论文部分内容阅读
斜率算法对哈特曼-夏克传感器非常重要,而基于归一化互相关斜率算法的哈特曼-夏克传感器可满足强噪声情况下点光源和扩展目标的斜率探测需求。为解决互相关斜率处理器可移植性差的问题,提出用多核中央处理器(CPU)来实现斜率探测。在优化归一化互相关算法的基础上,编程实现斜率探测,并从多核并行和运算向量化两方面优化程序,提升斜率提取速度。当模板分辨率为9 pixel×9 pixel时,优化程序在Intel(R)Core(TM)i7-3770k四核计算机上运行,完成400个15 pixel×15 pixel的子孔径斜率提取,用时约为340μs。搭建了自适应光学系统,对优化后的斜率算法进行系统闭环测试。结果表明,使用多核CPU的相关哈特曼-夏克自适应光学系统能在强噪声环境下进行有效的波前校正。
The slope algorithm is very important for the Hartmann-Shack sensor, while the Hartmann-Shack sensor based on the normalized cross-correlation slope algorithm can meet the slope detection needs of point light sources and extended targets under strong noises. In order to solve the problem that the cross-correlation slope processor has poor portability, a multi-core central processing unit (CPU) is proposed to realize the slope detection. On the basis of optimizing the normalized cross-correlation algorithm, the slope detection is programmed, and the program is optimized from the aspects of multi-core parallelization and vector quantization to improve the speed of slope extraction. When the template resolution is 9 pixel × 9 pixel, the optimizer runs on an Intel ® Core ™ i7-3770k quad-core computer and completes 400 sub-aperture slopes of 15 pixels × 15 pixels in approximately 340 μs . The adaptive optics system is built and the closed-loop test of the optimized slope algorithm is carried out. The results show that the relevant Hartmann-Shack adaptive optics system using multicore CPUs can perform effective wavefront correction in strong noisy environments.