...
  • 主题
高级搜索  >
历史搜索 清空历史
首页> 外文期刊>Journal of Residuals Science & Technology >Research on Intelligent Prediction of the Optical Lens Module’s Vision-based Assembly Quality
【24h】

Research on Intelligent Prediction of the Optical Lens Module’s Vision-based Assembly Quality

机译：基于光学镜头模块视觉组装质量的智能预测研究

获取原文
           
页面导航
  • 摘要
  • 著录项
  • 相似文献
  • 相关主题

摘要

: The assembly quality of the optical lens module has an important influence on the optical image quality.This paper focuses on the optical lens module’s quality prediction with vision-based compensation assembly and develops a PSO-BP neural network to predict the vision-based compensation assembly quality with on-line adaptation to the changing assembly conditions.Firstly,in the optical lens module’s assembly system based on machine vision compensation,the image information is used to predict the optical lens module’s assembly accuracy. Image information, in a certain extent,not only reflects the degree of the optical lens module’s assembly positioning error, image information is also easy to be integrated with the computer program, such as artificial intelligence, expert system,and so on. Therefore, this paper selects the main image parameters to predict the optical lens module’s assembly quality.Secondly,a PSO-BP neural network algorithm is proposed to predict the optical lens module’s assembly quality,which is an improved BP network algorithm based on PSO learning mechanism.The particle’s position is used to change the weight of BP neural network ,the particle’s velocity is used to update the weights,and avoid the intrinsic defects of neural network such as the slow convergence rate and the poor global covergence. Thirdly,samples are applied to train the PSO-BP neural network,and conduct the optical lens module’s assembly quality prediction experiments.Experimental results show that the PSO-BP neural network not only predicts the optical lens module’s vision-based compensation assembly quality with a less error, but also is not sensitive to the random initial population.
机译：光学透镜模块的组装质量对光学图像质量有重要影响。本文重点研究了基于视觉的补偿组件对光学透镜模块的质量预测，并开发了PSO-BP神经网络来预测基于视觉的补偿在线适应不断变化的装配条件的装配质量。首先，在基于机器视觉补偿的光学透镜模块的装配系统中，图像信息用于预测光学透镜模块的装配精度。图像信息在一定程度上不仅反映了光学透镜模块的装配定位误差的程度，而且图像信息还易于与计算机程序集成在一起，例如人工智能，专家系统等。因此，本文选择主要的图像参数来预测光学透镜模块的组装质量。其次，提出了一种PSO-BP神经网络算法来预测光学透镜模块的组装质量，它是一种基于PSO学习机制的改进的BP网络算法。粒子的位置用于改变BP神经网络的权重，粒子的速度用于更新权重，避免了神经网络固有的缺陷，例如收敛速度慢和全局覆盖性差。第三，通过样本训练PSO-BP神经网络，进行光学透镜模块的装配质量预测实验。实验结果表明，PSO-BP神经网络不仅可以预测光学透镜模块的视觉补偿装配质量。误差较小，但对随机初始种群不敏感。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱：kefu@zhangqiaokeyan.com

京公网安备：11010802029741号 ICP备案号：京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有

  • 客服微信

  • 服务号