微分算法在异性纤维分拣中的应用

Differentiation Algorithm in Application of Detection of Foreign Fiber

图像处理技术是机器视觉研究中的关键技术之一,为了更加有效地分拣棉花中的头发丝,提出了微分算法。该方法结合棉花阈值范围和线阵像机的特性,在最小二乘法的基础上提出了专门针对头发丝的另外一种算法。实验结果表明,该方法能够有效地清除棉花中存在的头发丝,很大程度地提高了系统的拣出率。

The technology of the image processing is the first of key technology in the visual research of the machine,for the purpose of detecting the hair in cotton,puts forward a method of differentiation algorithm in the detection of foreign fiber. According to this method,the distribution of cotton in the RGB color model can be used,combining threshold of cotton with the linear array camera＇s characteristic,advances the arithmetic of combination with upper and lower thresholds with the non-linear threshold,and proposes a specifical method for hair.