灰度分布曲线与冰勺头部劈裂检测

Gray Distribution Curve and Splitting Detection of Ice Scoop Head

冰勺头部劈裂会随其产生部位光照程度和开口大小对检测效果造成差异。为进一步消除现有方法对冰勺头部两侧光照较弱以及开口较小劈裂的漏检误检情况,设计一种基于灰度分布曲线的劈裂检测算法。算法在垂直方向计算冰勺头部的灰度均值,再对冰勺头部进行灰度校正并提取凹线段绝对幅值,最后基于准则计算劈裂位置的特征阈值。实验结果表明,所设计算法降低了劈裂的漏检率,能够提高冰勺出售前对劈裂缺陷的检测效果,消除安全隐患,具有一定的实际应用价值。

The splitting of the head of the ice scoop will have different detection results with the light intensity and opening size of the position where it is produced.In order to further eliminate the false detection of weak illumination on both sides of the ice scoop head and small opening in the existing methods,a split detection algorithm based on gray distribution curve is designed.The algorithm calculates the average gray level of the ice scoop head in the vertical direction,then corrects the gray level of the ice scoop head and extracts the absolute amplitude of the concave line segment.Finally,the characteristic threshold of the splitting position is calculated based on the criterion.The experimental results show that the designed algorithm can reduce the missed detection rate of splitting,improve the detection effect of splitting defects before the ice scoop is sold,and eliminate potential safety hazards,which has certain practical application value.