基于图像钢球缺陷检测有效面积分析

摘要

Based on the image detection of steel ball surface defects,the accuracy and efficiency of detection are seriously impacted by the effective area selected improperly.Firstly,the surface detection of steel ball optimal range is studied to determine the ideal arc length;secondly,in view of the sphere zone theory establishes effective area and zonal overlapping relationship model,in order to restrict detection times by reducing the overlap area;finally,the formula of effective arc covering efficiency is utilized to direcdy control arc length redundancy caused by the overlap area to improve the detection efficiency and laid the important theoretical basis for system development of ball diameter series detection.%基于图像检测钢球表面缺陷时有效面积选择不当严重影响检测精度和效率,首先对钢球表面检测最佳范围进行研究确定理想弧长;其次基于球带理论建立了有效面积与球带重叠关系模型,以此控制展开次数来降低重叠面积;最后利用有效孤长覆盖效率公式来直观地控制孤长冗余量所带来的重叠面积,达到提高检测效率的目的,并为钢球直径系列化检测系统开发奠定了重要理论基础.