摘要
针对轮胎图像高噪声的特点,本文提出一种轮胎规格号的提取方法。首先,采用最小二乘法拟合迭代圆心,取得圆心位置,采用极坐标变换、双线性插值拉伸对轮胎图像进行预处理;随后,采用灰度加窗变换方法,降低图像维度,构造锯齿波;最后,利用锯齿波参数获取规格号字符。实验结果表明:该方法能够有效的提取轮胎规格号。
A method of tire specification extraction was presented. First, in image pre-processing, iterative least square method was used to fit accurate positioning of the circle center, and coordinate transformation and bilinear interpolation stretching were used to preprocess tire image. Then, window transform was used to reduce image dimensions from two-dimensional to one-dimensional vector, and thus one-dimensional vector was proposed to construct the sawtooth with extreme points. Finally, by using the parameters, the characteristics of the waveform were described to extract character region of tire specifications. Experimental results show that this method can effectively extract the tire specification.
出处
《光电工程》
CAS
CSCD
北大核心
2010年第3期24-28,共5页
Opto-Electronic Engineering
关键词
积分图像
加窗变换
锯齿波
轮胎规格号
integral image
window transform
sawtooth
tire specification