A tea beverage dispensing method and device based on intelligent visual guidance

Abstract

The application relates to a tea beverage blending method and device based on intelligent visual guidance, which comprises the following steps: adding oolong tea leaves into a brewing vessel, the periphery of the brewing vessel being provided with multiple cameras, and collecting image information in an oolong tea soaking process through the cameras; injecting water with a preset temperature into the brewing vessel from top to bottom; identifying foam pattern information of the oolong tea in the soaking process based on the image information, wherein the foam pattern information comprises foam coverage, foam uniformity, overall characteristics of the foam pattern, foam particle size distribution characteristics and foam layer thickness uniformity; judging the brewing degree of the oolong tea based on the foam pattern information; and dynamically adjusting the brewing process according to the brewing degree. By introducing the foam pattern information, the tea soup state can be finely identified, and it can be judged whether the tea leaf quality is good or whether there is an abnormal condition in the brewing process. The quality evaluation can be completed only by relying on image processing.

Description

Technical Field

[0001] This application relates to the field of food processing and beverage preparation technology, and in particular to a tea preparation method and equipment based on intelligent vision guidance. Background Technology

[0002] Oolong tea is one of the six traditional tea categories in my country, and it is highly popular due to its unique processing and rich aroma. Oolong tea is a semi-fermented tea, combining the fragrance of green tea with the mellowness of black tea. Its production process is complex, mainly including sun-drying, shaking, fixation, rolling, and roasting. The finished tea has a tightly rolled or bead-like appearance, a dark green to yellowish color, a high and lasting aroma, a bright golden-yellow liquor, a mellow and sweet taste, a long-lasting aftertaste, and red edges and green centers on the infused leaves, making it highly visually appealing and of excellent drinking quality.

[0003] Oolong tea is rich in tea polyphenols, flavonoids, aromatic esters, and a certain amount of saponins and proteins. These components form abundant and stable surface foam during hot water steeping. Especially in the initial pouring stage, the foam forms rapidly, is full-bodied, and has good visual characteristics, reflecting the freshness of the raw materials, the brewing intensity, and the release of components. Furthermore, the release of aroma compounds in oolong tea is relatively concentrated, with a clear extraction window. Even slight under- or over-brewing can lead to insufficient aroma or increased bitterness, affecting the balance of flavor. Therefore, the stable blending of high-quality oolong tea beverages still highly depends on the experience and control of the tea master. Slight errors can result in the loss of aroma, an unbalanced flavor, or poor brewing performance.

[0004] In modern commercial settings, such as chain tea shops, intelligent self-service tea brewing equipment, and high-end customized tea production lines, the demands of multiple batches of oolong tea raw materials, changing customer tastes, and rapid delivery mean that relying solely on human experience is insufficient to consistently guarantee product quality and efficiency. Traditional procedural control methods based on quantitative, temperature-controlled, and time-controlled processes also lack the ability to perceive the tea infusion's state in real time, easily leading to under- or over-extraction.

[0005] In recent years, image intelligent analysis technology has been used for tea beverage preparation, but it has mostly focused on identifying the color of the tea liquor. However, oolong tea liquor has problems such as high transparency, weak texture, and sensitivity to light, making it difficult to capture images with high fidelity and causing the analysis results to be easily interfered with. Therefore, it is necessary to provide a tea beverage preparation system that can accurately perceive the state of the tea liquor and achieve efficient and stable blending without relying on liquor color recognition. Summary of the Invention

[0006] The purpose of this application is to solve the problems mentioned above, such as the high transparency of oolong tea soup, weak texture, sensitivity to light, difficulty in image acquisition and fidelity, and easy interference with analysis results, which make it impossible to accurately perceive the state of the tea soup and achieve efficient and stable blending.

[0007] According to one aspect of this application, a tea preparation method based on intelligent vision guidance is provided, comprising:

[0008] S100. Add oolong tea leaves to a brewing vessel. Multiple cameras are installed around the brewing vessel to capture image information of the oolong tea steeping process. The sidewalls of the brewing vessel are made of transparent material.

[0009] S200. Water at a preset temperature is poured into the brewing vessel from top to bottom;

[0010] S300. Based on the image information, identify the foam morphology information of oolong tea during the steeping process. The foam morphology information includes: foam coverage, foam uniformity, overall foam morphology characteristics, foam particle size distribution characteristics, and foam layer thickness uniformity.

[0011] S400. Determine the brewing degree of oolong tea based on the foam morphology information;

[0012] S500: Dynamically adjust the brewing process according to the brewing degree.

[0013] Preferably, S400 includes:

[0014] The degree of release of surfactants in tea leaves is determined based on the foam coverage rate.

[0015] Based on the foam uniformity, the extraction balance and surface flow of the tea soup are determined.

[0016] Based on the overall characteristics of the foam morphology, the consistency of tea quality, the physical stability of the tea soup, and the suitability of the current brewing stage can be determined.

[0017] Based on the foam particle size distribution characteristics, determine whether the release of active molecules in the tea soup is sufficient and stable;

[0018] The viscosity of the tea soup and the intensity of local extraction are determined based on the uniformity of the foam layer thickness.

[0019] Preferably, the image information includes a top view and a side view, and the step of identifying the foam morphology information of oolong tea during the steeping process based on the image information includes:

[0020] S310. Based on the top view, obtain the foam coverage rate during the oolong tea steeping process, specifically including:

[0021] S311. Extract the foam region using image segmentation technology, and calculate the relationship between the foam region and the surface of the brewing vessel.

[0022] The ratio of the surface area covered is denoted as the foam coverage rate;

[0023] S320. Based on the top view, obtain the foam uniformity during the oolong tea steeping process, specifically including:

[0024] S321. Divide the upper surface of the brewing vessel into M×N regions, and record the number of bubbles n1, n2, n3, ..., n in each region. m×n Calculate the ratio of the foam area to the area of ​​each region, and record it as...

[0025] The foam density in this area;

[0026] S322. Calculate the variance of the foam density in all areas, and denot it as the foam uniformity.

[0027] S330. Based on the top view, obtain the overall characteristics of the foam morphology during the oolong tea steeping process, specifically including:

[0028] S331. Calculate the area S and perimeter A of the foam region, based on... Calculate the foam region

[0029] The ratio of the area to the perimeter of the region is denoted as the edge regularity R1 of the foam region;

[0030] S332. Sort the foam edge contours according to pixel connection order, calculate the average curvature continuity of the curve, and record...

[0031] For edge continuity;

[0032] S333. Calculate the minimum bounding rectangle of the foam region and obtain the major axis a of the minimum bounding rectangle.

[0033] And the minor axis b, according to Calculate the edge deviation R2 of the foam region;

[0034] S340. Based on the side view, obtain the foam particle size distribution characteristics during the oolong tea steeping process, specifically including:

[0035] S341. Identify the diameter of each bubble using edge detection and contour analysis algorithms, and count the diameter of each bubble.

[0036] S342. Set a preset foam diameter window, and count the percentage of foams within the target diameter window.

[0037] Characteristics of foam particle size distribution;

[0038] S350. Based on the side view, obtain the uniformity of the foam layer thickness during the oolong tea steeping process, specifically including:

[0039] S351. Divide the side view image of the foam region into several regions along the horizontal direction, calculate the foam layer thickness in each region, and calculate the mean μ and standard deviation σ of the foam layer thickness values. calculate

[0040] The uniformity of foam thickness;

[0041] S360. Based on the side view, determine whether there are particulate matter during the oolong tea steeping process, specifically including:

[0042] S361. Apply grayscale threshold segmentation or a deep learning-based target detection model to the foam region to distinguish between air bubbles.

[0043] Bubbles and particulate matter;

[0044] S362. Calculate the area, quantity, and diameter of all particles.

[0045] Preferably, determining the brewing level of oolong tea based on the foam morphology information includes:

[0046] S410. Determine whether the following indicators meet the corresponding preset threshold range:

[0047] Does the foam coverage rate meet the preset foam coverage rate threshold range?

[0048] Does the foam uniformity meet the preset foam uniformity threshold range?

[0049] Whether the foam particle size distribution characteristics meet the preset foam particle size threshold range;

[0050] Whether the thickness uniformity of the foam layer meets the preset thickness uniformity threshold range;

[0051] S420. If any one of the conditions is not met, then based on the direction and magnitude of the deviation, identify the current brewing state as one of the following states:

[0052] Initial extraction stage: The foam coverage is lower than the preset foam coverage threshold range, the foam particle size distribution characteristics do not meet the preset foam particle size threshold range, and the foam layer thickness uniformity does not meet the preset thickness uniformity threshold range.

[0053] Insufficient component release stage: The foam coverage rate meets the preset foam coverage rate threshold range, but the foam particle size distribution characteristics do not meet the preset foam particle size threshold range, and the foam layer thickness uniformity does not meet the preset thickness uniformity threshold range.

[0054] Over-extraction stage: The foam coverage rate is higher than the preset foam coverage rate threshold range, but the foam particle size distribution characteristics do not meet the preset foam particle size threshold range, and the foam layer thickness uniformity does not meet the preset thickness uniformity threshold range.

[0055] Accordingly, S500 includes:

[0056] S510. If all the indicators meet the corresponding threshold range, it is determined that the current brewing level has reached the target state, the steeping ends, and the tea soup and tea leaves are separated.

[0057] S520. If the process is identified as "initial extraction stage" or "insufficient component release stage", then extend the soaking time.

[0058] S530. If the "over-extraction stage" is identified, the brewing process will be terminated and the tea soup will be quickly discharged.

[0059] Preferably, the foam morphology information further includes foam color, which is obtained based on the grayscale value of the foam region;

[0060] Accordingly, S410 further includes:

[0061] S411. Determine whether the foam color meets the preset foam color range;

[0062] The S420 also includes:

[0063] S421. If any one of the conditions is not met, then based on the direction and magnitude of the deviation, identify the current brewing state as one of the following states:

[0064] Initial extraction stage: The foam coverage is lower than the preset foam coverage threshold range, the foam particle size distribution characteristics do not meet the preset foam particle size threshold range, the foam layer thickness uniformity does not meet the preset thickness uniformity threshold range, and the foam color is lower than the preset foam color range.

[0065] Insufficient component release stage: The foam coverage rate meets the preset foam coverage rate threshold range, but the foam particle size distribution characteristics do not meet the preset foam particle size threshold range, the foam layer thickness uniformity does not meet the preset thickness uniformity threshold range, and the foam color is lower than the preset foam color range.

[0066] Over-extraction stage: The foam coverage rate is higher than the preset foam coverage rate threshold range, but the foam particle size distribution characteristics do not meet the preset foam particle size threshold range, the foam layer thickness uniformity does not meet the preset thickness uniformity threshold range, and the foam color is higher than the preset foam color range.

[0067] Preferably, step S421, which identifies the current brewing state as one of the following states, further includes:

[0068] The tea soup is too strong or the aroma is too intense: the foam thickness uniformity does not meet the preset thickness uniformity threshold range, the foam color is higher than the preset foam color range, and the foam uniformity does not meet the preset foam uniformity threshold range.

[0069] Accordingly, S500 includes:

[0070] S540. If the tea soup is identified as "too strong or too fragrant", then diluting additives are added, wherein the diluting additives include syrup and fruit juice.

[0071] Preferably, step S421, which identifies the current brewing state as one of the following states, further includes:

[0072] Poor tea quality or abnormal brewing: The edge regularity is lower than the preset edge regularity threshold, the edge continuity is lower than the preset edge continuity threshold, and the edge deviation is higher than the preset edge deviation threshold.

[0073] Accordingly, S500 includes:

[0074] S550 If the problem is identified as "poor tea quality or abnormal brewing", an error message will be issued, the current brewing process will be paused, and manual intervention or switching of raw materials will be required for re-brewing.

[0075] Preferably, step S421, which identifies the current brewing state as one of the following states, further includes:

[0076] The tea leaves are not fully dispersed and the concentration is uneven: the thickness uniformity of the foam does not meet the preset thickness uniformity threshold range;

[0077] Accordingly, S500 includes:

[0078] S560. If it is identified as "tea leaves are not sufficiently dispersed or have uneven concentration", then control the stirring component to extend from top to bottom below the surface of the tea liquid in the brewing vessel.

[0079] S561, Start the stirring operation for the preset duration;

[0080] S562. After stirring is completed, control the stirring component to remove the tea soup.

[0081] Preferably, step S421, which identifies the current brewing state as one of the following states, further includes:

[0082] The tea soup contains impurities: the particulate matter coverage exceeds a preset particulate matter coverage threshold; or, the maximum particulate matter diameter is greater than a preset particle size threshold.

[0083] Accordingly, S500 includes:

[0084] S570. If it is identified that "the tea soup contains impurities", then the tea soup in the brewing vessel is filtered.

[0085] Preferably, the foam morphology information further includes the foam growth rate, which is obtained by calculating the increment of the foam coverage within a preset time interval.

[0086] Accordingly, S421 further includes identifying the current brewing state as one of the following states:

[0087] Water temperature too low: The foam growth rate is lower than the preset foam growth rate threshold range;

[0088] Water temperature too high: The foam growth rate exceeds the preset foam growth rate threshold range;

[0089] S430. Determine whether the increment of the foam coverage rate exceeds the preset coverage rate increment window;

[0090] The S500 includes:

[0091] S580. If "water temperature is too high" is detected, a first compensating water flow is injected, wherein the temperature of the first compensating water flow is lower than the preset temperature in S200.

[0092] S581. If the water temperature is identified as "too high", then a second compensating water flow is injected, wherein the temperature of the second compensating water flow is higher than the preset temperature in S200.

[0093] This application offers the following advantages: By introducing foam morphology information, it achieves precise identification of the tea infusion state, accurately determining whether the tea quality is good or whether there are any abnormalities during brewing. It enhances the intelligence and objectivity of tea quality judgment, avoiding subjective bias caused by manual judgment; it improves the sensitivity and accuracy of anomaly detection, enabling timely identification of brewing abnormalities caused by tea deterioration, extraction abnormalities, or equipment contamination; it eliminates the need for chemical detection or complex sensors, relying solely on image processing to complete quality assessment, simplifying the system structure; it supports dynamic adjustments during the brewing process, such as triggering tea filtration, adjusting water temperature, water volume, or extraction time, improving brewing stability and beverage consistency; and it adapts to the identification needs of various tea types and grades, possessing good versatility and scalability. Attached Figure Description

[0094] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0095] Figure 1 This is a logic block diagram of a tea brewing method based on intelligent vision guidance according to an embodiment of this application. Detailed Implementation

[0096] To facilitate understanding of this application, a more complete description will be provided below with reference to the accompanying drawings. Preferred embodiments of this application are shown in the drawings. However, this application can be implemented in many different forms and is not limited to the embodiments described herein. Rather, these embodiments are provided to provide a more thorough and complete understanding of the disclosure of this application.

[0097] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs. The terminology used herein in the specification of this application is for the purpose of describing particular embodiments only and is not intended to be limiting of the application. The term "and / or" as used herein includes any and all combinations of one or more of the associated listed items.

[0098] While the foam state appears to be a physical phenomenon, it can indirectly reflect various microscopic changes in oolong tea during the brewing process, encompassing dimensions such as the degree of release of active substances, the progress of oxidation reactions, and the physical properties and clarity of the tea liquor. Especially for oolong tea, which has a high degree of fermentation, rich components, and complex aroma, foam characteristics have the following indicative value:

[0099] 1. Degree of release of active ingredients

[0100] A large number of foams with a stable structure indicates that the surface-active substances such as tea polyphenols, flavonoids, tea saponins, and proteins in the tea leaves are fully released, resulting in a higher concentration of tea soup and often a more pronounced aroma and fuller taste.

[0101] Sparse or rapidly bursting foam indicates that the active ingredients have not been fully released or have nearly dissolved completely. The soaking time may be insufficient or excessive, which may affect the harmony of the flavor.

[0102] 2. Oxidation state and soup color evolution

[0103] The color of the foam gradually changes from light yellow to golden yellow, orange-yellow, or even reddish-brown (with slight variations depending on the type of oolong tea and the degree of roasting): This reflects the enzymatic or non-enzymatic browning process of polyphenolic oxides and flavonoids in the tea soup, and is one of the microscopic signs of the deepening of the soup color.

[0104] Brown deposits or darkening of color at the edge of the foam: This indicates that high temperature may have caused rapid extraction and browning. This phenomenon is common in roasted oolong tea when brewed at high temperatures. It suggests that the water temperature and brewing time should be properly controlled to avoid a harsh or bitter taste.

[0105] 3. Changes in fluidity and viscosity

[0106] The foam distribution flows naturally with the liquid surface, and is uniform and stable: this indicates that the tea soup has good fluidity, moderate viscosity, and a balanced extraction process;

[0107] Uneven foam accumulation or stagnation: This may be caused by increased viscosity of the tea soup or excessive dissolution of some components (such as pectin released from the tips of high-roasted rock tea leaves), or it may indicate broken tea leaves or excessive sedimentation at the bottom of the leaves.

[0108] 4. Suspended particles and clarified state

[0109] If fine particles or sediment are visible in the foam, it indicates that there are a lot of suspended matter or fragments in the tea soup, which may affect the clarity of the soup color and the smoothness of the taste. This is especially unacceptable in low-roasted light-aroma oolong teas and may indicate the need for subsequent filtration or adjustment of the amount of tea.

[0110] Clear and transparent foam with few impurities indicates high tea quality, complete processing, and suitable brewing conditions.

[0111] Please refer to Figure 1 One embodiment of this application provides a tea brewing method based on intelligent vision guidance, including:

[0112] S100. Add oolong tea leaves to a brewing vessel. Multiple cameras are installed around the brewing vessel to capture image information during the oolong tea steeping process. The sidewalls of the brewing vessel are made of a transparent material. It should be noted that by installing multiple cameras around the brewing vessel and using a transparent sidewall structure, image information can be acquired to comprehensively capture the formation, diffusion, distribution, and changes in foam during the oolong tea steeping process.

[0113] S200. Pour water at a preset temperature into the brewing vessel from top to bottom. In this step, it's important to note that pouring water at a preset temperature from top to bottom offers the following significant advantages compared to traditional side or bottom pouring methods: It simulates the water flow pattern of hand-brewing; the top-down pouring method is closer to the "high pour" technique in traditional tea ceremony, helping to stimulate the release of active ingredients and aroma from the tea leaves, promoting their natural unfolding and even dispersion. The water flow disturbance created by top-down pouring effectively enhances the floating and rotating motion of the tea leaves in the water, improving steeping efficiency. It also enhances the controllability of foam formation conditions; the impact of the water flow on the tea leaf surface easily induces foam formation, and combined with the preset temperature parameters, it helps to form an appropriate amount of observable foam, providing ideal conditions for subsequent visual recognition.

[0114] In summary, this step not only optimizes the brewing environment but also lays a solid foundation for subsequent visual recognition and intelligent control, making it one of the key steps in achieving standardized and intelligent tea blending.

[0115] S300. Based on image information, identify the foam morphology information of oolong tea during the steeping process. The foam morphology information includes: foam coverage, foam uniformity, overall foam morphology characteristics, foam particle size distribution characteristics, and foam layer thickness uniformity. In this step, it should be noted that through intelligent identification and feature extraction of foam areas in the image information, the system can quantitatively obtain multi-dimensional foam morphology parameters, providing a data foundation for judging and controlling the brewing level. Specifically, this includes:

[0116] Based on the foam coverage, the degree of release of surface-active components in tea can be judged. This parameter can reflect the leaching intensity of tea and is a key basis for evaluating the fullness of brewing.

[0117] Based on the uniformity of foam, we can judge the extraction balance and surface flow of the tea soup, thereby understanding whether the distribution of substances released from the tea leaves is uniform.

[0118] Based on the overall characteristics of the foam morphology, the consistency of tea quality, the physical stability of the tea soup, and the suitability of the current brewing stage can be judged. For example, the stability and order of the foam structure can be identified by the regularity, continuity, and deviation of the edges.

[0119] Based on the characteristics of foam particle size distribution, it can be determined whether the release of active molecules in the tea soup is sufficient and stable. If the particle size distribution is concentrated, it indicates that the extraction process is relatively ideal.

[0120] The viscosity of the tea soup and the intensity of local extraction can be judged by the uniformity of the foam layer thickness. This indicator helps to assess local concentration changes and the thickness of the tea soup.

[0121] This step, through the fusion analysis of the above-mentioned multidimensional foam characteristic parameters, enables the visual monitoring of key physical and chemical changes during the brewing process, providing a scientific basis for subsequent intelligent judgment and adjustment strategies, and improving the intelligence and precision of the tea blending process.

[0122] S400. Determine the brewing level of oolong tea based on foam morphology information. In this step, it should be noted that by analyzing the extracted foam morphology information, the system can perform real-time evaluation of the oolong tea brewing process, thereby determining whether the current brewing stage has reached the set standard or requires adjustment. Specifically, this includes:

[0123] If the foam coverage continues to rise and tends to stabilize, it indicates that the release of surfactants in the tea leaves is approaching saturation, and the brewing process is nearing completion.

[0124] If the uniformity of the foam decreases or there are obvious regional differences, it may indicate uneven tea extraction or disturbance from water injection, requiring adjustment of the water flow or stirring method.

[0125] If the overall characteristics of the foam morphology (such as edge regularity, continuity, and deviation) show that the foam structure is messy or broken, it may mean that the tea quality is poor or that there are impurities in the tea soup.

[0126] If the foam particle size distribution is too dispersed, or large foam particles persist, it indicates that the release process is unstable and the brewing is insufficient.

[0127] If the uniformity of the foam layer thickness decreases, and local accumulation or blank areas appear, it may indicate that the tea leaves are not properly stacked or that local extraction is too strong.

[0128] This step, through comprehensive judgment of foam parameters, enables the system to accurately identify brewing progress and abnormal status, providing a basis for subsequent automatic control and personalized blending, thereby achieving scientific and intelligent control of the tea brewing process.

[0129] S500. Dynamically adjust the brewing process according to the brewing level. In this step, it should be noted that the system dynamically adjusts key parameters of the brewing process based on the brewing level determined by the aforementioned foam morphology information, in order to improve the consistency of the tea's taste and the level of intelligence in the brewing process. This includes, but is not limited to:

[0130] If the system detects that the current brewing level is insufficient, it can automatically extend the steeping time or add water a second time to enhance the release of active ingredients.

[0131] If the current foam structure is identified as unstable or uneven, the system can improve the foam state and enhance the extraction balance by adjusting the water temperature, water flow rate, or stirring frequency.

[0132] If the system identifies that the tea is of poor quality, such as low foam edge regularity or high foam deviation, it can issue a warning or switch to an adaptive brewing mode to optimize flavor release.

[0133] If the system identifies "impurities in the tea soup", such as particulate matter coverage exceeding the threshold or the maximum particle diameter exceeding the threshold, the system will activate the filtration module to filter the tea soup.

[0134] If the system recognizes that the brewing process has reached its optimal state, it can automatically terminate the brewing process and prompt the user to dispensing the beverage, thus avoiding over-extraction.

[0135] This step ensures that the system can respond in real time based on the recognition results, making the brewing process more precise and flexible, and improving the adaptability and user experience of the intelligent tea blending system.

[0136] The technical solution implemented in this embodiment, by introducing foam morphology information, achieves precise identification of the tea infusion state, accurately determining whether the tea quality is good or whether there are any abnormalities during the brewing process. This enhances the intelligence and objectivity of tea infusion quality judgment, avoiding subjective biases caused by manual judgment; improves the sensitivity and accuracy of anomaly detection, enabling timely identification of brewing abnormalities caused by tea deterioration, extraction abnormalities, or equipment contamination; eliminates the need for chemical detection or complex sensors, relying solely on image processing to complete quality assessment, simplifying the system structure; supports dynamic adjustments during the brewing process, such as triggering tea filtration, adjusting water temperature, water volume, or extraction time, improving brewing stability and beverage consistency; and adapts to the identification needs of various tea types and grades, possessing good versatility and scalability.

[0137] In one specific embodiment, the image information includes a top view and a side view. Based on the image information, identifying the foam morphology information of oolong tea during the steeping process includes:

[0138] S310. Based on the top view, obtain the foam coverage rate during the oolong tea steeping process, specifically including:

[0139] S311. Extract the foam region using image segmentation technology, and calculate the ratio of the foam region to the coverage area of ​​the upper surface of the brewing vessel, which is denoted as the foam coverage rate.

[0140] In step S310, it should be noted that the foam coverage reflects the release capacity of surface-active ingredients in tea leaves. A higher coverage usually corresponds to a more thorough soaking state. Therefore, by acquiring images from a top-down angle and extracting the foam area, the degree of foam coverage on the liquid surface can be assessed, providing a quantitative basis for judging the brewing progress and the effect of ingredient release.

[0141] In step S311, it should be noted that an image segmentation method (such as threshold segmentation, morphological processing, or depth segmentation network) is used to accurately extract the foam region from the background, and the coverage value is calculated by pixel area. A high coverage value indicates that the active substances in the tea leaves are fully released and the foam formation is rich and stable.

[0142] S320. Based on the top view, obtain the foam uniformity during the oolong tea steeping process, specifically including:

[0143] S321. Divide the upper surface of the brewing vessel into M×N regions and record the amount of foam in each region.

[0144] n1, n2, n3, ..., n m×n Calculate the ratio of the foam area to the area of ​​each region, denoted as .

[0145] The foam density in this area;

[0146] S322. Calculate the variance of foam density for all regions, denoted as foam uniformity.

[0147] In step S320, it should be noted that foam uniformity can reveal the spatial distribution consistency of foam, reflecting the balance between liquid surface tension and brewing extraction. Uniform distribution usually means stable brewing water flow and good blade unfolding.

[0148] In step S321, it should be noted that by dividing the grid area and statistically analyzing the local foam density, a more detailed spatial analysis of the foam distribution can be achieved, avoiding the overall average value from masking local anomalies.

[0149] In step S322, it should be noted that the variance of foam density is used to measure its distribution uniformity. The smaller the variance, the more uniform the foam distribution, which helps to assess the stability of liquid surface tension and the contact state of water blades.

[0150] S330. Based on the top view, obtain the overall characteristics of the foam morphology during the oolong tea steeping process, specifically including:

[0151] S331. Calculate the area S and perimeter A of the foam region, according to... Calculate the area of ​​the foam region

[0152] The ratio of the perimeter to the edge regularity is denoted as R1 of the bubble region.

[0153] S332. Sort the foam edge contours according to pixel connection order, calculate the average curvature continuity of the curve, and record...

[0154] For edge continuity;

[0155] S333. Calculate the minimum bounding rectangle of the foam region, obtain the major axis a and minor axis b of the minimum bounding rectangle, and then... Calculate the edge deviation R2 of the foam region.

[0156] In step S330, it should be noted that the overall characteristics of the foam morphology are mainly used to judge the consistency of tea quality and the stability of the brewing process, including edge regularity, continuity and deviation, which can reflect the integrity of the foam structure and the interfacial behavior of active substances.

[0157] In step S331, it should be noted that regularity measures whether the foam boundary is smooth and round. Complex and broken edges usually indicate uneven activity of the tea soup or fluctuations in tea quality.

[0158] In step S332, it should be noted that edge continuity reflects the stability during the foam formation process. Poor continuity may indicate problems such as unstable interfacial tension or abnormal gas-liquid ratio.

[0159] In step S333, it should be noted that the deviation indicates whether the foam structure has deformed or deviated from the ideal circular structure. An excessively high deviation usually indicates abnormal tea release behavior or uneven liquid surface disturbance.

[0160] S340. Based on the side view, obtain the foam particle size distribution characteristics during the oolong tea steeping process, specifically including:

[0161] S341. Identify the diameter of each bubble using edge detection and contour analysis algorithms, and count the diameter of each bubble.

[0162] S342. Preset a foam diameter window, and count the proportion of foam particles within the target diameter window, which is recorded as the foam particle size distribution characteristic.

[0163] In step S340, it should be noted that particle size distribution is an important indicator for measuring foam stability and bubble formation mechanism. Particle size concentration indicates that the foam structure is stable and the release behavior is standardized, which helps to identify whether the brewing is sufficient.

[0164] In step S341, it should be noted that edge algorithms such as Sobel and Canny are used to identify the outer contour of the foam, and then the diameter is extracted using the minimum circumcircle or ellipse to quantify the size of each foam.

[0165] In step S342, it should be noted that by screening the proportion of foam within a specific diameter range, the concentration of foam can be assessed to avoid abnormally large or small bubbles dominating the assessment of foam stability.

[0166] S350. Based on the side view, obtain the uniformity of foam layer thickness during the oolong tea steeping process, specifically including:

[0167] S351. Divide the side view image of the foam area into several equal regions along the horizontal direction. Calculate the foam layer thickness in each region, and calculate the mean μ and standard deviation σ of the foam layer thickness values. Calculate the uniformity of foam thickness.

[0168] In step S350, it should be noted that the spatial consistency of the foam layer thickness is related to the viscosity of the tea soup surface and the equilibrium state of the liquid surface tension, which is a key indicator for evaluating local extraction intensity and interfacial behavior.

[0169] In step S351, it should be noted that the upper and lower boundaries of the foam are extracted using image segmentation and edge fitting algorithms to obtain the thickness value. Then, the mean and standard deviation of the thickness of all regions are calculated to characterize the thickness uniformity. The smaller the standard deviation, the more stable the layer thickness.

[0170] S360. Based on the side view, determine whether there are particulate matter during the oolong tea steeping process, specifically including:

[0171] S361. Apply grayscale threshold segmentation or a deep learning-based target detection model to the foam area to distinguish between bubbles and particulate matter.

[0172] S362. Calculate the area, quantity, and diameter of all particles.

[0173] In step S360, it should be noted that the presence of impurities in the tea soup (such as fine tea dust, tea powder or foreign matter) may seriously affect the drinking experience and the purity of the tea soup. Therefore, timely identification through visual detection is of great significance.

[0174] In step S361, it should be noted that the grayscale feature method is suitable for scenes with obvious contrast, while deep models (such as YOLO, Mask uniformity R-CNN, etc.) are suitable for complex backgrounds and can achieve high-precision particle recognition and separation.

[0175] In step S362, it should be noted that these characteristic indicators help to determine the degree of impurities and are used to compare with the set threshold to determine whether to trigger filtration or issue a brewing abnormality warning.

[0176] Furthermore, judging the brewing level of oolong tea based on foam morphology information includes:

[0177] S410. Determine whether the following indicators meet the corresponding preset threshold range:

[0178] Does the foam coverage rate meet the preset foam coverage rate threshold range?

[0179] Does the foam uniformity meet the preset foam uniformity threshold range?

[0180] Whether the foam particle size distribution characteristics meet the preset foam particle size threshold range;

[0181] Does the uniformity of the foam layer thickness meet the preset thickness uniformity threshold range?

[0182] In this step, it should be noted that the judgment process is used to quantitatively compare foam morphology information to objectively assess the current brewing state of the oolong tea. The system pre-sets ideal threshold ranges for each indicator, corresponding to foam coverage, foam uniformity, foam particle size distribution characteristics, and foam layer thickness uniformity, respectively. By comparing with real-time collected data, it determines whether each indicator is within the normal range, providing a basis for subsequent brewing strategy decisions.

[0183] S420. If any one of the conditions is not met, then based on the direction and magnitude of the deviation, identify the current brewing state as one of the following states:

[0184] Initial extraction stage: foam coverage is lower than the preset foam coverage threshold range, foam particle size distribution characteristics do not meet the preset foam particle size threshold range, and foam layer thickness uniformity does not meet the preset thickness uniformity threshold range.

[0185] Insufficient component release stage: The foam coverage rate meets the preset foam coverage rate threshold range, but the foam particle size distribution characteristics do not meet the preset foam particle size threshold range, and the foam layer thickness uniformity does not meet the preset thickness uniformity threshold range.

[0186] Over-extraction stage: The foam coverage rate is higher than the preset foam coverage rate threshold range, but the foam particle size distribution characteristics do not meet the preset foam particle size threshold range, and the foam layer thickness uniformity does not meet the preset thickness uniformity threshold range.

[0187] In this step, it's important to note that the direction and degree of deviation from different foam indicators reflect the specific stages and state changes in the brewing process. The system identifies the deviation patterns of each indicator to classify the brewing state, improving the accuracy and intelligence of the identification. Specifically, this includes:

[0188] Initial extraction stage: In this stage, foam has not yet been generated in large quantities, the coverage is low, and the particle size distribution and thickness uniformity are unstable, indicating that the tea leaves have not fully unfolded or the active ingredients have not been released in large quantities;

[0189] Insufficient component release stage: Although the foam coverage is at a normal level in this stage, the particle size distribution and layer thickness uniformity are abnormal, indicating that the tea component release process has not reached a balanced and sufficient state.

[0190] Over-extraction stage: This stage usually involves excessive foam coverage, accompanied by an imbalance in foam particle size distribution and uneven thickness, indicating that the tea has been over-extracted and the flavor may become bitter. Brewing should be stopped immediately.

[0191] Accordingly, the S500 includes:

[0192] S510. If all indicators meet the corresponding threshold range, the brewing process is determined to have reached the target state, and the steeping is stopped, separating the tea liquor from the tea leaves. It should be noted that when all foam morphology indicators are within the ideal range, it indicates that the effective components of the tea have been released evenly and stably, and the foam state has reached an ideal level. Stopping the steeping operation at this point yields the best flavor while avoiding over-steeping that could lead to a decline in quality. Therefore, the system's automatic control device separates the tea liquor from the tea leaves, ending the brewing process.

[0193] S520. If the system identifies the process as "initial extraction stage" or "insufficient component release stage," then extend the steeping time. In this step, it should be noted that the system determines, based on the state recognition results, that the tea leaves have not fully released their flavor or the foam state is unstable. Therefore, it automatically extends the steeping time. Extending the steeping time provides conditions for further release of active substances, improving the consistency and concentration of the tea's flavor, while avoiding the problem of insufficient steeping under manual judgment.

[0194] S530. If the system detects an "over-extraction stage," the brewing process is terminated, and the tea liquor is quickly discharged. It should be noted that in this step, once the system detects that the foam indicators show the brewing has entered an over-extraction state (such as excessive coverage, broken or uneven foam), it will immediately terminate the brewing and quickly separate the tea liquor through the control components. This operation effectively prevents the tea liquor from becoming bitter or astringent due to over-extraction, ensuring beverage quality and improving the stability and intelligent response capabilities of the automatic brewing system.

[0195] The technical solution implemented in this embodiment can accurately extract foam morphology information of oolong tea during the steeping process based on multi-view image information, including multiple dimensions such as coverage, uniformity, particle size distribution, morphological characteristics, layer thickness, and whether it contains particulate impurities, thereby constructing a comprehensive brewing state recognition system. By comparing various indicators with preset thresholds and combining deviation trends for intelligent judgment, the specific stage of brewing (such as initial extraction, insufficient component release, over-extraction, etc.) can be effectively distinguished, thereby achieving dynamic control of brewing time. This method not only improves the automation and standardization level of oolong tea brewing, but also significantly enhances the consistency of tea flavor and quality stability. It is applicable to various application scenarios such as intelligent tea drinking equipment and commercial automatic tea brewing machines, and has significant practical value and promotion prospects.

[0196] In an optional embodiment, the foam morphology information also includes foam color, which is obtained based on the grayscale value of the foam area. Foam color, as a characterizing factor of tea infusion concentration changes, can reflect the degree of release of pigments (such as theaflavins, thearubigins, and theabrownins) in tea leaves. Since foam adheres to the surface of the tea infusion, its color often exhibits different shades of gray depending on the color of the tea infusion. Therefore, statistical analysis of the grayscale values ​​of the foam area can obtain the average color level of the foam, providing supplementary information for further evaluation of brewing concentration and maturity.

[0197] Accordingly, S410 also includes:

[0198] S411. Determine whether the foam color meets the preset foam color range.

[0199] In this step, it's important to note that the uniformity system is pre-defined with a grayscale range representing the ideal foam color. This range is typically based on empirical data or tea quality standards and represents the expected visual characteristics of the foam. Comparing the average grayscale value of the foam area in the real-time image with this range determines whether the tea pigment release has reached the ideal state. If the foam grayscale value is too high (i.e., the color is too light), it may indicate insufficient component release; if the grayscale value is too low (i.e., the color is too dark), it may suggest over-extraction or excessive tea concentration.

[0200] The S420 also includes:

[0201] S421. If any one of the conditions is not met, then based on the direction and magnitude of the deviation, identify the current brewing state as one of the following states:

[0202] Initial extraction stage: foam coverage is lower than the preset foam coverage threshold range, foam particle size distribution characteristics do not meet the preset foam particle size threshold range, foam layer thickness uniformity does not meet the preset thickness uniformity threshold range, and foam color is lower than the preset foam color range.

[0203] Insufficient component release stage: The foam coverage rate meets the preset foam coverage rate threshold range, but the foam particle size distribution characteristics do not meet the preset foam particle size threshold range, the foam layer thickness uniformity does not meet the preset thickness uniformity threshold range, and the foam color is lower than the preset foam color range.

[0204] Over-extraction stage: The foam coverage rate is higher than the preset foam coverage rate threshold range, but the foam particle size distribution characteristics do not meet the preset foam particle size threshold range, the foam layer thickness uniformity does not meet the preset thickness uniformity threshold range, and the foam color is higher than the preset foam color range.

[0205] In this step, it's important to note that foam color uniformity, as a supplementary indicator, can be used in conjunction with foam structure characteristics to identify the brewing state, thereby improving the accuracy of state classification. For example, in the initial extraction stage, because the tea leaves haven't fully unfurled and less tea polyphenols and pigments are released, the foam color is usually lighter and the gray value is higher. In the insufficient component release stage, although the foam surface area and distribution may be normal, a lighter color still indicates insufficient internal release. In the over-extraction stage, the foam color is darker and often has a lower gray value, indicating that the tea soup is too strong and the flavor tends to be bitter. Combining this with other abnormal foam structure indicators can strengthen the confirmation of the over-extraction state.

[0206] The technical solution implemented in this embodiment introduces foam color parameters on top of the original foam morphology analysis, enabling visual quantitative analysis of the release of pigment components during oolong tea brewing, thus enhancing the system's judgment of brewing level. Foam color, as an indirect representation of flavor release, supplements the "concentration level" information that traditional structural indicators struggle to reflect, facilitating more precise identification of brewing stages and dynamic adjustment of control strategies. This solution improves the intelligent response capability and flavor assurance level of the automatic brewing system, demonstrating good scalability and practical value.

[0207] Furthermore, S421 also includes identifying the current brewing state as one of the following states:

[0208] The tea soup is too strong or the aroma is too intense: the foam thickness uniformity does not meet the preset thickness uniformity threshold range, the foam color is higher than the preset foam color range, and the foam uniformity does not meet the preset foam uniformity threshold range.

[0209] In this step, it's important to note that this state is used to identify problems such as an overly strong tea flavor and flavor imbalance caused by excessive release during brewing. Specifically, if the foam thickness uniformity does not meet the threshold, it may indicate that the foam structure is beginning to collapse or aggregate, suggesting a large release of tea cell contents and reduced foam stability; if the foam color is higher than the preset range, it reflects that the tea concentration is too high and pigments are excessively released; abnormal foam uniformity (such as a mixture of large and small bubbles, or a loose structure) further suggests an imbalance in the concentration of aroma and flavor substances, especially an overly concentrated concentration of aromatic oils, resulting in an overly strong aroma and a poor drinking experience. Therefore, based on the above characteristics, it can be identified as a state of "overly strong tea and overly strong aroma".

[0210] Accordingly, the S500 includes:

[0211] S540. If the tea soup is identified as "too strong" or "too strong in aroma", diluents are added, including syrups and fruit juices.

[0212] In this step, it's important to note that to address the issue of overly strong tea, the system can automatically trigger the blending module to introduce flavoring and diluting ingredients, such as syrups and fruit juices, to reduce bitterness and balance the overall flavor. Syrups not only enhance sweetness and mask bitterness but also improve the mouthfeel by adjusting the tea's viscosity; fruit juices introduce natural fruit aromas that complement the tea's fragrance, mitigating the pungent smell caused by an overly strong aroma. Furthermore, by preset dilution ratios or intelligently controlling the amount added, personalized flavor adjustments and stable output can be achieved.

[0213] The technical solution implemented in this embodiment adds a mechanism to identify "overly strong tea soup and overly intense aroma" based on traditional foam structure recognition, and includes a strategy for adding dilution additives. This not only enables refined identification of the tea soup state but also allows for dynamic correction of brewing quality through automation. This solution effectively mitigates flavor deviations caused by over-extraction, improves product consistency and user satisfaction, demonstrates the intelligent brewing system's proactive intervention capability in controlling end-user flavor, and possesses good application adaptability and commercial prospects.

[0214] Furthermore, S421 also includes identifying the current brewing state as one of the following states:

[0215] Poor tea quality or abnormal brewing: edge regularity is lower than the preset edge regularity threshold, edge continuity is lower than the preset edge continuity threshold, and edge deviation is higher than the preset edge deviation threshold.

[0216] In this step, it's important to note that analyzing the morphological characteristics of the foam edges through image recognition can reflect the stability and uniformity of the foam's structure during brewing. Low edge regularity indicates irregular foam formation, potentially stemming from impurities or insufficient content in the tea leaves. Low edge continuity suggests breakage or discontinuity during foam formation, indicating instability in the extraction process. High edge deviation signifies asymmetrical foam expansion, possibly caused by equipment malfunction, abnormal water flow, or tea quality issues. Joint analysis of these features can accurately identify problems at the raw material or equipment level.

[0217] Accordingly, the S500 includes:

[0218] S550 If the problem is identified as "poor tea quality or abnormal brewing", an error message will be issued, the current brewing process will be paused, and manual intervention or switching of raw materials will be required for re-brewing.

[0219] The technical solution implemented in this embodiment enables early identification and response to raw material defects or equipment malfunctions, effectively preventing the output of substandard tea. By automatically pausing and prompting for manual intervention, system safety and product quality control capabilities are enhanced, ensuring user experience and product consistency.

[0220] Furthermore, S421 also includes identifying the current brewing state as one of the following states:

[0221] Tea leaves are not fully dispersed and the concentration is uneven: the thickness uniformity of the foam does not meet the preset thickness uniformity threshold range.

[0222] In this step, it should be noted that the uniformity of foam thickness reflects the uniformity of the distribution of dissolved substances in the tea soup. If this indicator does not meet the set range, it means that the tea leaves are unevenly distributed in the water, resulting in floating or sedimentation, which leads to local concentrations that are too high or too low, thereby affecting the taste of the tea soup and the stability of foam formation.

[0223] Accordingly, the S500 includes:

[0224] S560. If it is identified as "tea leaves are not sufficiently dispersed or the concentration is uneven", then control the stirring component to extend from top to bottom below the surface of the tea liquid in the brewing vessel.

[0225] S561, Start the stirring operation for the preset duration;

[0226] S562. After stirring is complete, control the stirring component to remove the tea soup.

[0227] The technical solution implemented in this embodiment compensates for the uneven distribution of tea leaves through an automated stirring mechanism, which can improve dissolution efficiency and tea soup homogeneity, stabilize foam structure, improve the taste of beverages, and further enhance the adaptability and intelligence of the brewing process.

[0228] Furthermore, S421 also includes identifying the current brewing state as one of the following states:

[0229] The tea soup contains impurities: the particulate matter coverage exceeds the preset particulate matter coverage threshold; or, the maximum particle diameter is greater than the preset particle size threshold.

[0230] In this step, it's important to note that the presence of visible particles (such as tea leaves, foreign objects, etc.) in the brewed tea will directly affect the appearance and taste of the beverage. The system uses image recognition to determine the particle coverage and maximum particle size; when these exceed a set threshold, it can be determined that impurities are present in the tea.

[0231] Accordingly, the S500 includes:

[0232] S570. If the tea soup is identified as containing impurities, the tea soup in the brewing vessel shall be filtered.

[0233] The technical solution implemented in this embodiment removes impurities through automated filtration, ensuring the clarity and purity of the tea soup, improving product quality and consumer satisfaction, and enhancing the cleanliness and intelligent maintenance capabilities of the brewing system.

[0234] Furthermore, the foam morphology information also includes the foam growth rate, which is obtained by calculating the increment of foam coverage within a preset time interval.

[0235] Accordingly, S421 also includes identifying the current brewing state as one of the following states:

[0236] Water temperature too low: The foam growth rate is lower than the preset foam growth rate threshold range;

[0237] Water temperature too high: The foam growth rate exceeds the preset foam growth rate threshold range;

[0238] S430. Determine whether the increase in foam coverage exceeds the preset coverage increment window.

[0239] In this step, it's important to note that the foam growth rate reflects the activity of the extraction reaction. A significantly low foam growth rate may indicate insufficient water temperature, resulting in slow release of the active ingredients from the tea leaves; conversely, a high growth rate may suggest excessively high water temperature, causing the tea leaves to release aroma and other substances too quickly, disrupting the brewing rhythm. Incremental calculations can quickly pinpoint water temperature anomalies.

[0240] The S500 includes:

[0241] S580 If "water temperature is too high" is identified, a first compensating water flow is injected, wherein the temperature of the first compensating water flow is lower than the preset temperature in S200;

[0242] S581. If the water temperature is identified as "too high", then a second compensating water flow is injected, wherein the temperature of the second compensating water flow is higher than the preset temperature in S200.

[0243] The technical solution implemented in this embodiment dynamically adjusts the water temperature by monitoring the foam growth rate, thereby achieving closed-loop feedback regulation of temperature control during the brewing process. This effectively stabilizes the foam growth trend, ensures that the extraction process operates within the optimal thermal range, and enhances the flavor reproduction of the tea soup and the system's adaptive control capability.

[0244] The present invention also provides a tea brewing method and apparatus based on intelligent vision guidance, applied to the above-mentioned tea brewing method based on intelligent vision guidance, comprising:

[0245] The brewing vessel, used to hold tea leaves and water, is the physical unit that holds the tea in the entire tea preparation process. It can be made of transparent or semi-transparent material to allow the image acquisition module to obtain clear image data.

[0246] The image acquisition module, located around the perimeter of the brewing vessel, includes a top-view camera and a side-view camera, used to acquire real-time images of the foam during the brewing process. The top view primarily observes the distribution range and edge morphology of the foam, while the side view is used to assess dimensional information such as the thickness and changing trend of the foam layer.

[0247] The image processing and recognition module is used to process, segment, and analyze the acquired image information, and extract and recognize the following foam morphology information:

[0248] Foam coverage: This indicates the proportion of the brewing liquid surface covered by foam.

[0249] Foam layer thickness: reflects the vertical accumulation of foam;

[0250] Foam color: used to back-calculate concentration and during the extraction stage;

[0251] Foam particle size distribution characteristics: representing the particle size composition of air bubbles in foam;

[0252] Foam morphological characteristics include edge regularity, continuity, and deviation.

[0253] Foam growth rate: Reflects the dynamic growth rate of foam and is related to the brewing reaction activity.

[0254] The judgment and control module, based on the above foam morphology parameters, performs the following functions:

[0255] Determine the current concentration and brewing level of the tea;

[0256] Identify any abnormal conditions during the brewing process (such as over-extraction, uneven concentration, abnormal water temperature, etc.);

[0257] Issue corresponding control commands to guide subsequent regulatory operations such as adding water, stirring, draining liquid, or prompting intervention.

[0258] The water injection component, based on the instructions of the judgment and control module, precisely injects water of different temperatures or volumes into the brewing vessel to achieve: initial water injection for brewing; temperature compensation (such as when the water temperature is too low or too high); and dilution control.

[0259] The stirring component, when the system determines that the tea leaves are not sufficiently dispersed or the tea soup concentration is uneven, is controlled to extend downwards below the surface of the tea soup and perform a stirring operation for a preset duration before automatically retracting. This structure helps improve the uniformity of contact between tea leaves and water, and alleviates the problem of excessively high or low concentrations in certain areas.

[0260] The filter component is used to identify the presence of impurities or large tea particles in the tea infusion. If the particle coverage exceeds the limit or the particle size exceeds the standard, the component is triggered to perform a filtration operation, improving the cleanliness of the tea infusion and the drinking experience.

[0261] The auxiliary ingredient addition component is triggered by the system when indicators such as foam color and thickness indicate that the tea soup concentration is too high or the aroma is too strong. This component injects diluents (such as syrup or fruit juice) to adjust the flavor concentration and meet different user preferences.

[0262] The drainage component and waste liquid container are used to automatically discharge a portion of the tea soup or waste (such as tea leaves or overly concentrated liquid) according to the control strategy. The discharged liquid enters the waste liquid container located below, preventing contamination of the main channel and achieving automatic cleaning.

[0263] The time control component, based on the concentration assessment results, dynamically adjusts the brewing time, automatically extending or terminating the steeping time. This avoids over-extraction and adapts to the personalized brewing needs of different teas.

[0264] The tea preparation equipment provided by this invention has the following beneficial effects:

[0265] Enhance the level of intelligence: By judging the state of the tea soup through multi-dimensional image parameters, the process can be upgraded from "passive brewing" to "intelligent blending".

[0266] Enhance tea consistency and quality control: A visual recognition-based feedback loop system can dynamically adjust key brewing parameters to ensure stable tea flavor.

[0267] It supports adaptive handling of multiple abnormal working conditions: such as uneven concentration, insufficient extraction, and excessive tea residue, etc., can be automatically identified and intervened, reducing reliance on manual intervention.

[0268] Adapting to personalized beverage needs: Customized beverage preparation can be achieved through adjustable components such as temperature control, stirring, and ingredient addition.

[0269] Reduce failure risk and operating costs: Improve system self-maintenance capabilities and safety through waste liquid management and anomaly alert mechanisms.

[0270] It is suitable for a variety of tea types and extraction methods: not limited to oolong tea, but can be extended to green tea, black tea, flower tea and other fields, with broad market adaptability.

[0271] The following is a preferred embodiment provided in this application, describing the intelligent vision-guided tea brewing method provided by the present invention, applicable to the automated brewing of oolong tea. The method includes the following steps:

[0272] Step 1: Setting up the brewing vessel and acquiring images

[0273] A cylindrical brewing vessel made of transparent material, with a top diameter of approximately 80mm, was selected. A top-view camera and two side-view cameras were installed on three sides of the vessel, with resolutions of 1000×1000 and 1920×1080 pixels respectively. These were used to acquire top-view and side-view image data of the tea leaves during the brewing process.

[0274] Step 2: Add tea leaves and hot water

[0275] Add 3g of dried oolong tea leaves to the brewing vessel. Depending on the recipe, you can also add syrup or fruit juice. Pour approximately 200ml of 90℃ hot water from top to bottom through the water inlet, and moisten the tea leaves before pouring.

[0276] Step 3: Image Acquisition and Foam Recognition

[0277] Image acquisition was initiated immediately after water injection, and the images were processed and analyzed as follows:

[0278] Step 3.1, Foam Coverage Identification

[0279] The tea soup area in the top view was extracted into a circular region (with a radius of approximately 400 pixels), and the image was segmented using a U-Net neural network.

[0280] The area of ​​the foam region A1 = 410,000 pixels, and the total area of ​​the circular region A2 ≈ 502,655 pixels;

[0281] The foam coverage rate R = A1 / A2 ≈ 81.6%, which is greater than the set threshold of 80%, and is therefore determined to be "the cup is full of foam".

[0282] Step 3.2, Identification of Foam Distribution Uniformity

[0283] The top view is divided into a 5×5 grid, resulting in 25 regions.

[0284] The foam density in each area is [15%, 12%, 10%, ...];

[0285] Calculate the variance σ 2 The information entropy H≈3.4 and information entropy H≈4.5 indicate that the foam distribution is relatively uniform.

[0286] Step 3.3: Identification of Overall Foam Morphology Features

[0287] Foam area = 62800 pixels, edge perimeter = 1470 pixels;

[0288] The perimeter-to-area ratio R1 = 2.74, which indicates slight irregularity.

[0289] The connectivity score is 1.0 (no breaks), the deviation ratio is approximately 0.086, and the shape is approximately symmetrical.

[0290] Step 3.4: Identification of Foam Particle Size Distribution

[0291] The side view calibration scale is 1mm = 12 pixels;

[0292] 47 bubbles were detected, with an average particle size of ≈2.3 mm and a median particle size of ≈2.1 mm;

[0293] The proportion of bubbles with a diameter of 1-4mm is approximately 87.2%, which is considered a qualified bubble diameter.

[0294] Step 3.5: Identification of foam layer thickness uniformity

[0295] The average thickness of the foam was measured to be approximately 5.03 mm.

[0296] The standard deviation σ≈0.19mm and the coefficient of variation CV≈0.038 indicate that the layer thickness is stable and the structure is good.

[0297] Step 3.6, Impurity Identification

[0298] The total area of ​​particles in the foam is 1200 pixels, accounting for 8% of the total area.

[0299] The maximum pixel size is approximately 250 pixels, and the average pixel size is approximately 80 pixels, indicating that the pixel density is too high.

[0300] Step 4: Judging the brewing degree

[0301] Based on the bubble indicator and the set threshold range, the current data is judged as follows:

[0302] Coverage rate meets the requirement: This indicates that active substances such as tea polyphenols are fully released;

[0303] Good uniformity: extraction is balanced and liquid surface tension is stable;

[0304] Particle size is within acceptable limits: foam structure is stable;

[0305] The morphological indicators are slightly irregular, but the edge structure is reasonable;

[0306] The layer thickness is moderate, the CV value is low, and the structure is stable.

[0307] The detected particle density is higher than the threshold, indicating a large amount of tea residue.

[0308] The brewing state is determined as follows: the target state has been reached, but filtration needs to be performed.

[0309] Step 5: Control Logic and Operation Response

[0310] The control module triggers the "End Brewing" command, controlling the draining component to transfer the tea liquid to vessel 2;

[0311] At the same time, with high particle coverage, the system switches to the filtration channel and filters through a microporous membrane.

[0312] If the foam color is too dark or the foam is too thick, the following may be triggered depending on the degree of deviation:

[0313] S540: The auxiliary material addition module injects fruit juice for dilution;

[0314] S580: Inject cold water to lower the temperature of the soup and inhibit oxidation.

[0315] This implementation method achieves the following technical advantages: It offers rich recognition dimensions and high analysis accuracy, using six major categories of visual indicators—coverage, uniformity, morphological parameters, bubble diameter, thickness, and particle size—to perform detailed analysis of the brewing state. It provides real-time operation response and a high degree of automation, dynamically determining whether to continue steeping, add water or ingredients, or perform filtration or stirring based on the foam state. Objective judgment replaces subjective experience, achieving standardized, visualized, and intelligent control of the brewing process, making it particularly suitable for deployment in commercial automated tea beverage systems.

[0316] The embodiments described above are merely illustrative of several implementations of this application, and while the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the patent application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the scope of protection of this application. Therefore, the scope of protection of this patent application should be determined by the appended claims.

Claims

1. A tea preparation method based on intelligent vision guidance, characterized in that, include: S100. Add oolong tea leaves to a brewing vessel. Multiple cameras are installed around the brewing vessel to capture image information of the oolong tea steeping process. The sidewalls of the brewing vessel are made of transparent material. S200. Water at a preset temperature is poured into the brewing vessel from top to bottom; S300. Based on the image information, identify the foam morphology information of oolong tea during the steeping process. The foam morphology information includes: foam coverage, foam uniformity, overall foam morphology characteristics, foam particle size distribution characteristics, and foam layer thickness uniformity. S400. Determining the brewing level of oolong tea based on the foam morphology information includes: The degree of release of surfactants in tea leaves is determined based on the foam coverage rate. Based on the foam uniformity, the extraction balance and surface flow of the tea soup are determined. Based on the overall characteristics of the foam morphology, the consistency of tea quality, the physical stability of the tea soup, and the suitability of the current brewing stage can be determined. Based on the foam particle size distribution characteristics, determine whether the release of active molecules in the tea soup is sufficient and stable; The viscosity of the tea soup and the local extraction intensity are determined based on the uniformity of the foam layer thickness. S500: Dynamically adjust the brewing process according to the brewing degree.

2. The tea preparation method based on intelligent vision guidance according to claim 1, characterized in that, The image information includes a top view and a side view. The process of identifying the foam morphology of oolong tea during the steeping process based on the image information includes: S310. Based on the top view, obtain the foam coverage rate during the oolong tea steeping process, specifically including: S311. Extract the foam region using image segmentation technology, calculate the ratio of the foam region to the coverage area of ​​the upper surface of the brewing vessel, and record it as the foam coverage rate. S320. Based on the top view, obtain the foam uniformity during the oolong tea steeping process, specifically including: S321, Divide the upper surface of the brewing vessel into... Each region is recorded with a specific number of bubbles. Calculate the ratio of the foam area of ​​each region to the area of ​​that region, and record it as the foam density of that region; S322. Calculate the variance of the foam density in all areas, and denot it as the foam uniformity. S330. Based on the top view, obtain the overall characteristics of the foam morphology during the oolong tea steeping process, specifically including: S331. Calculate the area S and perimeter A of the foam region, based on... Calculate the ratio of the area to the perimeter of the foam region, and denot it as the edge regularity of the foam region. ; S332. Sort the foam edge contours according to the pixel connection order, calculate the average curvature continuity of the curve, and denot it as edge continuity. S333. Calculate the minimum bounding rectangle of the foam region, obtain the major axis a and minor axis b of the minimum bounding rectangle, and according to... Calculate the edge deviation of the foam region. ; S340. Based on the side view, obtain the foam particle size distribution characteristics during the oolong tea steeping process, specifically including: S341. Identify the diameter of each bubble using edge detection and contour analysis algorithms, and count the diameter of each bubble. S342. Preset a foam diameter window, and count the proportion of foams within the foam diameter window, which is recorded as the foam particle size distribution characteristic. S350. Based on the side view, obtain the uniformity of the foam layer thickness during the oolong tea steeping process, specifically including: S351. Divide the side view image of the foam region into several equal regions along the horizontal direction, calculate the foam layer thickness in each region, and calculate the average value of the foam layer thickness. and standard deviation ,according to Calculate the uniformity of the foam thickness; S360. Based on the side view, determine whether there are particulate matter during the oolong tea steeping process, specifically including: S361. Apply grayscale threshold segmentation or a deep learning-based target detection model to the foam area to distinguish between bubbles and particulate matter. S362. Calculate the area, quantity, and diameter of all particles.

3. The tea brewing method based on intelligent vision guidance according to claim 2, characterized in that, The step of determining the brewing level of oolong tea based on the foam morphology information includes: S410. Determine whether the following indicators meet the corresponding preset threshold range: Does the foam coverage rate meet the preset foam coverage rate threshold range? Does the foam uniformity meet the preset foam uniformity threshold range? Whether the foam particle size distribution characteristics meet the preset foam particle size threshold range; Whether the thickness uniformity of the foam layer meets the preset thickness uniformity threshold range; S420. If any one of the conditions is not met, then based on the direction and magnitude of the deviation, identify the current brewing state as one of the following states: Initial extraction stage: The foam coverage is lower than the preset foam coverage threshold range, the foam particle size distribution characteristics do not meet the preset foam particle size threshold range, and the foam layer thickness uniformity does not meet the preset thickness uniformity threshold range. Insufficient component release stage: The foam coverage rate meets the preset foam coverage rate threshold range, but the foam particle size distribution characteristics do not meet the preset foam particle size threshold range, and the foam layer thickness uniformity does not meet the preset thickness uniformity threshold range. Over-extraction stage: The foam coverage rate is higher than the preset foam coverage rate threshold range, but the foam particle size distribution characteristics do not meet the preset foam particle size threshold range, and the foam layer thickness uniformity does not meet the preset thickness uniformity threshold range. Accordingly, S500 includes: S510. If all the indicators meet the corresponding threshold range, it is determined that the current brewing level has reached the target state, the steeping ends, and the tea soup and tea leaves are separated. S520. If the process is identified as "initial extraction stage" or "insufficient component release stage", then extend the soaking time. S530. If the "over-extraction stage" is identified, the brewing process will be terminated and the tea soup will be quickly discharged.

4. The tea brewing method based on intelligent vision guidance according to claim 3, characterized in that, The foam morphology information also includes foam color, which is obtained based on the grayscale value of the foam region. Accordingly, S410 further includes: S411. Determine whether the color of the foam meets the preset foam color range; The S420 also includes: S421. If any one of the conditions is not met, then based on the direction and magnitude of the deviation, identify the current brewing state as one of the following states: Initial extraction stage: The foam coverage is lower than the preset foam coverage threshold range, the foam particle size distribution characteristics do not meet the preset foam particle size threshold range, the foam layer thickness uniformity does not meet the preset thickness uniformity threshold range, and the foam color is lower than the preset foam color range. Insufficient component release stage: The foam coverage rate meets the preset foam coverage rate threshold range, but the foam particle size distribution characteristics do not meet the preset foam particle size threshold range, the foam layer thickness uniformity does not meet the preset thickness uniformity threshold range, and the foam color is lower than the preset foam color range. Over-extraction stage: The foam coverage rate is higher than the preset foam coverage rate threshold range, but the foam particle size distribution characteristics do not meet the preset foam particle size threshold range, the foam layer thickness uniformity does not meet the preset thickness uniformity threshold range, and the foam color is higher than the preset foam color range.

5. The tea brewing method based on intelligent vision guidance according to claim 4, characterized in that, S421 further includes identifying the current brewing state as one of the following states: The tea soup is too strong or the aroma is too intense: the foam thickness uniformity does not meet the preset thickness uniformity threshold range, the foam color is higher than the preset foam color range, and the foam uniformity does not meet the preset foam uniformity threshold range. Accordingly, S500 includes: S540. If the tea soup is identified as "too strong" or "too strong in aroma", then diluents are added, including syrup and fruit juice.

6. The tea brewing method based on intelligent vision guidance according to claim 4, characterized in that, S421 further includes identifying the current brewing state as one of the following states: Poor tea quality or abnormal brewing: The edge regularity is lower than the preset edge regularity threshold, the edge continuity is lower than the preset edge continuity threshold, and the edge deviation is higher than the preset edge deviation threshold. Accordingly, S500 includes: S550 If the problem is identified as "poor tea quality or abnormal brewing", an error message will be issued, the current brewing process will be paused, and manual intervention or switching of raw materials will be required for re-brewing.

7. The tea preparation method based on intelligent vision guidance according to claim 4, characterized in that, S421 further includes identifying the current brewing state as one of the following states: The tea leaves are not fully dispersed and the concentration is uneven: the thickness uniformity of the foam does not meet the preset thickness uniformity threshold range; Accordingly, S500 includes: S560. If it is identified as "tea leaves are not sufficiently dispersed or have uneven concentration", then control the stirring component to extend from top to bottom below the surface of the tea liquid in the brewing vessel. S561, Start the stirring operation for the preset duration; S562. After stirring is completed, control the stirring component to remove the tea soup.

8. The tea brewing method based on intelligent vision guidance according to claim 4, characterized in that, S421 further includes identifying the current brewing state as one of the following states: The tea soup contains impurities: the particulate matter coverage exceeds a preset particulate matter coverage threshold; or, the maximum particulate matter diameter is greater than a preset particle size threshold. Accordingly, S500 includes: S570. If it is identified that "the tea soup contains impurities", then the tea soup in the brewing vessel is filtered.

9. The tea brewing method based on intelligent vision guidance according to claim 4, characterized in that, The foam morphology information also includes the foam growth rate, which is obtained by calculating the increment of the foam coverage within a preset time interval. Accordingly, S421 further includes identifying the current brewing state as one of the following states: Water temperature too low: The foam growth rate is lower than the preset foam growth rate threshold range; Water temperature too high: The foam growth rate exceeds the preset foam growth rate threshold range; S430. Determine whether the increment of the foam coverage rate exceeds the preset coverage rate increment window; The S500 includes: S580. If "water temperature is too high" is detected, a first compensating water flow is injected, wherein the temperature of the first compensating water flow is lower than the preset temperature in S200. S581. If the water temperature is identified as "too high", then a second compensating water flow is injected, wherein the temperature of the second compensating water flow is higher than the preset temperature in S200.

Images