人工神经网络技术在光电检测中的应用  

Applications of Artificial Neural Network to Opto-Electric Measurements

作　　者：张广军[1] 

机构地区：[1]北京航空航天大学自动控制系

出　　处：《北京航空航天大学学报》2001年第5期564-568,共5页Journal of Beijing University of Aeronautics and Astronautics

摘　　要：人工神经网络不仅被广泛用于人工智能、自动控制、机械人、统计学等领域的信息处理中 ,还越来越多地被应用于光电检测 .综述了人工神经网络技术在光电检测中的最新应用进展 ,重点讨论了人工神经网络技术在谷物蛋白质含量近红外在线监测、结构光三维视觉检测、焊缝缺陷射线实时成像识别、表面粗糙度及球半径精密测量应用中的优点和局限性 。Artificial neural network technology is getting wider and wider scale application. Following a brief introduction on neural network, its recent applications to opto-electric measurements were described in detail, including on-line monitoring grain protein based on NIR, structured light based 3D vision inspection, X ray real time detection of welding seam defect, and precision measurements of surface roughness and sphere radius. By summarizing its advantages and limitations in these applications, suggestions and comments were given for further research efforts in this field.

关 键 词：人工神经网络 缺陷检查 表面粗糙率 蛋白质含量 视觉检测 球半径 光电检测 

分 类 号：TP183[自动化与计算机技术—控制理论与控制工程]