Abstract: Ear defects are a common type of defect in bar rolling and are difficult to eliminate. In order to achieve intelligent prediction and guide the elimination of ear defects, this article observes that during the rolling process, the part of the bar in contact with air will be oxidized by air, forming a black strip like surface. The narrower the width of the black band, the easier it is to produce ears. However, there is currently almost no research on the automatic measurement of the width of the black belt. In actual production, the measurement of the width of the Huff surface mainly relies on manual visual inspection, which has a high degree of subjective randomness. This article proposes a method based on high-speed cameras to quickly capture images. By recognizing the images, the width of the oxidation black belt on the Huff surface can be intelligently measured, which can provide data support for the optimization of related subsequent process parameters and has important engineering significance for further guiding industrial production.

Key words: Rolling of automotive bars; Width of oxidation black belt; Image processing; Intelligent measurement

摘要： 耳子缺陷是棒材轧制的常见缺陷种类，且不易消除，为实现耳子缺陷的的智能化预判 并指导消除，文章通过观察发现，在轧制过程中，棒材与空气接触的部分会被空气氧化，形 成一个黑色条状表面，而该黑带宽度越窄，越容易产生耳子。但是该黑带宽度的自动测量， 目前几乎没有相关研究。实际生产中，对于哈夫面宽度的测量主要依靠人工目测，主观随机 性较大。文章提出了一种基于高速相机快速抓取图像，通过对图像进行识别，可以对哈夫面 氧化黑带宽度进行智能测量，能够为相关后续工艺参数的优化提供数据支撑，也对进一步指 导工业生产具有重要的工程意义。

关键词: 汽车用棒材轧制；氧化黑带宽度；图像处理；智能测量