resolutions between 2 1 1 and 2 1 2 factorised solely into primes larger than 100 concluded that the performance cost is of approximately 2.6 − 3.1 times that expected of a resolution power of 2.…”

“…While many high performance algorithms are available [1][2][3][4][5], less information is available on the precise factors influencing performance. In order to support our ongoing research into high power laser projectors this paper benchmarks the widely used GS algorithm.…”