To realize the non-destructive identification of Panax notoginseng powder in different parts, this research proposes a non-destructive identification method based on the electronic nose and time-domain feature extraction. First, the electronic nose technology combined with statistical analysis method was used to collect and extract nine time-domain characteristics of the response information of Panax notoginseng whole root powder, tap root powder, rhizome powder, and fibrous powder, including the data at 110 s, the mean value between 101–120 s, the maximum value, minimum value, integral value, differential value, skewness factor, kurtosis factor, and standard deviation between 0–120 s. Next, three classical feature selection method was used to reduce the data dimension. Subsequently, the classification models of support vector machine (SVM), least-square support vector machine (LSSVM), and extreme learning machine (ELM) were established based on original data, multi-feature data, and feature selection data. Finally, the Grey Wolf Optimization (GWO) algorithms were introduced to optimize the parameters of the classification model. The results show that the GWO-CARS-LSSVM achieved the best modeling effect, and the classification accuracy on the test set was 97.92%. This study provides a theoretical basis and technical support for rapid identification of adulteration of Panax notoginseng powder.

The existing classification methods for Panax notoginseng taproots suffer from low accuracy, low efficiency, and poor stability. In this study, a classification model based on image feature fusion is established for Panax notoginseng taproots. The images of Panax notoginseng taproots collected in the experiment are preprocessed by Gaussian filtering, binarization, and morphological methods. Then, a total of 40 features are extracted, including size and shape features, HSV and RGB color features, and texture features. Through BP neural network, extreme learning machine (ELM), and support vector machine (SVM) models, the importance of color, texture, and fusion features for the classification of the main roots of Panax notoginseng is verified. Among the three models, the SVM model performs the best, achieving an accuracy of 92.037% on the prediction set. Next, iterative retaining information variables (IRIVs), variable iterative space shrinkage approach (VISSA), and stepwise regression analysis (SRA) are used to reduce the dimension of all the features. Finally, a traditional machine learning SVM model based on feature selection and a deep learning model based on semantic segmentation are established. With the model size of only 125 kb and the training time of 3.4 s, the IRIV-SVM model achieves an accuracy of 95.370% on the test set, so IRIV-SVM is selected as the main root classification model for Panax notoginseng. After being optimized by the gray wolf optimizer, the IRIV-GWO-SVM model achieves the highest classification accuracy of 98.704% on the test set. The study results of this paper provide a basis for developing online classification methods of Panax notoginseng with different grades in actual production.

The classification of the taproots of Panax notoginseng is conducive to improving the economic added value of its products. In this study, a real-time sorting robot system for Panax notoginseng taproots was developed based on the improved DeepLabv3+ model. The system is equipped with the improved DeepLabv3+ classification model for different grades of Panax notoginseng taproots. The model uses Xception as the taproot feature extraction network of Panax notoginseng. In the residual structure of the Xception network, a group normalization layer with deep separable convolution is adopted. Meanwhile, the global maximum pooling method is added in the Atrous Spatial Pyramid Pooling (ASPP) part to retain more texture information, and multiple shallow effective feature layers are designed to overlap in the decoding part to minimize the loss of features and improve the segmentation accuracy of Panax notoginseng taproots of all grades. The model test results show that the Xception-DeepLabv3+ model performs better than VGG16-U-Net and ResNet50-PSPNet models, with a Mean Pixel Accuracy (MPA) and a Mean Intersection over Union (MIoU) of 78.98% and 88.98% on the test set, respectively. The improved I-Xce-DeepLabv3+ model achieves an average detection time of 0.22 s, an MPA of 85.72%, and an MIoU of 90.32%, and it outperforms Xce-U-Net, Xce-PSPNet, and Xce-DeepLabv3+ models. The system control software was developed as a multi-threaded system to design a system grading strategy, which solves the problem that the identification signal is not synchronized with the grading signal. The system test results show that the average sorting accuracy of the system is 77% and the average false detection rate is 21.97% when the conveyor belt running speed is 1.55 m/s. The separation efficiency for a single-channel system is 200–300 kg/h, which can replace the manual work of three workers. The proposed method meets the requirements of current Panax notoginseng processing enterprises and provides technical support for the intelligent separation of Panax notoginseng taproots.

Identifying the grade of Gastrodia elata in the market has low efficiency and accuracy. To address this issue, an I-YOLOX object detection algorithm based on deep learning and computer vision is proposed in this paper. First, six types of Gastrodia elata images of different grades in the Gastrodia elata planting cooperative were collected for image enhancement and labeling as the model training dataset. Second, to improve feature information extraction, an ECA attention mechanism module was inserted between the backbone network CSPDarknet and the neck enhancement feature extraction network FPN in the YOLOX model. Then, the impact of the attention mechanism and application position on model improvement was investigated. Third, the 3 × 3 convolution in the neck enhancement feature extraction network FPN and the head network was replaced by depthwise separable convolution (DS Conv) to reduce the model size and computation amount. Finally, the EIoU loss function was used to predict boundary frame regression at the output prediction end to improve the convergence speed of the model. The experimental results indicated that compared with the original YOLOX model, the mean average precision of the improved I-YOLOX network model was increased by 4.86% (97.83%), the model computation was reduced by 5.422 M (reaching 3.518 M), the model size was reduced by 20.6 MB (reaching 13.7 MB), and the image frames detected per second increased by 3 (reaching 69). Compared with other target detection algorithms, the improved model outperformed Faster R-CNN, SSD-VGG, YOLOv3s, YOLOv4s, YOLOv5s, and YOLOv7 algorithms in terms of mean average precision, model size, computation amount, and frames per second. The lightweight model improved the detection accuracy and speed of different grades of Gastrodia elata and provided a theoretical basis for the development of online identification systems of different grades of Gastrodia elata in practical production.