摘要
针对目前国内铸坯表面缺陷检测方法落后、检测效率低的情况,应用图像处理技术,设计了铸坯表面缺陷自动检测系统方案.研究了适合高温高辐射条件下的图像采集方案和算法,采用基于BP神经网络的模式识别方法对铸坯表面缺陷图像进行识别与分类,能够有效地提高铸坯质量管理.
Aim at the unreliable and low efficient method on billet inspection,this paper proposed an image processing approach to inspect billet surface defect automatically.This approach also proposed the image acquisition solution and the algorithm in the high temperature and high radiation environment.With BP neural network,the system can recognize and classify the billet surface defect automatically,improve the billet quality management effectively.
出处
《机械与电子》
2010年第12期38-41,共4页
Machinery & Electronics
关键词
图像处理
表面缺陷
缺陷检测
BP神经网络
image processing
surface defect
defect inspection
BP neural network
