计算机视觉在金刚石净度检测中的应用研究

Research on application of computer vision in the detection of diamond clarity

净度是指金刚石所包含的杂质、裂隙及生长纹等的多少。金刚石的机械性质与晶体内部的杂质含量之间存在着重要的相互关系,为此给出了一种用计算机视觉系统代替人眼对金刚石净度进行自动检测的方法。通过对金刚石彩色图像和瑕疵的颜色特性的分析,按照视觉心理感知特性选择HIS色彩空间,结合色调H和饱和度S作为特征参数进行识别,实现了对金刚石净度的自动检测。经过MAT-LAB仿真,实验结果表明:随着瑕疵的增多,该系统的准确率可高达98%,为金刚石净度检测提供了一条新途径。

Clarity is an important factor in the evaluation of the quality of diamond.Usually the clarity of diamond is determined through experienced technicians.This means of measurement is imprecise,inconsistent and time consuming.A computer vision system which can test automatically to replace the human eyes was developed.Through the analysis of the diamond color images and color characteristics of the flaws,hue and saturation was chosen as the parameters.The system is also capable for clarity grading.The results ...