中国激光, 2013, 40 (8): 0808002, 网络出版: 2013-07-09   

基于RANSAC-LSSVM回归的惯性约束聚变光学元件损伤在线检测技术

RANSAC-SVM Based Online Inspection Technique for Optics Damage in Intertial Confinement Fusion System
作者单位
1 哈尔滨工业大学电气学院, 黑龙江 哈尔滨 150001
2 中国工程物理研究院激光聚变研究中心, 四川 绵阳 621900
摘要
针对惯性约束聚变(ICF)光学元件损伤问题,提出了一种基于随机抽样一致性(RANSAC)及最小二乘支持向量机(LSSVM)回归的高精度检测方法。建立了损伤区域总灰度与实际尺寸的回归模型,通过该回归模型对待检测损伤区域的尺寸进行预测,得到损伤区域的高精度尺寸。为剔除回归模型建立过程中离群样本点的影响,采用RANSAC方法对训练样本进行优化选择。针对抽样组中样本数对检测精度及检测效率的影响进行了相关实验,确定了抽样组中样本数的合适区间。RANSAC-LSSVM方法可通过改变误差评价函数得到不同评价体系下的最优回归模型。实验证明,在传统像素级检测方法的基础上,该方法将损伤尺寸检测的平均相对误差降低了近90%。
Abstract
A high precision method, based on random sampling consensus (RANSAC) and least squares support vector machine (LSSVM) regression, is advanced to improve the inspection for inertial confinement fusion (ICF) optics damage. In present study, the measurement precision of damaged areas is improved significantly, via analyzing the regression model of grey value and size. RANSAC is adopted for optimizing the training samples to avoid the effect of outliers. In addition, special experiment is proceeded to assess the effect of sample size on the measurement precision and the efficiency, and the appropriate interval of sample size is obtained as well. RANSAC-LSSVM can get optimal regression models which is adopted in different evaluative systems, by using different error evaluation functions. The results show that RANSAC-LSSVM reduces the mean relative error by 90% approximately, in measuring of damaged areas, compared with traditional pixel-level detection.
参考文献

[1] 李大伟, 赵元安, 贺洪波, 等. 光学元件激光损伤阈值的指数拟合法以及测试误差分析[J]. 中国激光, 2008, 35(2): 273-275.

    Li Dawei, Zhao Yuan′an, He Hongbo, et al.. Exponential fitting of laser damage threshold and analysis of testing errors[J]. Chinese J Lasers, 2008, 35(2): 273-275.

[2] 李熙斌, 王海军, 袁晓东, 等. CO2激光辐照对熔石英表面形貌与应力分布的影响[J]. 中国激光, 2011, 38(9): 0903002.

    Li Xibin, Wang Haijun, Yuan Xiaodong, et al.. Effects of CO2 laser irradiation on surface morphology and stress distribution of fused silica[J]. Chinese J Lasers, 2011, 38(9): 0903002.

[3] 王雪, 谢志江, 孙红岩, 等. 大口径精密光学元件表面疵病检测系统研究[J]. 仪器仪表学报, 2006, 27(10): 1262-1265.

    Wang Xue, Xie Zhijiang, Sun Hongyan, et al.. Study on automatic flaw inspection system for large caliber precision optical surface[J]. Chinese J Scientific Instrument, 2006, 27(10): 1262-1265.

[4] 邓燕, 许乔, 柴立群, 等. 光学元件亚表面缺陷的全内反射显微检测[J]. 强激光与粒子束, 2009, 21(6): 835-840.

    Deng Yan, Xu Qiao, Chai Liqun, et al.. Total internal reflection microscopy: a subsurface defects identification technique in optically transparent components[J]. High Power Laser and Particle Beams, 2009, 21(6): 835-840.

[5] 彭志涛. 强激光复杂光机组件光学元件激光损伤在线检测技术研究[D], 绵阳: 中国工程物理研究院,2011. 1-8,16-19.

[6] 于忠党, 王龙山. 基于回归型最小二乘支持向量机卷积模板的椒盐噪声开关滤波器[J]. 光学学报, 2008, 29(1): 163-168.

    Yu Zhongdang, Wang Longshan. Salt pepper noise switching filter based on LS-SVR convolution mask[J]. Acta Optica Sinica, 2008, 29(1):163-168.

[7] 郭志明, 陈立平, 黄文倩, 等. 基于FLS-SVM背景预测的红外弱小目标检测[J]. 激光与光电子学进展, 2012, 49(2): 021201.

    Guo Zhiming, Chen Liping, Huang Wenqian, et al.. Application of genetic algorithm-least squares support vector regression with near infrared spectroscopy for prediction of nicotine content in tobacco[J]. Laser & Optoelectronics Progress, 2012, 49(2): 021201.

[8] 赵树森,陈思嘉,沈京玲. 用支持向量机识别毒品的太赫兹吸收光谱[J]. 中国激光, 2009, 36(3): 752-757.

    Zhao Shusen, Chen Sijia, Shen Jingling. Identification of terahertz absorption spectra of illicit drugs using support vector machines[J]. Chinese J Lasers, 2009, 36(3): 752-757.

[9] 刘贵喜, 刘冬梅, 刘凤鹏, 等. 一种稳健的特征点配准算法[J]. 光学学报, 2008, 28(3): 454-461.

    Liu Guixi, Liu Dongmei, Liu Fengpeng, et al.. A robust image registration algorithm based on feature points matching[J]. Acta Optica Sinica, 2008, 28(3): 454-461.

[10] 张宁, 常雷, 徐熙平. 基于机器视觉的三维重建技术研究[J]. 激光与光电子学进展, 2012, 49(5): 051001.

    Zhang Ning, Chang Lei, Xu Xiping. Research on the technology of three-dimensional reconstruction based on machine vision[J]. Laser & Optoelectronics Progress, 2012, 49(5): 051001.

[11] 张际, 李大海. 光学元件损伤暗场成像检测的算法[J]. 中国激光, 2006, 33(8): 1109-1111.

    Zhang Ji, Li Dahai. Algorithm of optics damage inspection from its dark-field image[J]. Chinese J Lasers, 2006, 33(8): 1109-1111.

[12] 解亚平,孙志红,成泽, 等. 光学元件损伤在线检测中的图像处理[J]. 强激光与粒子束,2006,18(7): 1085-1089.

    Xie Yaping, Sun Zhihong,Cheng Ze, et al.. Image processing in online inspection of damage in optics[J]. High Power Laser and Particle Beams, 2006,18(7): 1085-1089.

[13] Feng Bo, Chen Fengdong, Liu Bingguo. Segmentation of small defects in final optics damage online inspection images[C]. IEEE International Conference on Image Analysis and Signal Processing, 2012. 110-113.

[14] 彭志涛, 魏晓峰, 元浩宇, 等. 全内反射照明光学元件损伤检测信噪比分析[J]. 红外与激光工程, 2011, 40(6): 1111-1114.

    Peng Zhitao, Wei Xiaofeng, Yuan Haoyu, et al.. Signal noise ratio of total internal reflection edge illumination for optics damage inspection[J]. Infrared and Laser Engineering, 2011, 40(6): 1111-1114.

[15] 肖冰. 大口径光学元件表面疵病自动化检测系统关键问题讨论与研究[D]. 杭州:浙江大学, 2010.

    Xiao Bing. Research on the Key Issues of Automatic Defects Measurement on Large Fine Optics [D]. Hangzhou: Zhejiang University, 2010.

[16] 沈沁梅, 周卫东, 李科学. 基于遗传神经网络的激光诱导击穿光谱元素定量分析技术[J]. 中国激光, 2011, 38(3): 0315001.

    Shen Qinmei, Zhou Weidong, Li Kexue. Quantative elemental analysis using laser induced breakdown spectroscopy and neuro-genetic approach[J]. Chinese J Lasers, 2011, 38(3): 0315001.

冯博, 陈凤东, 刘炳国, 彭志涛, 刘楠, 刘国栋, 孙和义. 基于RANSAC-LSSVM回归的惯性约束聚变光学元件损伤在线检测技术[J]. 中国激光, 2013, 40(8): 0808002. Feng Bo, Chen Fengdong, Liu Bingguo, Peng Zhitao, Liu Nan, Liu Guodong, Sun Heyi. RANSAC-SVM Based Online Inspection Technique for Optics Damage in Intertial Confinement Fusion System[J]. Chinese Journal of Lasers, 2013, 40(8): 0808002.

本文已被 3 篇论文引用
被引统计数据来源于中国光学期刊网
引用该论文: TXT   |   EndNote

相关论文

加载中...

关于本站 Cookie 的使用提示

中国光学期刊网使用基于 cookie 的技术来更好地为您提供各项服务,点击此处了解我们的隐私策略。 如您需继续使用本网站,请您授权我们使用本地 cookie 来保存部分信息。
全站搜索
您最值得信赖的光电行业旗舰网络服务平台!