CAJ | 학술논문

为实现工业现场中轴承防尘盖表面缺陷的自动检测，提出一种基于机器视觉技术的检测方法。采用蓝色同轴光源作为检测系统所用光源，克服金属反光；采用最小二乘法拟合轴承外圆，根据轴承型号比例分割出防尘盖区域，利用Otsu阈值分割和Roberts边缘提取处理图像，每2°统计值为1的点的数目，与模板轴承此数据比较，求出相差角度，由此将防尘盖字符、非字符区域分离，两部分是否存在缺陷分开判别，互不干扰。实际测试表明：检测系统采集到的轴承图像清晰，缺陷检测算法正确率在96%以上，可实现轴承防尘盖表面缺陷的自动检测。
위실현공업현장중축승방진개표면결함적자동검측，제출일충기우궤기시각기술적검측방법。채용람색동축광원작위검측계통소용광원，극복금속반광；채용최소이승법의합축승외원，근거축승형호비례분할출방진개구역，이용Otsu역치분할화Roberts변연제취처리도상，매2°통계치위1적점적수목，여모판축승차수거비교，구출상차각도，유차장방진개자부、비자부구역분리，량부분시부존재결함분개판별，호불간우。실제측시표명：검측계통채집도적축승도상청석，결함검측산법정학솔재96%이상，가실현축승방진개표면결함적자동검측。
To realize the automatic detection of bearing shield surface, this paper proposes a method based on machine vision. It uses the blue coaxial light to overcome the metal reflection;it uses the least squares method to fit the bearing outer circle. According to the bearing type, it segments the bearing. Using Otsu’s method and Roberts edge extraction it processes the shield image. It calculates the points when value is 1 per 2°. Compared with the template data, it obtains the phase angle, then separates the character region and no-character region;there is no interference when the two parts defect is detected. Experiments show that:the captured image is unambiguous and the correct rate of detection algorithm is more than 96%. It can realize the automatic detection of bearing shield surface.