摘要
从贝叶斯推理的角度提出一种计算线性模型参数的方法,并将该方法应用到高速纸币清分系统的纸币图像边界检测中,实验结果表明,可以通过少量观测数据实现图像边界线性模型的参数估计,从而完成边界的快速检测。检测精度高,同时针对纸币边缘破损的情况,提出了有效避免旧纸币边缘破损对纸币定位影响的方法。
A linear model parameter estimation method is proposed based on Bayesian treatment, we estimate the linear model parameter of image edge use this method. It was used in high speed paper currency image edge detection. Experiment result show that it can finish the parameters fast estimation and the detection of image edge depend on few observed data with a high detect precision. At the same time , a method avoid the negative impact of currency＇s broken edge is proposed to solve the problem.
出处
《机电一体化》
2010年第6期80-83,共4页
Mechatronics
关键词
线性模型
纸币清分
边界检测
linear model currency sorter edge detection
