基于神经网络的芳香类中草药品质检测技术

Research on Aromatic Chinese Herbal Medicine Quality Testing Technology Based on Reinforced Neural Network

研究利用芳香类中草药的气味数据分析,准确检测中草药品质的特性。然而对于被长期储存或运输的中草药,气味数据特征不明显,传统的模式识别方法直接提取气味特征作为"指纹"数据,输入到识别模型中进行品质检测,受不明显的气味特征影响,存在品质检测准确度不高的问题。为解决上述问题,提出了加强型神经网络的中草药品质检测技术。利用气味数据在变换域主成分特征更明显的特点,对采集到的气味数据进行Gabor变换以加强气味特征,克服特征不明显对检测的影响,将提取到的加强型气味特征作为"指纹"数据输入到神经网络中分析,完成品质的检测。实验表明,改进方法能够有效解决中草药因长期储存或运输造成的气味特征不明显的影响,实现中草药品质特征的准确检测。

Research the aromatic data analysis of Chinese herbal medicine to accurately detect the quality problem of Chinese herbal medicine. This paper put forward the Chinese herbal medicine quality testing technology based on reinforced neural network. Using the features that in the transform domain, the smell principal components are more obvious, the Gabor transform was carried out with thecollected smell data to enhance tht smell characteristics. The reinforced smell features were extracted as a fingerprint data, and then were inputted to the neural network for analysis, completing the quality detection. Experimental results show that this method can effectively solve the problen of non - obvious smell of Chinese herbal medicine caused by long - term storage or transportation.