基于机器视觉的板材表面缺陷检测方法

Research on Wood Surface Detection Based on Computer Vision

基于机器视觉技术对板材表面缺陷检测进行研究,设计了在线检测的硬件系统和软件处理流程,介绍了图像预处理、图像分割和目标提取等处理方法并对缺陷图像进行相应处理。利用C#软件对板材缺陷在线检测进行人机交互界面设计和在线调试及在线检测。实验结果表明,本检测方法可行,误检率低,可很好地应用于木材自动化生产过程的在线检测。

This paper presents a study into board surface defect detection based on machine vision technology.On-line detection hardware system and software processing procedure were designed in the study. This paper also described image preprocessing, image segmentation, and object extraction techniques and covered appropriate processing of defect images. C# software was utilized to develop MMI interfaces for on-line board surface defect detection. On- line debugging and on- line detection were carried out. It has been proven by experimentation that this detection approach was feasible and registered a low error detection rate, and may be thus used with advantage in on-line detection during automatic wood production process.