一种基于图像的轮胎表面破损识别算法  

Based on Image Tire Surface Subtle Crack Identification Technology

作　　者：吕聪颖[1] 赵刚彬[1] 

机构地区：[1]南阳理工学院计算机与信息工程学院,河南南阳473000

出　　处：《科技通报》2012年第12期204-207,共4页Bulletin of Science and Technology

基　　金：国家自然科学基金青年科学基金项目(81101490)

摘　　要：针对由于车辆轮胎的裂痕特征在像素上与轮胎本体极为相似,当裂痕比较细微的情况下,形态特征的采集也极为困难,传统的基于像素差异和形态检测的方法对轮胎的细微裂痕检测效果一直很低的问题,提出了基于图像的轮胎表面细微裂痕识别技术。通过去除图像中的噪声和孤立点对测量准确率的影响,利用不规则扭曲像素裂痕映射检测方式,通过对轮胎图像像素变形和缺失的准确测量,实现轮胎表面细微裂痕的检测。实验证明,这种检测方式能够快速、准确地检测出轮胎表面的细微裂痕,取得了令人满意的效果。For as the vehicle tyre cracks characteristics in pixel and tyre is very similar,when cracks under the condition of the more subtle,morphological characteristics of the acquisition also extremely difficult,the traditional based on pixel differences and form detection method of tiny cracks on the tyres detection effect has been very low.In order to solve the appeal,put forward the tire surface based on image subtle crack identification technology.Through the remove the noise of the image and isolated to the influence of the measurement accuracy point,of irregular twisted pixel crack detection mode,through to the tyre pixels deformation and the lack of accurate measurement,realize the tire surface crack detection subtle.Experiments show that the detection means to be able to rapid and accurate detection of subtle tire surface crack,satisfactory results have been obtained.

关 键 词：细微裂痕 轮胎检测 扭曲像素 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]