图像处理技术在焊接接头硬度测试中的应用  

Application of image processing in hardness measurement of welded joint

作　　者：刚铁[1] 赵雪梅[1] 杨浩[2] 杨双羊 

机构地区：[1]哈尔滨工业大学,150001 [2]北京航空航天大学,100191 [3]海洋石油工程股份有限公司,天津市300452

出　　处：《焊接》2008年第12期1-4,共4页Welding & Joining

摘　　要：硬度是材料的重要力学性能之一,也是衡量焊接接头质量的重要指标之一。针对目前硬度测试需要在目镜下人为测量,受操作者主观因素影响较大,误差较大,多点测量时数据量较大,工作效率极低的实际情况,研究了基于特征提取思想的图像处理技术在硬度测试中的应用,并以搅拌摩擦焊接头各区域硬度测试为例,进行了实验验证。实验结果表明,即使在简易实验设备条件下,通过图像处理技术的辅助,可以极大地提高硬度测量的精度和工作效率。It’s acknowledged that the hardness is an important property of materials; it is also one of crucial parameters in quality control of welded joints. Due to the human error which includes cognitive bias,there is a low inaccuracy in conventional hardness testing,particularly,the operation efficiency was usually limited by mass of data. This paper concerns a study on the applications of image processing in hardness test based on feature extraction technology,and hardness of a friction stir welded joint by both conventional method and feature extraction algorithm were compared. The results show that the method we recommended has higher accuracy and efficiency ,even under the simple experiment conditions.

关 键 词：硬度测量 图像处理 焊接接头 

分 类 号：TG115.51[金属学及工艺—物理冶金]