| 摘要: |
| 目的：利用机器视觉技术建立苦豆子颗粒外观质量评价方法，并对不同苦豆子颗粒进行外观质量评价。方法：利用机器视觉技术通过对苦豆子颗粒进行数字图像采集、特征提取、特征降维、对照特征指纹向量建立、相似度计算等，建立苦豆子颗粒外观对照特征指纹向量，将参比颗粒外观特征图谱与苦豆子颗粒对照指纹图谱进行相似度计算，以相似度值为标准对参比颗粒外观质量进行评价。结果：建立了苦豆子颗粒对照特征指纹向量，通过不同计算方法相似度准确性评价确定巴氏距离法为本研究相似度计算方法，对参比标准颗粒相似度进行计算。对不同主药含量、不同粒度、不同水分的苦豆子颗粒作为参比颗粒分别进行了外观质量评价，结果显示3类参比苦豆子颗粒外观相似度均显著低于标准苦豆子颗粒相似度平均值。根据参比苦豆子颗粒相似度计算结果与中国药典有关要求的相关性情况，将0.8700作为苦豆子颗粒外观相似度评价的合格值。结论：本研究建立了基于机器视觉技术的苦豆子颗粒外观质量评价方法，确定了外观相似度评价合格值。该方法具有设备简单、操作简便、快速准确等优点，可以用于苦豆子颗粒生产中产品外观质量评价，并通过外观质量评价对苦豆子颗粒剂进行质量控制。 |
| 关键词: 机器视觉 苦豆子颗粒 外观质量评价 对照特征指纹向量 |
| DOI： |
| 投稿时间：2020-09-13修订日期：2021-03-12 |
| 基金项目:宁夏重点研发计划重大（重点）项目（2018BFG02006）；宁夏中药材开发与利用人才小高地项目 |
|
| Study on appearance Quality evaluation of Sophora alopecuroides granules based on machine vision technology |
|
| (ingxia Polytechnic) |
| Abstract: |
| Objective: The exterior quality evaluation method of Sophora alopecuroides granules was established by machine vision technology and the exterior quality evaluation of different bitter bean granules was carried out..Method: Alopecuroides particles to make use of machine vision technology through digital image acquisition, feature extraction, feature dimension reduction, contrast feature vector to establish fingerprint, similarity calculation, etc., to establish fingerprint grain appearance of Sophora alopecuroides contrast characteristics vector and the reference particle appearance feature maps and Sophora alopecuroides particle control fingerprint similarity calculation, on the basis of similarity value for reference to evaluate grain appearance quality.Result：The comparison characteristic fingerprint vector of Sophora alopecuroides granules was established, and the pasteurian distance method was used to calculate the similarity of reference standard granules by evaluating the similarity accuracy of different calculation methods.The appearance quality of Sophora alopecuroides granules with different contents, particle size and water content as reference particles was evaluated, and the results showed that the appearance similarity of the 3 kinds of reference Sophora alopecuroides granules was significantly lower than the average value of the standard.According to the correlation between the calculated results of reference Sophora alopecuroides granules and the relevant requirements of Chinese pharmacopoeia, 0.8700 was taken as the qualified value for the evaluation of appearance similarity of Sophora alopecuroides granules.Conclusion: In this study, the appearance quality evaluation method of Sophora alopecuroides granules based on machine vision technology was established and the qualified value of appearance similarity evaluation was determined. This method has the advantages of simple equipment, simple operation, rapid and accurate, and can be used to evaluate the product external appearance quality in the production of Sophora alopecuroides granules, and to control the quality of Sophora alopecuroides granules through the external appearance quality evaluation. |
| Key words: Machine vision Sophora alopecuroides granules Appearance quality evaluation Contrast feature fingerprint vector |
