Like the electronic tongue and electronic nose, the Baosheng electronic eye in Shanghai is also a sensory analysis instrument. It mimics the human eye's perception of samples and can conduct overall or partial visual analysis of the color and shape of complex surfaces such as food and its packaging. It can measure color changes during processes like food processing, film coating and polishing, fruit and vegetable preservation, pigment removal, and shelf - life storage, achieve visual quality control, and explore the internal relationship between food quality and color.
The process of identifying Chinese herbal medicines or authentic herbs using the electronic eye (machine vision system) mainly relies on image processing and artificial intelligence technologies. These technologies determine the type, quality, and presence of impurities in the herbs by analyzing their appearance, shape, color, texture, and other characteristics. The specific process usually includes the following steps:
Image Acquisition
The electronic eye system first photographs the Chinese herbal medicines through a high - precision camera. The image acquisition system typically has high resolution and high frame rate, enabling it to capture the details of the herbs in real - time. These images are then transmitted to a computer for subsequent processing.
Image Pre - processing
Image data may be affected by factors such as lighting and angle, so pre - processing steps are required. These steps include:
Denoising: Removing noise from the image to ensure the accuracy of subsequent analysis.
Grayscaling: Converting the image into a grayscale image to simplify calculations.
Contrast Adjustment: Enhancing key features in the image to improve 辨识度.
Image Smoothing: Reducing distortion caused by motion blur or shooting angles.
Feature Extraction
In the pre - processed image, the system extracts multiple features of the herbs:
Morphological Features: By analyzing the shape, size, and structure of the herbs, it can identify whether the herbs are the specified authentic herbs.
Texture Analysis: Analyzing the texture, patterns, or other micro - structural features on the surface of the herbs helps determine if they are counterfeit or non - standard herbs.
Color Recognition: Different Chinese herbal medicines or authentic herbs may have specific colors, which can be used as a basis for identification.
Size and Proportion: The size and shape proportion of the herbs can also be used as a basis for distinguishing different varieties.
Defect Detection
The electronic eye system can detect whether there are cracks, insect damage, mildew, impurities, and other defects on the surface of the herbs. This part is particularly important because the quality of Chinese herbal medicines is not only determined by their appearance, and the detection of defects directly affects the grade and usability of the herbs.
Flaw Identification: Such as cracks, discoloration, spots, etc.
Impurity Removal: The system can identify and remove non - standard impurities such as sand, leaves, etc.
Spoilage Detection: Including problems such as mildew, rot, or moisture.
Classification and Identification
After feature extraction and defect detection, the machine vision system compares the Chinese herbal medicines with the known standard herbs in the database and uses pattern recognition algorithms (such as the Convolutional Neural Network - CNN) for classification.
Herb Identification: By comparing with the authentic herbs in the database, the system determines the type of the herb. For example, through the analysis of shape, color, and texture, the system can identify whether it is angelica, astragalus, wolfberry, etc.
Authenticity Judgment: Judging whether the herb meets the standards of authentic herbs and whether it has the geographical signature features of authentic herbs. For example, by comparing with the images of herbs from the origin, the system can determine whether the herb comes from the designated production area.
Quality Evaluation
By analyzing the features extracted from the images, the electronic eye system can evaluate the quality of the herbs to determine whether they meet the quality standards. Quality evaluation usually includes:
Color Uniformity: Authentic herbs often have specific colors, and the electronic eye can detect color differences.
Surface Smoothness: Whether there are cracks, damages, or irregular surfaces.
Integrity: Checking whether the herbs are complete, without any defects or breakages.
Data Storage and Report Generation
The system stores the test results and generates a quality report. This report can include information such as the type of the herb, quality score, whether it meets the standards of authentic herbs, and whether it contains impurities. In this way, producers and suppliers can conduct quality control and traceability of the herbs.
Application Cases:
Identification of Chinese Herbal Medicine Types: The machine vision system can quickly identify and distinguish different Chinese herbal medicines. For example, it can automatically identify wolfberry, astragalus, codonopsis, etc., according to the shape and surface texture of the herbs.
Authentication of Authentic Herbs: Some Chinese herbal medicines such as gastrodia elata, wolfberry, and codonopsis have specific geographical indications. The machine vision system can assist in judging whether the herbs are genuine authentic herbs by comparing their appearance features.
Detection of Impurities and Defects: The electronic eye can effectively identify impurities (such as sand, sawdust, other plant residues) or surface damages in the herbs, thus helping to ensure the purity and quality of the herbs.
Automatic Sorting: On the production line, the vision system can automatically sort the herbs that meet the standards and those that do not according to the quality evaluation results of the herbs.
Advantages:
High Efficiency: Machine vision can efficiently and real - time complete the detection task of Chinese herbal medicines, greatly improving production efficiency.
High Precision: Through sophisticated image processing algorithms, the machine vision system can achieve subtle quality detection to ensure the quality of the herbs.
Reduction of Human Error: Automated detection avoids subjective factors in manual detection, ensuring the consistency of test results.
Intelligence: Through artificial intelligence algorithms, the system can self - learn and optimize, gradually improving the accuracy and efficiency of identification.
In general, electronic eye technology can effectively identify and conduct quality inspections of Chinese herbal medicines and authentic herbs through image processing and artificial intelligence algorithms, ensure that the herbs meet the standards, reduce errors in human operations, and enhance the automation of the entire production line.
