用BP神经网络构成三维视觉测量系统的新方法  

A new method for 3-D vision measurement system by the use of BP neural network

作　　者：吕乃光[1] 冯迪[1] 邓文怡 桑新柱 

机构地区：[1]北京机械工业学院电子信息工程系

出　　处：《华中科技大学学报（自然科学版）》2002年第5期19-21,共3页Journal of Huazhong University of Science and Technology(Natural Science Edition)

基　　金：北京市教育委员会科技发展计划资助项目 (99KJ4 3)

摘　　要：提出一种基于人工神经网络的新的三维视觉测量方法 .建立在三角测量原理上的传统的三维立体视觉测量技术会带入成像畸变非线性误差 ,这种新方法可以消除非线性因素的影响 .采用正弦函数作为传递函数加快了网络的训练速度 ,提高了学习精度 ,对空间点位和尺寸测量的结果验证了方法的可行性 .对影响系统测量精度的因素进行了讨论 ,并与传统立体视觉测量方法进行了比较 .A new method using an artificial neural network for 3 D vision measurement is presented. The traditional method for 3 D vision measurement based on the principle of trigonometry introduces the nonlinear error because of imaging distortion, and the new method can avoid the influence of nonlinearity. The training speed and precision are improved through the introduction of sine function as transfer function. The experiments have been completed. From the results the feasibility of this method is proved. The factors affecting accuracy of measurement system are discussed and the comparison with the traditional stereo vision system is presented.

关 键 词：三维测量 三维视觉 人工神经网络 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]