Iterative wavefront shaping is a powerful tool to overcome optical scattering and enable the focusing of diffusive light, which has exciting potential in many applications that desire localized light delivery at depths in tissue-like complex media. Unsatisfactory performance and efficiency, however, have been a long-standing problem, and the large discrepancy between theoretical and experimental results has hindered the wide applications of the technology. Currently, most algorithms guiding the iterative search for optimum phase compensation rely heavily on randomness to achieve solution diversity. It is similar to black-box optimization, in which the mechanism for arriving at a good solution is unclear. The lack of clear guidance on the new solution generation process considerably affects the efficiency of optimization. Therefore, we propose a probability-based iterative algorithm that combines the genetic algorithm and ant colony optimization to develop new solutions based on a probability map. Thanks to the clearer guidance provided by the probability map and the reduced involvement of randomness, we can obtain optimization results with optimal efficiency for single and multiple focuses behind scattering media. In addition, with the proposed algorithm, we also demonstrate higher adaptability in an unstable scattering environment and more spatially uniform optical focusing in the field of view. This study advances the state-of-the-art in the practice of iterative wavefront shaping. More importantly, the significant improvement in optimization efficiency and adaptability, if further engineered, can potentially inspire or open up wide applications that desire localized and enhanced optical delivery in situ.
ASJC Scopus subject areas
- Atomic and Molecular Physics, and Optics
- Computer Networks and Communications