中文摘要：

烟叶人工分级易受主观因素影响,分级结果存在主观差异性,需要建立与人眼视觉感受相吻合的客观分级方法.本文针对烟叶颜色属性进行分级判断,给出一种基于CIEDE2000的均匀色差颜色分级方法.该方法包含了亮度、色调和饱和度综合色差视觉感受的图像评价标准,将待测烟叶与标准烟叶库中所有标准烟叶分别做色差计算,根据距离准则,寻找与当前待测烟叶色差最小的标准烟叶,同时对这个最小距离进行阈值判定,来确定当前待测烟叶的等级.实验结果表明根据CIEDE2000色差公式的颜色分级方法优于RGB颜色分级方法,与人眼视觉感受更为吻合,为建立烟叶自动分级系统提供了新的依据.

英文摘要：

Tobacco artificial classification is susceptible by subjective elements, the classification results are subjective differences, and therefore need to adopt an intelligent classification approach which is not influenced by graders subjective factors and consistent with the perception of the human vision. The paper mainly classify the tobacco by color , and presents a method based on CIEDE2000 uniform color difference. The method contains the brightness, hue and saturation integrated visual perception of color image evaluation standard, will be under test with the standard repository of tobacco leaf were all standard tobacco color difference calculation, according to the distance criterion, looking for and the current tobacco leaf under test minimum standard tobacco color, at the same time on the minimum distance threshold value to judge, to determine the current level of the tobacco leaf under test. The current level of tobacco to be measured by looking for the smallest chromatic aberration. The experimental results show that CIEDE2000 color difference formulas is better than the RGB color classification method, which is more consistent with the human visual experience, and provides a new basis for the establishment of tobacco automatic grading system.