基于区域生长的电路板图像分割算法的研究

在印刷电路板质量检测中,自动光学检测应用越来越普遍。其原理是采集印刷电路板的图像,运用数字图像处理技术对图像进行分析,由计算机对电路板制造中产生的缺陷进行自动识别。在实际中,由于电路板材质不均匀的特点,采集到的图像包含很多背景噪声,难以进行缺陷识别。本文介绍了一种基于区域生长的电路板图像分割算法。经实验证明,运用该方法对电路板图像进行分割,可以完整地保留电路板的特征,而有效的除去背景噪声。

Research of printed circuit board image segmentation based on region growing

In printed circuit board inspection system,automated optical inspection is more and more adopted.The principle is that according to the analysis results of printed circuit board image by using digital image processing methods,the computer will point out the defects automatically.Practically,because of the uneven feather of the board material,the image has much background noise,and the defects can't be found.The method using region growing in printed circuit board image segmentation is presented in this paper.The experiments indicate that the method extracts the characteristics well and removes the background noise effectively.