基于OpenCV的中药饮片“辨色”识别研究

Research of OpenCV-based Recognition System of Herbal Decoction Pieces

为利用计算机视觉技术实现对中药饮片的种类识别与质量鉴定,文章从颜色特征入手研究了OpenCV在中药饮片识别中的应用。首先介绍了OpenCV计算机视觉库、饮片识别材料及实验设备,然后论述了OpenCV在饮片图像分割、颜色特征提取的具体方法,最后使用OpenCV机器学习库中的支持向量机算法对5种饮片进行"辨色"识别。实验结果表明,在相同的图像采集环境下,系统能够准确识别出饮片的种类。该研究对于利用计算机视觉进行饮片种类识别具有积极意义,为进一步研发基于计算机视觉的中药饮片质量鉴定系统奠定基础。

For the realization of species identification and quality evaluation of herbal decoction pieces by using computer vision technology,the OpenCV-based identification method of herbal decoction pieces was studied starting with the color characteristics.This paper first introduces the OpenCV computer vision library,the identification material of herbal decoction pieces and the experimental equipment,and then discusses the specific method of OpenCV in the herbal decoction pieces image segmentation and color feature extraction.Finally,use the SVM machine learning algorithm in OpenCV machine learning library to identify five kinds of frequently used herbal decoction pieces.The results show that the system can obtain good recognition effect in the same image acquisition environment.The study has positive significance for the identification of herbal decoction pieces,which lays the foundation for the further development of herbal decoction pieces quality identification system based on computer vision.