摘要
轮胎生产过程中广泛采用纠偏系统防止半成品胶带发生跑偏。纠偏系统是一个实时或准实时的系统,其性能的好坏主要取决于图像处理模块。本文为了更好地保留图像信息,减少噪声对图像的影响,首先利用小波包对图像的低频和高频部分进行分解;然后,采用Birge-Massart惩罚函数确定的阈值对胎面图像进行有效去噪;接着,利用Robert算子进行边缘检测,提取出胎面边缘和机架边缘;最后,经过算法计算出胎面的偏移角度。
The deviation-correcting system is widely used in the tire production process because the semi-finished tire products often turn aside from the center line. The whole system is a real-time or part real-time system, and its quality is mainly decided by the imagery processing module. In order to preserve the information of tire images and decrease the noise pollution in this paper,a wavelet packet is used to split the low- and high-frequency part of the images first, then thresholding is used to decrease the noise of images effectively, which is determined through the Brige-Massart punishment function, and the Robert operator is used to detect tire edges and their characteristics. Finally the displacement angle is computed through the algorithm.
出处
《计算机工程与科学》
CSCD
北大核心
2009年第3期52-54,共3页
Computer Engineering & Science
关键词
图像压缩
小波包去噪
边缘检测
特征提取
image compression
wavelet packet de-noising
edge detect
feature extraction
