Tea fermentation degree measuring method with image recognition function

Abstract

The invention discloses a tea leaf fermentation degree determination method with image recognition, which comprises the following steps: shooting tea leaves through a high-definition camera and a camera in different tea leaf processing stages according to different tea leaves to obtain high-definition images of the tea leaves, preprocessing the collected tea leaf images, and determining the fermentation degree of the tea leaves according to the preprocessed tea leaf images. Through denoising, contrast enhancement and color correction technologies, an acquired image is preprocessed, the image is converted into a proper color space for extracting color features in the tea image, and dominant hue and color distribution on the surface of tea are analyzed. The invention relates to the technical field of tea leaf detection, and the method can realize non-destructive, real-time and accurate tea leaf fermentation degree monitoring, and automatically judges the fermentation state of tea leaves by shooting tea leaf images and analyzing visual features of colors, forms and textures of the tea leaves in combination with a machine learning algorithm. Therefore, the automation level and the production efficiency of the tea processing process are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of tea detection, in particular to a method for determining the fermentation degree of tea with image recognition. Background Art

[0002] When processing tea, it needs to be fermented, and the determination of the fermentation degree of tea is crucial for the quality assessment and classification of tea. The fermentation degree of tea directly affects the aroma, taste, color and other characteristics of the tea. If the fermentation degree is too low, the tea may lack a rich flavor; if the fermentation degree is too high, it may lead to a bitter or unpleasant flavor. Therefore, by accurately measuring the fermentation degree, the production process can be adjusted to ensure the stable quality of the tea.

[0003] Traditional methods for determining tea fermentation degree rely primarily on manual observation or chemical testing. These methods are often time-consuming and require human intervention, making real-time, accurate quality monitoring difficult. Furthermore, existing technologies lack an automated detection method that can incorporate visual features such as tea surface color, morphology, and texture. With the rapid development of image recognition technology, applying image recognition to tea fermentation degree determination has become a critical technical requirement for improving tea production efficiency and quality. Summary of the Invention

[0004] In response to the above situation, in order to make up for the above-mentioned existing defects, the present invention provides a method for determining the fermentation degree of tea based on image recognition. This method can realize non-destructive, real-time and accurate monitoring of the fermentation degree of tea. By taking tea images and analyzing their visual characteristics of color, shape and texture, combined with machine learning algorithms, the fermentation status of tea can be automatically judged, thereby improving the automation level and production efficiency of the tea processing process.

[0005] The present invention provides the following technical solution: A method for determining the fermentation degree of tea leaves with image recognition is proposed by the present invention, comprising:

[0006] Step 1: Tea Image Collection

[0007] For different types of tea, at different stages of tea processing, high-definition cameras are used to photograph the tea to obtain high-definition images of the tea. The collected images should have sufficient resolution and clarity to accurately extract color, shape and texture features.

[0008] Step 2: Image preprocessing

[0009] The collected tea images are preprocessed by using denoising, contrast enhancement and color correction techniques to improve image quality and ensure the accuracy of subsequent feature extraction.

[0010] Step 3: Color Analysis

[0011] The image is converted to a suitable color space to extract the color features in the tea image. The color depth of the tea is determined by analyzing the main color tone and color distribution on the tea surface.

[0012] Linear relationship between tea color and fermentation degree: C = k × F

[0013] C represents the color depth of the tea leaves (can be a color difference value or a chromaticity value).

[0014] F represents the degree of fermentation of tea (usually measured by fermentation time, temperature and humidity factors).

[0015] k represents a constant related to tea type and processing method.

[0016] This formula shows that the color depth C is proportional to the fermentation degree F. The color depth of tea is usually proportional to its fermentation degree. The darker the color, the higher the fermentation degree (the specific constant k needs to be determined through experimental data, combined with the specific type of tea and the fermentation environment).

[0017] Step 4: Morphological Analysis

[0018] Through image processing algorithms, the morphological characteristics of tea leaves are extracted, focusing on the edges, curvature and surface texture of the leaves. During the fermentation process, the morphology of the tea leaves will undergo certain changes. Image recognition technology can identify these subtle morphological changes and compare them with standard models of different fermentation stages to further determine the degree of fermentation.

[0019] Step 5: Texture Analysis

[0020] A texture analysis algorithm is used to extract the texture features of the tea surface. The surface texture features of tea vary at different fermentation stages, and the texture complexity and number of details usually increase with the increase of fermentation degree.

[0021] Step 6: Feature Fusion and Machine Learning Model Training

[0022] Features such as color, shape, and texture are integrated to construct a multidimensional feature vector. Through a large amount of labeled tea fermentation degree image data, the model is trained using deep learning and machine learning algorithms, so that it can automatically determine the fermentation degree of tea based on image features. The trained model is used to accurately predict the fermentation status of tea.

[0023] Step 7: Fermentation Degree Prediction and Feedback

[0024] The trained model is used to analyze tea images collected in real time to obtain a prediction of the tea's fermentation degree. Based on the prediction results, the system automatically provides feedback to guide production personnel to make corresponding operational adjustments to ensure that the tea fermentation process meets the standards.

[0025] Step 8: Result display and automated control

[0026] The fermentation degree classification results of tea are displayed through a graphical user interface system (GUI). By connecting the system with the control system of the tea production line, the prediction results can directly affect the automatic adjustment of the production process and realize intelligent production.

[0027] Preferably, the image conversion method includes but is not limited to HSV and Lab color spaces.

[0028] Preferably, the texture analysis algorithm includes but is not limited to gray-level co-occurrence matrix, local binary pattern, wavelet transform and fractal dimension.

[0029] Preferably, deep learning and machine learning algorithms include but are not limited to convolutional neural networks (CNN) and support vector machines (SVM).

[0030] The present invention proposes a method for determining the degree of fermentation of tea leaves with image recognition, which has the following advantages:

[0031] 1. This method can realize the rapid detection of the fermentation degree of tea, avoid the time-consuming problem of traditional manual detection methods, and can monitor the fermentation process of tea in real time.

[0032] 2. This method is a non-destructive detection method combined with image recognition technology. It will not cause any damage to the tea leaves and preserves the original quality of the tea leaves.

[0033] 3. This method integrates the color, shape, and texture of tea leaves through multi-dimensional image features and combines them with machine learning models to improve the accuracy of fermentation degree detection.

[0034] 4. This method can be combined with the automated control system of the tea production line to achieve fully automatic and intelligent monitoring of tea fermentation degree, thereby improving production efficiency and reducing labor costs. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation of the present invention. In the accompanying drawings:

[0036] Figure 1 This is a schematic diagram of the overall detection process of the tea fermentation degree determination method with image recognition in this scheme;

[0037] Figure 2 This is a schematic diagram of the system architecture of the tea fermentation degree determination method with image recognition in this scheme;

[0038] Figure 3This is a schematic diagram of the data flow and control feedback of the tea fermentation degree determination method with image recognition in this scheme. DETAILED DESCRIPTION

[0039] The technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments; based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0040] It should be noted that the words "front", "rear", "left", "right", "up" and "down" used in the following description refer to directions in the accompanying drawings, and the words "inside" and "outside" refer to directions toward or away from the geometric center of a specific component, respectively.

[0041] The present invention proposes a method for determining the degree of fermentation of tea leaves with image recognition, which specifically comprises the following steps:

[0042] Step 1: Tea Image Collection

[0043] 1. Equipment preparation: prepare high-definition cameras and cameras (video equipment with a resolution of no less than 1080p), ensure that the equipment has high resolution and clarity, and ensure that image details can be accurately captured.

[0044] 2. Choose the shooting location. Place the tea leaves in an environment with a uniform background and sufficient light to avoid light reflection and shadow interference, and ensure that all parts of the tea leaves, such as the edges and surfaces of the leaves, are clearly visible.

[0045] 3. Shooting angle: Choose a suitable shooting angle based on the type and shape of the tea leaves to ensure that the tea leaves in the image are fully displayed, which is convenient for subsequent image processing.

[0046] Step 2: Image preprocessing

[0047] 1. Denoising: Use filtering algorithms such as median filtering and mean filtering to denoise the image and remove random noise generated during the acquisition process.

[0048] 2. Contrast enhancement: The image contrast is enhanced through histogram equalization technology, making the details of the tea leaves clearer.

[0049] 3. Color correction: adjust the color deviation of the image to ensure accurate color restoration and use white balance algorithm to avoid color distortion.

[0050] Step 3: Color Analysis

[0051] 1. Color space conversion: convert the collected tea images from RGB color space to HSV (hue, saturation, value) or Lab (brightness, green-red, blue-yellow) color space. Select the appropriate color space for the collected tea images to better extract the color information of the tea leaves.

[0052] 2. Color depth calculation: Calculate the color depth C of the tea according to the changes in the surface hue of the tea leaves. The color depth is calculated using the formula C = k × F, where C is the color depth, F is the fermentation degree, and k is a constant for the type of tea and the processing method. Adjust the value of the constant k based on experimental data to ensure a proportional relationship between color depth and fermentation degree.

[0053] Step 4: Morphological Analysis

[0054] 1. Morphological feature extraction: Use image processing algorithms such as edge detection and contour extraction to extract the morphological features of tea leaves, focusing on the edges, curvature, and damage of the leaves.

[0055] 2. Detection of changes in fermentation degree: As fermentation progresses, the shape of the tea leaves may undergo slight changes. Image recognition technology can capture these changes and automatically determine the fermentation degree of the tea leaves by comparing them with standard models at different fermentation stages.

[0056] Step 5: Texture Analysis

[0057] 1. Texture feature extraction: Common texture analysis algorithms, such as gray-level co-occurrence matrix (GLCM), local binary pattern (LBP), and wavelet transform, are used to extract the texture features of the tea surface.

[0058] 2. Texture complexity analysis. The texture complexity and number of details of tea leaves vary at different fermentation stages. By analyzing the complexity and changing trends of texture features, the fermentation process of tea leaves can be judged.

[0059] Step 6: Feature Fusion and Machine Learning Model Training

[0060] 1. Feature fusion: The features extracted from color, shape, texture and other steps are fused to construct a multi-dimensional feature vector to more comprehensively describe the fermentation degree of tea.

[0061] 2. Model training: Using labeled tea fermentation degree image data, deep learning, such as convolutional neural network (CNN), and traditional machine learning, such as support vector machine (SVM) algorithm, are used to train the model, enabling the model to automatically predict the fermentation degree based on the input tea image features.

[0062] 3. Verification and tuning: Through cross-validation and adjustment of model parameters, the prediction accuracy is continuously improved to ensure the reliability and stability of the model.

[0063] Step 7: Fermentation Degree Prediction and Feedback

[0064] 1. Real-time prediction: During the tea production process, the system inputs the real-time collected tea images into the trained model to predict the fermentation degree of the tea.

[0065] 2. Automatic feedback: Based on the fermentation degree predicted by the model, the system will automatically provide adjustment suggestions to guide production personnel to take necessary actions, such as adjusting temperature and humidity, fermentation time, etc., to ensure that the fermentation process of tea meets the standards.

[0066] Step 8: Result display and automated control

[0067] 1. Graphical user interface (GUI) displays the predicted results of tea fermentation degree, which is convenient for production personnel to view and operate.

[0068] 2. Automated control: connect the prediction results with the control system of the tea production line, automatically adjust production parameters according to the predicted value of fermentation degree, and realize intelligent production.

[0069] 3. Real-time monitoring and adjustment. The system provides real-time monitoring function to ensure that the production process is always in the best fermentation state.

[0070] It should be noted that, in this document, relational terms such as first and second, etc., are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the terms "comprises," "comprising," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, material, or apparatus that includes a list of elements includes not only those elements but also other elements not explicitly listed, or elements inherent to such process, method, material, or apparatus.

[0071] While embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions, and variations may be made to these embodiments without departing from the principles and spirit of the invention, and that the scope of the invention is defined by the appended claims and their equivalents.

Claims

1. A method for determining the degree of fermentation of tea leaves with image recognition, characterized in that: Specifically include the following steps: Step 1: Tea image collection, for different teas, at different stages of tea processing, tea is photographed with high-definition cameras to obtain high-definition images of the tea. The collected images should have sufficient resolution and clarity to accurately extract color, shape and texture features. Step 2: Image preprocessing is performed on the collected tea images. De-noising, contrast enhancement and color correction techniques are used to preprocess the collected images to improve image quality and ensure the accuracy of subsequent feature extraction. Step 3: Color analysis converts the image into a suitable color space to extract the color features in the tea image and judge the color depth of the tea by analyzing the main color tone and color distribution on the surface of the tea. Linear relationship between tea color and fermentation degree: C = k × F C represents the color depth of the tea leaves (can be a color difference value or a chromaticity value). F represents the degree of fermentation of tea (usually measured by fermentation time, temperature and humidity factors). k represents a constant related to tea type and processing method. Step 4: Morphological analysis: Through image processing algorithms, the morphological characteristics of the tea leaves are extracted, focusing on the edges, curvature and surface texture of the leaves. During the fermentation process, the morphology of the tea leaves will undergo certain changes. Image recognition technology can identify these subtle morphological changes and compare them with standard models of different fermentation stages to further determine the degree of fermentation. Step 5: Texture analysis. Use a texture analysis algorithm to extract the texture features of the tea surface. The surface texture features of tea vary at different fermentation stages, and the texture complexity and number of details generally increase with the degree of fermentation. Step 6: Feature fusion and machine learning model training: features such as color, shape, and texture are integrated to construct a multidimensional feature vector. A large amount of labeled tea fermentation degree image data is used to train the model using deep learning and machine learning algorithms, enabling it to automatically determine the fermentation degree of tea based on image features. The trained model is used to accurately predict the fermentation status of tea. Step 7: Fermentation degree prediction and feedback. The trained model analyzes the real-time collected tea images to obtain the tea fermentation degree prediction results. Based on the prediction results, the system automatically provides feedback to guide production personnel to make corresponding operational adjustments to ensure that the tea fermentation process meets the standards. Step 8: Result display and automatic control. The fermentation degree classification results of tea are displayed through the graphical user interface system (GUI). The system is connected to the control system of the tea production line. The prediction results can directly affect the automatic adjustment of the production process and realize intelligent production.

2. The method for determining the degree of fermentation of tea leaves with image recognition according to claim 1, wherein: Image conversion methods include but are not limited to HSV and Lab color spaces.

3. The method for determining the degree of fermentation of tea leaves with image recognition according to claim 1, wherein: Texture analysis algorithms include but are not limited to gray-level co-occurrence matrix, local binary pattern, wavelet transform and fractal dimension.

4. The method for determining the degree of fermentation of tea leaves with image recognition according to claim 1, wherein: Deep learning and machine learning algorithms include but are not limited to convolutional neural networks (CNN) and support vector machines (SVM).

Images