中药饮片切面纹理特征提取研究

目的:研究中药饮片切面纹理特征提取方法。方法:利用基于Tamura方法研究中药饮片切面纹理粗糙度、对比度、方向度、线性度、规则度、粗略度等特征参数提取。结果:由于Tamura方法更加符合人眼视觉习惯与心理感知,对比试验表明多种参数结合的Tamura方法能够较好的描述中药饮片切面纹理特点。结论:该方法可以局部反应纹理局部细节,从而实现了将视觉感知用科学、定量的形式表示,为基于机器视觉的中药饮片自动辨识及分级等应用提供理论依据,由于受到饮片颜色、饮片背景、饮片形状等因素的影响,基于Tamura纹理方法提取中药饮片切面纹理特征参数,需要多种特征参数融合才能够较为精确的表示纹理特征。

Texture Feature Extraction Research of Sections of Traditional Chinese Medicine

Objective:To study the texture feature extraction method of the cut crude TCM.Methods:The parameters of the cut crude TCM was studied in the aspect of texture coarseness,contrast,directionality,linelikeness,regularity and roughness based on the Tamura method.Results:because the Tamura method is more in line with the human visual habits and psychological perception,the comparison experiment shows that a variety of parameters of the Tamura method can be better to describe the characteristics of t cut crude TCM.Conclusion:The method can be used to extract the cut crude TCM local texture details,enabling the visual perception with a scientific,quantitative representation,and it provides theoretical basis for machine vision-based the cut crude TCM automatic identification and classification,due to cut crude shades,cut crude the impact of factors such as the color,the background,the shape of cut crude,In order to represent more accurately the extracted texture features,the integration of a variety of algorithms will be a good method.