基于机器视觉的发动机包覆层表面缺陷检测技术

Detecting surface defects of rocket motor cladding based on machine vision

固体火箭发动机包覆层表面的缺陷检测是保证其使用安全性的重要手段。本文应用机器视觉技术实现发动机包覆层表面缺陷的自动检测。用工业相机采集发动机包覆层表面的图像,结合图像特点,通过多次实验分析,选择中值滤波方法进行滤波,采用图像模式匹配技术将缺陷从图像背景中分割出来,并将以上功能进行整合,设计缺陷自动检测系统。实验结果表明该方法在发动机包覆层表面缺陷检测方面有较好的效果。

Surface-defect detection of rocket motor cladding is an important way to ensure the safety in application. A method for detecting surface defects of rocket motor cladding based on machine vision is proposed in this paper. The surface images of rocket motor cladding are collected by industrial cameras. Combining image characteristics and analyzing experiment results, the median filter is chosen to denoise. By adopting the image pattern matching techniques, the defects are segmented from background. Based on the above operation , a system for detecting surface defects of rocket motor cladding is designed. The experiment results show the proposed detection method is feasible and effective