摘要
提出以激光束的横模结构特别是基模所占的比例作为激光光束质量的评价依据,并研究了测量方法.建立了基于Hopfield神经网络原理的非线性网络,使用CCD采集激光束的光斑图像作为网络的输入,通过计算该网络的能量函数,实行调整训练,获得网络的动力学稳定状态,此时网络中各阶横模的比例即为测量结果.实验采集了一束光的多幅光斑图像,经预处理后输入该网络,可获得模式结构数据,其中基模分量为69%.利用所得结果合成一幅光斑,与输入的原光斑图像的相对误差为3.53%.
The distribution of transverse modes of the laser, especially the ratio of fundamental mode is proposed to evaluate laser beam quality, and the measuring method is studied. The beam spot images captured by a CCD detector are input into a nonlinear network which bases on the Hopfield network principle. The dynamics steady state is obtained by computing the energy function of the network and performing training arrangement. The proportion of all ranks of transverse mode in the steady state is the measurement result. During the experiment, a laser diodepumped solid-state laser was used to emit laser beam, and the captured beam spot images were input into the nonlinear network after image processing, and the proportion of fundamental mode was 69 %. The results were used to compose a new spot digital image, and the relative error compared the original image was 3.53%.
出处
《中国激光》
EI
CAS
CSCD
北大核心
2008年第4期555-558,共4页
Chinese Journal of Lasers
基金
国家重点实验室基金(51438010205DZ0101)资助项目
关键词
测量
激光束
横模
非线性网络
CCD
measurement
laser beam
transverse mode
nonlinear network
CCD
