基于亚像素差异化判断的珠宝光泽度的判断

Judgment of Jewelry Gloss Based on the Subpixel Differentiation

在对珠宝光泽度的判断过程中,采用传统的方法判断时,由于传感器的离散性和噪声的干扰,造成了像素级误差过大的问题。提出了基于亚像素差异化判断的珠宝光泽度判断方法。利用分阶段的高精度亚像素特征点提取方法,将珠宝光泽度图像边缘特征离散作为亚像素级特征点,组建了分类规则,根据Dempster组合规则满足高置信度、低冲突性的情况,将其融合于加权分配冲突法中,并解决证据冲突的问题,从而完成了对珠宝光泽度的精确判断。实验仿真证明,提出的基于亚像素差异化判断的珠宝光泽度的判断方法,精准度高,实用性强。

In the process of judgment of jewelry gloss, adopt the method of traditional judgment, due to the discreteness of sensors and noise interference, has created the problem of pixel level error is too large. Therefore, is proposed based on subpixel differentiation judge jewelry gloss judgment method. Using phased high precision sub-pixel feature point extraction method, the jewelry gloss discrete image edge character as a pixel level and feature point, set up the classification rules, according to the rules of Dempster combination to meet high confidence level, low conflicting situation, its integration in weighted allocation conflict law, and solve the problem of conflict evidence, thus complete the accurate judgment of jewelry glossiness. Experimental simulation show that the proposed based on pixel differentiation and judge jewelry gloss judgment method, high accuracy, strong practicability.