基于加权最小二乘法的纸币图像倾斜校正方法  

Banknote Image Skew Correction Method Based on Weighted Least Squares

作　　者：奕科杰 薛凌云[1] YI Kejie XUE Lingyun(School of Life Information and Instrument Engineering, Hangzhou Dianzi University, Hangzhou Zhejiang 310018, China)

机构地区：[1]杭州电子科技大学生命信息与仪器工程学院,浙江杭州310018

出　　处：《杭州电子科技大学学报（自然科学版）》2017年第3期68-72,共5页Journal of Hangzhou Dianzi University：Natural Sciences

摘　　要：提出了一种基于加权最小二乘法的纸币图像倾斜校正方法.根据纸币图像特点对行列进行扫描并获取一定数量纸币边缘点,通过加权最小二乘法进行边缘直线拟合以获取纸币倾斜角度,利用倾斜角度实现了纸币图像的倾斜校正,确保了纸币图像特征区域正确分割提取.仿真结果表明,该算法检测精度较高,计算量小且抗干扰性好,对各20组纸币无残缺与残缺边缘倾斜检测平均绝对误差分别为0.06°和0.45°,平均耗时为1.1ms,有效提高了纸币图像倾斜检测精度与效率.This paper presents a method of banknote image skew correction based on weighted least squares. According to the characteristics of banknote image, it extracts a number of banknote edge points with line scan and row scan. Then by using weighted least square method to fit the edge points, it calculates the edge linear equation and the tilt angle of the edge line. Finally, it uses the tilt angle to realize tilt correction to ensure that feature area of banknote image can be correctly extracted. As a result, the experiment results show that the algorithm which has advantages of high detection accuracy, low computational complexity and good anti-disturbance ability. The mean absolute error of 20 groups of banknote image samples with non-defective edge and defective edge is 0.06 degree and 0.45 degree respectively, the average detection time is 1. lms. It can effectively improve accuracy and efficiency of banknote image tilt detection.

关 键 词：纸币图像 加权最小二乘法 倾斜校正 

分 类 号：TP274[自动化与计算机技术—检测技术与自动化装置]