一种汽车刹车油管环形接头缺陷检测新方法  

A New Method for Detecting Defect of Annular Joint of Automotive Brake Fluid Pipe

作　　者：张泽文 郑志军 白永洁 徐值 余卉卉 张巨豪 殷安民[1,2] Zhang Zewen;Zheng Zhijun;Bai Yongjie

机构地区：[1]宁波大学机械工程与力学学院,浙江宁波315211 [2]浙江省零件轧制成形技术研究重点实验室,浙江宁波315211 [3]宁波亚大自动化科技有限公司,浙江宁波315143 [4]宁波福士汽车部件有限公司,浙江宁波315145

出　　处：《机械制造》2023年第10期63-69,共7页Machinery

基　　金：浙江省自然科学基金资助项目(编号:LY21E050004);宁波市科技创新2025重大专项(编号:2022Z055);宁波大学产学研合作项目(编号:HK2021000304)。

摘　　要：汽车刹车油管环形接头是刹车系统的重要安全零部件,制造过程中产生的缺陷形貌尺寸较小,表面需经镀膜处理,具有表面高反光特性,导致微小缺陷检出困难。对此,提出一种汽车刹车油管环形接头缺陷检测新方法。采用一种新型感兴趣区域融合差值法,基于图像融合原理,针对环形感兴趣区域进行不同图层的权值计算,保证融合的精准度。通过引入拉普拉斯锐化及差值处理,增强对缺陷区域的表征。利用中值滤波对椒盐噪声的作用特性,实现图像噪声抑制。通过不同方法不同缺陷图像的处理结果分析,确认所提出的新方法可以有效提高图像对比度,抑制非缺陷噪声,具有更好的输出质量。The annular joint of the automotive brake fluid pipe is an important safety component of the brake system.The defect generated during the manufacturing process is small in size,the surface needs to be coated,and has a highly reflective property,which makes it difficult to detect small defect.In this regard,a new method was proposed to detect the defect of the annular joint of the automobile brake fluid pipe.Based on the principle of image fusion,a new interest region fusion difference method was used to calculate the weight of different layers for the annular region of interest to ensure the accuracy of fusion.The characterization of defective region was enhanced by introducing Laplacian sharpening and difference processing.Using the characteristic of median filter on pretzel noise,image noise suppression was realized.By analyzing the processing results of different defective images with different methods,it is confirmed that the proposed new method can effectively improve the image contrast,suppress the non-defective noise and has better output quality.

关 键 词：刹车油管 环形接头 缺陷检测 

分 类 号：U463.55[机械工程—车辆工程]