Baking process and device for coffee beans based on blade baking method

Abstract

The embodiment of the invention provides a coffee bean baking process and device based on a blade baking method. In one specific embodiment, the device comprises a fluidized bed type full hot air baking machine, a controller and a baking identification assembly. According to the embodiment, the baking recognition assembly is adopted to obtain the image data of the coffee beans in the baking machine in real time and extract the average size and the average density of the coffee beans, so that the real-time baking degree of the coffee beans is accurately recognized, a baker can more accurately recognize the baking degree, the flow of searching the optimal baking curve by the baker is simplified, and the baking efficiency is improved. And the workload of a baker for searching an optimal baking curve is reduced. A baker can formulate a proper baking scheme by combining the baking curve according to the variety, the production place and the expected flavor characteristics of the coffee beans.

Description

Technical Field

[0001] The present invention relates to the field of coffee bean roasting, and more specifically, to a coffee bean roasting process and device based on a knife edge roasting method. Background Art

[0002] When using a roasting technique such as the Blade Roast method for ultra-light roasting, the degree of roasting needs to be precisely controlled to ensure that the coffee beans are neither over-charred nor retain the off-flavors of the raw beans, thereby retaining the original flavor of the coffee while showing a more delicate and layered taste.

[0003] The existing roasting process generally requires roasters to estimate the roasting degree of coffee beans by observing the color changes, roasting temperature, and roasting time through the observation window on the roaster. This method is largely limited to the roaster's personal experience and intuition, and lacks objective and precise standards to measure the precise relationship between various parameters in the roasting process and the final roasting degree. Therefore, it is difficult to ensure the consistency and repeatability of each roasting.

[0004] To this end, roasters are usually required to test the best roasting curve so that they can formulate a suitable roasting plan based on the variety, origin and expected flavor characteristics of the coffee beans and the roasting curve. However, in the process of exploring the best roasting curve, roasters have to constantly conduct roasting experiments based on their own experience and intuition; the roasting experiment process lacks a clear and systematic rule, which leads to a significant increase in the number of trial and error and costs.

[0005] Especially in the process of ultra-light roasting, roasters need to explore the best state of each type of bean as much as possible while ensuring that the coffee beans are fully cooked. This critical point is very subtle and difficult to accurately reach through simple color value judgment or roasting time control. Once over-roasted, the unique flavor characteristics of ultra-light roasting will be lost; while under-roasting may cause the coffee beans to have a raw bean taste or astringency. The roaster can only determine the current roasting degree based on his own experience, which brings great difficulties to the roaster's exploration of the optimal roasting curve. Summary of the invention

[0006] The purpose of the present disclosure is to provide a roasting process and device for coffee beans based on a blade roasting method that can identify the current roasting degree in real time, so as to solve at least one of the problems existing in the prior art.

[0007] In order to achieve the above objectives, the present invention adopts the following technical solutions:

[0008] A first aspect of the present disclosure provides a coffee bean roasting device based on a blade roasting method, comprising:

[0009] Fluidized bed full hot air roaster, used to roast green coffee beans in full hot air roasting mode and collect real-time roasting temperature;

[0010] A controller, used to obtain a target baking degree and a target baking temperature; read the real-time baking temperature collected by the baking machine and control the baking machine to be preheated to the target baking temperature, and control the baking temperature of the baking machine based on a PID algorithm, with the real-time baking temperature as a feedback amount of the PID algorithm and the target baking temperature as a given amount of the PID algorithm;

[0011] a roasting identification component, used to acquire image data of coffee beans, extract a first average size and a first average density of the coffee beans, so as to identify a real-time roasting degree of the coffee beans and output the result to the controller;

[0012] The controller is further configured to generate and output a baking completion instruction when the real-time baking degree is equal to the target baking degree.

[0013] Further, the baking recognition component includes an image collector, a feature extraction module, and a recognition module;

[0014] The image collector is used to collect the initial image of the coffee beans in the roaster;

[0015] The feature extraction module is used to extract a first average size and a first average density of coffee beans in the initial image;

[0016] The recognition module is used to input the first average size and the first average density of the coffee beans into a trained machine learning model to obtain the real-time roasting degree of the coffee beans.

[0017] Further, the feature extraction module includes a conversion module, a shape extraction module, a contour recognition module, a density calculation module, and a size calculation module;

[0018] The conversion module is used to convert the initial image into a grayscale image;

[0019] The shape extraction module is used to perform binarization processing on the grayscale image, extract the shape features of the coffee beans based on a preset threshold, and obtain a binary image containing the shape features of the coffee beans;

[0020] The contour recognition module is used to identify all connected areas in the binary image based on an edge detection algorithm, and extract contour areas of multiple coffee beans;

[0021] The density calculation module calculates the grayscale mean of each identified connected area to obtain a calculated density of each coffee bean; and calculates a first average density of the coffee beans according to the calculated density;

[0022] The size calculation module is used to extract the contour area of ​​each coffee bean, calculate the area and perimeter of each coffee bean respectively, calculate the average area of ​​the coffee beans according to the area of ​​each coffee bean, and calculate the average perimeter of the coffee beans according to the perimeter of each coffee bean; and calculate the first average size of the coffee beans based on the average area and the average perimeter.

[0023] Further, calculating the first average size of the coffee beans based on the average area and the average perimeter includes:

[0024] According to the formula

[0025]

[0026] Calculate the average major axis length of the coffee beans and the average minor axis length of coffee beans in, is the average circumference of the coffee beans; is the average area of ​​the coffee beans;

[0027] According to the formula

[0028]

[0029] Calculate a first average size of the coffee beans

[0030] Furthermore, the feature extraction module also includes an alarm module;

[0031] The size calculation module is further used to output the first average size of the coffee beans to the alarm module;

[0032] The alarm module is used to determine whether the first average size of the coffee beans is within the preset size range according to the first average size of the coffee beans, and output an alarm signal if it is not within the preset size range.

[0033] The preset size range is from 13 mesh to 18 mesh.

[0034] Further, the baking recognition component also includes a training module;

[0035] The training module is used to obtain a plurality of images of coffee beans at different roasting stages to obtain model training images; the model training images are configured to be labeled with roasting degrees of the coffee beans;

[0036] The feature extraction module is further used to extract the second average size and the second average density of the coffee beans in each of the model training images to obtain a first data set;

[0037] Dividing the first data set into a training set and a test set;

[0038] Build a machine learning model and input the training set into the machine learning model training to obtain a trained machine learning model.

[0039] Furthermore, the machine learning model is a neural network model.

[0040] Furthermore, the target roasting degree is very light roasting.

[0041] The target baking temperature is 200°C to 250°C.

[0042] A second aspect of the present disclosure provides a roasting process for coffee beans based on a blade roasting method, comprising:

[0043] Roast the green coffee beans using a fluidized bed full hot air roaster and collect the real-time roasting temperature;

[0044] Obtaining a target baking degree and a target baking temperature through a controller; reading the real-time baking temperature collected by the baking machine and controlling the baking machine to be preheated to the target baking temperature, and controlling the baking temperature of the baking machine based on a PID algorithm, with the real-time baking temperature as a feedback amount of the PID algorithm and the target baking temperature as a given amount of the PID algorithm;

[0045] Acquiring image data of coffee beans through a roasting identification component, extracting a first average size and a first average density of the coffee beans to identify a real-time roasting degree of the coffee beans and outputting the data to the controller;

[0046] When the real-time baking degree is equal to the target baking degree, the controller generates and outputs a baking completion instruction.

[0047] Further, the acquiring of image data of coffee beans by the roasting identification component, extracting a first average size and a first average density of the coffee beans to identify the real-time roasting degree of the coffee beans and outputting the data to the controller includes:

[0048] The roasting identification component is used to collect an initial image of coffee beans in a roaster; a first average size and a first average density of the coffee beans in the initial image are extracted; and the first average size and the first average density of the coffee beans are input into a trained machine learning model to obtain a real-time roasting degree of the coffee beans.

[0049] The beneficial effects of the present disclosure are as follows:

[0050] The present disclosure adopts a roasting recognition component to obtain image data of coffee beans in a roaster in real time, extract the average size and average density of the coffee beans, and thus accurately identify the real-time roasting degree of the coffee beans. The roaster can more accurately grasp the current real-time roasting degree based on the current real-time roasting degree, and on this basis, combine the roaster's own roasting experience (such as judging the roasting degree by the color change of the coffee beans and the roasting temperature, etc.), so as to combine the roasting degree obtained based on objective laws with the roaster's own roasting experience, which can greatly improve the efficiency of exploring the optimal roasting curve.

[0051] Compared with the traditional roasting process which often relies on the experience of the roaster to judge the degree of roasting, the present invention reduces the difficulty of roasting through an automated and intelligent control method, making the roasting process more controllable. At the same time, since the roasting temperature and degree can be accurately controlled, the waste of coffee beans caused by over-roasting or under-roasting is reduced, and the roasting cost is reduced.

[0052] In addition, the present disclosure adopts a fluidized bed full hot air roaster, which distributes the hot air evenly in the roasting chamber by forced circulation of hot air. This uniformity can ensure that the coffee beans are heated evenly throughout the roasting process, avoiding the problem of uneven roasting caused by uneven hot air that may occur in a drum roaster, and the coffee bean expansion rate is high, thereby improving the coffee extraction rate.

[0053] The fluidized bed hot air roaster can precisely adjust the temperature, wind speed and roasting time. Operators can make precise adjustments based on different coffee bean characteristics and roasting requirements to achieve more consistent and high-quality roasting results. You can challenge the roasting limits according to product requirements and produce roasted products that are more characteristic of the coffee origin.

[0054] Fluidized bed hot air roaster uses forced circulation hot air, which can more effectively transfer heat to the surface and inside of coffee beans, thus completing the roasting process in a relatively short time. This not only improves production efficiency, but also reduces energy consumption. The fastest roasting time can be controlled within 3 minutes (traditional 8-18 minutes).

[0055] The fluidized bed hot air roaster can adapt to different types and batches of coffee beans because its operating parameters can be adjusted according to specific circumstances. This flexibility makes it perform well in handling diverse roasting needs.

[0056] Compared with the complex structure of semi-hot air (or full hot air) drum roasters, fluidized bed full hot air roasters are usually simpler in design and relatively easy to clean and maintain. This is of great significance for reducing downtime and improving production efficiency. Fluidized bed full hot air roasters are selected and promoted by Jinmilan for their advantages such as uniform hot air, controllable baking process, high energy efficiency, strong adaptability and easy cleaning and maintenance.

[0057] The present invention adopts a control method combining a fluidized bed full hot air roaster with a PID algorithm to achieve precise control of roasting temperature and roasting degree. The PID algorithm can dynamically adjust the heating power of the roaster according to the difference between the real-time roasting temperature and the target roasting temperature, thereby ensuring that the coffee beans reach a predetermined roasting degree during the roasting process, without being over-coked or retaining the off-flavor of the raw beans. BRIEF DESCRIPTION OF THE DRAWINGS

[0058] The specific implementation methods of the present disclosure are further described in detail below with reference to the accompanying drawings.

[0059] Figure 1 A schematic structural diagram of a coffee bean roasting device based on a blade roasting method according to the present invention is shown.

[0060] Figure 2 A schematic diagram showing the structure of a roasting identification component of the present disclosure is shown.

[0061] Figure 3 A structural schematic diagram of a feature extraction module disclosed in the present invention is shown.

[0062] Figure 4 A flow chart showing a roasting process of coffee beans based on the blade roasting method of the present disclosure.

[0063] Figure 5 A schematic diagram of the structure of a computer system that implements the controller, feature extraction module, and recognition module provided by an embodiment of the present disclosure is shown. DETAILED DESCRIPTION

[0064] In order to more clearly illustrate the present disclosure, the present disclosure is further described below in conjunction with the embodiments and drawings. Similar components in the drawings are represented by the same reference numerals. It should be understood by those skilled in the art that the content described below is illustrative rather than restrictive, and should not be used to limit the scope of protection of the present disclosure.

[0065] like Figure 1 As shown, the present invention provides a coffee bean roasting device based on a blade roasting method, comprising:

[0066] Fluidized bed full hot air roaster, used to roast green coffee beans in full hot air roasting mode and collect real-time roasting temperature;

[0067] A controller is used to obtain a target baking degree and a target baking temperature; read the real-time baking temperature collected by the baking machine and control the baking machine to preheat to the target baking temperature, and control the baking temperature of the baking machine based on a PID algorithm, with the real-time baking temperature as a feedback amount of the PID algorithm and the target baking temperature as a given amount of the PID algorithm;

[0068] a roasting identification component, used to obtain image data of coffee beans, extract a first average size and a first average density of the coffee beans, so as to identify a real-time roasting degree of the coffee beans and output the result to a controller;

[0069] The controller is further configured to generate and output a baking completion instruction when the real-time baking degree is equal to the target baking degree. In this embodiment, the controller is connected to a touch display module to display the baking completion instruction output by the controller for reference by the baker.

[0070] It should be noted that, in this embodiment, the PID algorithm is a proportional, integral, and derivative control algorithm, which is a feedback control algorithm based on proportional, integral, and derivative operations.

[0071] The fluidized bed full hot air baking machine in this embodiment is a fluidized bed hot air baking machine.

[0072] The present disclosure adopts a roasting recognition component to obtain image data of coffee beans in a roaster in real time, extract the average size and average density of the coffee beans, and thus accurately identify the real-time roasting degree of the coffee beans. The roaster can more accurately grasp the current real-time roasting degree based on the current real-time roasting degree, and on this basis, combine the roaster's own roasting experience (such as judging the roasting degree by the color change of the coffee beans and the roasting temperature, etc.), so as to combine the roasting degree obtained based on objective laws with the roaster's own roasting experience, which can greatly improve the efficiency of exploring the optimal roasting curve.

[0073] Compared with the traditional roasting process, which often relies on the experience of the roaster to judge the degree of roasting, which increases the difficulty and uncertainty of roasting, the present invention reduces the difficulty of roasting through an automated and intelligent control method, making the roasting process more controllable. At the same time, since the roasting temperature and degree can be accurately controlled, the waste of coffee beans caused by over-roasting or under-roasting is reduced, and the roasting cost is reduced.

[0074] In addition, the present disclosure adopts a fluidized bed full hot air roaster, which distributes the hot air evenly in the roasting chamber by forced circulation of hot air. This uniformity can ensure that the coffee beans are heated evenly throughout the roasting process, avoiding the problem of uneven roasting caused by uneven hot air that may occur in a drum roaster, and the coffee bean expansion rate is high, thereby improving the coffee extraction rate.

[0075] The fluidized bed hot air roaster can precisely adjust the temperature, wind speed and roasting time. Operators can make precise adjustments based on different coffee bean characteristics and roasting requirements to achieve more consistent and high-quality roasting results. You can challenge the roasting limits according to product requirements and produce roasted products that are more characteristic of the coffee origin.

[0076] Fluidized bed hot air roaster uses forced circulation hot air, which can more effectively transfer heat to the surface and inside of coffee beans, thus completing the roasting process in a relatively short time. This not only improves production efficiency, but also reduces energy consumption. The fastest roasting time can be controlled within 3 minutes (traditional 8-18 minutes).

[0077] The fluidized bed hot air roaster can adapt to different types and batches of coffee beans because its operating parameters can be adjusted according to specific circumstances. This flexibility makes it perform well in handling diverse roasting needs.

[0078] Compared with the complex structure of semi-hot air (full hot air) drum roasters, fluidized bed full hot air roasters are usually simpler in design and relatively easy to clean and maintain. This is of great significance for reducing downtime and improving production efficiency. Fluidized bed full hot air roasters are selected and promoted by Jinmilan for their advantages such as uniform hot air, controllable baking process, high energy efficiency, strong adaptability and easy cleaning and maintenance.

[0079] The present invention adopts a control method combining a fluidized bed full hot air roaster with a PID algorithm to achieve precise control of roasting temperature and roasting degree. The PID algorithm can dynamically adjust the heating power of the roaster according to the difference between the real-time roasting temperature and the target roasting temperature, thereby ensuring that the coffee beans reach a predetermined roasting degree during the roasting process, without being over-coked or retaining the off-flavor of the raw beans.

[0080] It should be noted that during the coffee bean roasting experiment, the roaster can dynamically adjust the target roasting temperature, and the roasting time can also be controlled by the roaster. The roasting degree output by the controller is mainly used as a reference for the roaster.

[0081] In this embodiment, there may be multiple target baking temperatures, and the controller may obtain the corresponding target baking temperature at the corresponding preset time to achieve multi-temperature automatic baking control.

[0082] In one possible implementation, the controller is further configured to obtain the first average size and the first average density of the coffee beans identified by the feature extraction module and output them for user reference.

[0083] The roaster can also judge the roasting degree of the coffee beans according to the roasting completion instruction, the first average size and the first average density of the coffee beans and in combination with his own experience.

[0084] In one possible implementation, Figure 2 As shown, the baking recognition component includes an image collector, a feature extraction module, and a recognition module;

[0085] An image collector for collecting an initial image of coffee beans in a roaster;

[0086] a feature extraction module, for extracting a first average size and a first average density of coffee beans in the initial image;

[0087] The recognition module is used to input the first average size and the first average density of the coffee beans into the trained machine learning model to obtain the real-time roasting degree of the coffee beans.

[0088] In one possible implementation, Figure 3 As shown, the feature extraction module includes a conversion module, a shape extraction module, a contour recognition module, a density calculation module, and a size calculation module;

[0089] The conversion module is used to convert the initial image into a grayscale image. In this embodiment, the initial color image containing coffee beans is read. Then, the color image is converted into a grayscale image using a color space conversion technology. In the grayscale image, the brightness information of the coffee beans is retained, while the color information is removed.

[0090] The shape extraction module is used to perform binarization processing on the grayscale image, extract the shape features of the coffee beans based on a preset threshold, and obtain a binary image containing the shape features of the coffee beans; this is achieved by setting a preset threshold, setting the part with a pixel value higher than the threshold to white (or black), and the part below the threshold to black (or white). In this way, the pixels in the grayscale image are divided into two categories: pixels above the threshold are set to white, indicating the coffee bean area; pixels below the threshold are set to black, indicating the background area. In this way, a binary image containing only black and white colors is obtained.

[0091] The contour recognition module is used to identify all connected areas in the binary image based on the edge detection algorithm and extract the contour areas of multiple coffee beans;

[0092] In this embodiment, an edge detection algorithm is applied to the binary image, and the edge detection algorithm may be a Canny edge detector, a Sobel operator, or a Laplacian operator, etc., to identify the edge between the coffee beans and the background. The edge detection algorithm can detect the location where the gray value in the image changes dramatically, thereby outlining the outline of the coffee beans.

[0093] Use a connected region analysis algorithm (such as depth-first search DFS, breadth-first search BFS or mark and clear algorithm) to identify all connected regions in the binary image. A connected region refers to a set of interconnected white pixels in the image, and each connected region corresponds to a coffee bean.

[0094] The density calculation module calculates the grayscale mean of each identified connected area to obtain the calculated density of each coffee bean; and calculates the first average density of the coffee beans according to the calculated density;

[0095] In this embodiment, the calculation process includes extracting the grayscale values ​​of all pixels in each connected region.

[0096] The grayscale values ​​of each connected region are summed and then divided by the number of pixels in the region to obtain the grayscale mean of the region.

[0097] In image processing, the density of coffee beans can be obtained by calculating the grayscale mean of coffee beans. The grayscale mean reflects the brightness level of the image area, and the brightness level is related to the reflectivity of the object, which is further related to the density of the object. Specifically, for the same variety of coffee beans, multiple different time points can be selected during the roasting process and the cooling process;

[0098] At each time point, the initial image of the coffee beans is obtained and the density value of the coffee beans is measured; the grayscale mean and the density value corresponding to the grayscale value of the coffee beans in the initial image are extracted; thereby, the corresponding relationship between the grayscale mean and the density value of the coffee beans is obtained; in addition, the present embodiment can also construct a machine learning model, and the grayscale mean and the density value corresponding to the grayscale value of the coffee beans obtained at each time point are constructed as a data set, and the data set is divided into a training set and a validation set, and the machine learning model is trained with the training set to obtain a trained machine learning model. Therefore, in the present embodiment, after obtaining the grayscale mean, it is input into the trained machine learning model to obtain the density of the corresponding coffee beans. The machine learning model can be a convolutional machine learning model, a fully connected neural network, etc., which is not limited here. In the present embodiment, the density of the coffee beans is automatically obtained based on the grayscale mean, which has a high computational efficiency relative to the traditional experimental method.

[0099] The size calculation module is used to extract the contour area of ​​each coffee bean, calculate the area and perimeter of each coffee bean respectively, calculate the average area of ​​the coffee beans according to the area of ​​each coffee bean, and calculate the average perimeter of the coffee beans according to the perimeter of each coffee bean; calculate the first average size of the coffee beans based on the average area and the average perimeter.

[0100] The conversion module converts the initial image into a grayscale image, which greatly simplifies the subsequent processing steps because the grayscale image only contains brightness information, reducing the complexity of data processing. At the same time, this conversion makes the subsequent binarization processing more efficient and can more accurately extract the shape features of coffee beans.

[0101] The present invention uses a shape extraction module to perform binarization processing on the grayscale image using a preset threshold, effectively separates the coffee beans from the background, and extracts clear shape features of the coffee beans.

[0102] The contour recognition module can accurately identify all connected areas in the binary image based on the edge detection algorithm, that is, the contour area of ​​the coffee beans. This step is crucial for the subsequent density and size calculations, ensuring the accuracy and reliability of the calculations.

[0103] The density calculation module calculates the grayscale mean of each identified connected area to obtain the calculated density of each coffee bean. It calculates the first average density of coffee beans based on these calculated densities. The size calculation module calculates the area and perimeter of each coffee bean by extracting the contour area of ​​each coffee bean, and calculates the first average size of the coffee beans based on these data, providing important data support for evaluating the roasting degree of coffee beans.

[0104] In one possible implementation, calculating the first average size of the coffee beans based on the average area and the average perimeter includes:

[0105] According to the formula

[0106]

[0107] Calculate the average major axis length of coffee beans and the average minor axis length of coffee beans in, is the average circumference of coffee beans; is the average area of ​​coffee beans;

[0108] According to the formula

[0109]

[0110] Calculate the first average size of coffee beans

[0111] In one possible implementation, the feature extraction module further includes an alarm module;

[0112] The size calculation module is further used to output the first average size of the coffee beans to the alarm module;

[0113] The alarm module is used to determine whether the first average size of the coffee beans is within a preset size range according to the first average size of the coffee beans, and output an alarm signal if it is not within a preset size range.

[0114] In one possible implementation, the preset size range is 13 mesh to 18 mesh.

[0115] When the size of coffee beans does not meet the requirements, it is difficult to reach the ideal roasting temperature. Smaller coffee beans can easily reach high temperatures quickly, while larger coffee beans require higher temperatures or longer times to ensure uniform roasting, which ultimately affects the quality of the coffee beans.

[0116] In one possible implementation, the roast identification component further includes a training module;

[0117] A training module, used for acquiring a plurality of images of coffee beans at different roasting stages to obtain model training images; the model training images are configured to be labeled with roasting degrees of the coffee beans;

[0118] The feature extraction module is further used to extract the second average size and the second average density of coffee beans in each model training image to obtain a first data set;

[0119] Divide the first data set into a training set and a test set;

[0120] Build a machine learning model and input the training set into the machine learning model training to obtain a trained machine learning model.

[0121] In one possible implementation, the machine learning model is a neural network model. It should be noted that the neural network model in this embodiment can be a convolutional neural network (CNN) model, etc., as long as it can identify the roasting degree of coffee beans, and there is no limitation here.

[0122] In a possible implementation, the target baking degree is very light baking, and the target baking temperature is 200°C to 250°C.

[0123] It should be noted that, in this embodiment, according to the SCAA roasting degree classification, the Agtron color value corresponding to Ultra-Light Roast is 85-95, and the roasting time is shorter than the first crack intensive period to retain the original flavor of the coffee beans. Generally, in the ultra-light roasting stage, the average density of the coffee beans decreases by 5%-8%, and the expansion rate is 10%-15%.

[0124] In a possible implementation, the coffee bean roasting device based on the blade roasting method further includes a storage warehouse for storing green coffee beans.

[0125] In this embodiment, within the appropriate humidity range, the coffee beans can be effectively prevented from becoming moldy due to excessive humidity, thereby maintaining their original flavor and quality.

[0126] The right temperature helps slow down the natural spoilage of coffee beans, prolong their shelf life, and maintain their freshness and aroma. If the temperature is too high, it may accelerate the oxidation process of coffee beans, resulting in loss of flavor; while if the temperature is too low, the moisture in the coffee beans may freeze, affecting the quality.

[0127] Coffee beans stored under suitable temperature and humidity conditions are more likely to achieve uniform roasting results during the roasting process. This is because a stable storage environment helps maintain the stability of the internal structure of the coffee beans, allowing them to be heated evenly during roasting.

[0128] Appropriate storage conditions help preserve the flavor substances in the coffee beans so that they can be fully released after roasting, improving the taste and aroma of the coffee.

[0129] It should be noted that the knife edge roasting method disclosed in the present invention is a technology of extremely light roasting through fluidized bed full hot air; in order to fully display the flavor and acidity of the coffee beans, a roasting process requires precise control of the roasting time. This control method is like dancing on the edge of a knife, and is called the knife edge roasting method.

[0130] In this embodiment, the fluidized bed full hot air baking machine includes:

[0131] Hot air circulation system, the hot air circulation system includes:

[0132] High-pressure centrifugal fan: Located at the bottom of the roaster, it is used to generate high-speed airflow to force hot air through the roasting chamber.

[0133] Heating module: uses electric heating tubes or gas burners, linked with the fan, to heat the air to the target temperature.

[0134] Air duct design: The annular guide structure is adopted to ensure that the hot air enters the bottom of the baking chamber evenly, and disperses the airflow through the porous distribution plate (aperture 1mm-3mm) to avoid local high temperature.

[0135] Waste heat recovery system: A heat exchanger is installed at the exhaust port to recover part of the heat energy for preheating fresh air and improving energy efficiency.

[0136] Baking room:

[0137] Material and structure: Double-layer stainless steel cavity, the inner layer is a high-temperature resistant ceramic coating (anti-adhesion), and the outer layer is a heat insulation layer (to reduce heat loss).

[0138] Coffee bean suspension area: The airflow through the porous plate at the bottom makes the coffee beans suspended and tumble, forming a fluidized bed state, ensuring that each bean is evenly heated.

[0139] Observation window: A high-temperature resistant glass window is installed on the top or side wall, which works with an image collector (such as a CCD camera) to capture the status of coffee beans in real time.

[0140] Coffee bean conveying system:

[0141] Feed port: Located at the top of the roasting chamber, the speed of feeding green beans is controlled by a vibrating feeder.

[0142] Discharge port: A quick discharge valve is designed at the bottom, which opens automatically after roasting is completed, and the coffee beans are transferred to the cooling plate through a spiral conveyor belt.

[0143] Cooling module: Integrated cyclone separator and cold air system, cools coffee beans to room temperature within 10 seconds to lock in the flavor.

[0144] Temperature control module

[0145] Temperature sensor: distributed inside the roasting chamber, monitors the air flow temperature and bean surface temperature in real time.

[0146] PID Controller: Dynamically adjusts baking temperature based on sensor feedback.

[0147] like Figure 4 As shown, another embodiment of the present invention provides a roasting process of coffee beans based on a blade roasting method, comprising:

[0148] Step S1, roasting the green coffee beans in a full hot air roasting mode by a fluidized bed full hot air roaster, and collecting the real-time roasting temperature;

[0149] Step S2, obtaining a target baking degree and a target baking temperature through a controller; reading a real-time baking temperature collected by a baking machine and controlling the baking machine to preheat to the target baking temperature, and controlling the baking temperature of the baking machine based on a PID algorithm, with the real-time baking temperature as a feedback amount of the PID algorithm and the target baking temperature as a given amount of the PID algorithm;

[0150] Step S3, acquiring image data of coffee beans through a roasting identification component, extracting a first average size and a first average density of the coffee beans to identify a real-time roasting degree of the coffee beans and outputting the data to a controller;

[0151] Step S4: When the real-time baking degree is equal to the target baking degree, the controller generates and outputs a baking completion instruction.

[0152] Acquiring image data of coffee beans through a roasting identification component, extracting a first average size and a first average density of the coffee beans to identify a real-time roasting degree of the coffee beans and outputting the data to a controller includes:

[0153] Through the roasting recognition component, an initial image of coffee beans in a roaster is collected; the first average size and the first average density of the coffee beans in the initial image are extracted; the first average size and the first average density of the coffee beans are input into a trained machine learning model to obtain the real-time roasting degree of the coffee beans.

[0154] like Figure 5As shown, a computer system suitable for implementing the controller, feature extraction module, and recognition module provided by the above-mentioned embodiment includes a central processing module (CPU), which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) or a program loaded from a storage portion into a random access memory (RAM). In the RAM, various programs and data required for the operation of the computer system are also stored. The CPU, ROM, and RAM are connected to each other via a bus. An input / output (I / O) interface is also connected to the bus.

[0155] The following components are connected to the I / O interface: an input part including a keyboard, a mouse, etc.; an output part including a liquid crystal display (LCD), a speaker, etc., a storage part including a hard disk, etc.; and a communication part including a network interface card such as a LAN card, a modem, etc. The communication part performs communication processing via a network such as the Internet. A drive is also connected to the I / O interface as needed. Removable media, such as a magnetic disk, an optical disk, a magneto-optical disk, a semiconductor memory, etc., are installed on the drive as needed so that the computer program read therefrom is installed into the storage part as needed.

[0156] In particular, according to this embodiment, the process described in the flowchart above can be implemented as a computer software program. For example, this embodiment includes a computer program product, which includes a computer program tangibly contained on a computer-readable medium, and the computer program includes program code for executing the method shown in the flowchart. In such an embodiment, the computer program can be downloaded and installed from a network through a communication part, and / or installed from a removable medium.

[0157] The flow chart and schematic diagram in the accompanying drawings illustrate the possible architecture, function and operation of the system, method and computer program product of the present embodiment. In this regard, each square box in the flow chart or schematic diagram can represent a part of a module, program segment or code, and a part of the above-mentioned module, program segment or code contains one or more executable instructions for realizing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the square box can also occur in a sequence different from that marked in the accompanying drawings. For example, two square boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each square box in the schematic diagram and / or flow chart, and the combination of the square boxes in the schematic diagram and / or flow chart can be implemented with a dedicated hardware-based system that performs a specified function or operation, or can be implemented with a combination of dedicated hardware and computer instructions.

[0158] In the description of the present disclosure, it should be noted that the orientation or positional relationship indicated by the terms "upper", "lower", etc. is based on the orientation or positional relationship shown in the accompanying drawings, and is only for the convenience of describing the present disclosure and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as a limitation on the present disclosure. Unless otherwise clearly specified and limited, the terms "installed", "connected", and "connected" should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection, or it can be indirectly connected through an intermediate medium, and it can be the internal connection of two elements. For those of ordinary skill in the art, the specific meanings of the above terms in the present disclosure can be understood according to the specific circumstances.

[0159] It should also be noted that, in the description of the present disclosure, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that the process, method, article or equipment including a series of elements includes not only those elements, but also includes other elements not explicitly listed, or also includes elements inherent to such process, method, article or equipment. In the absence of further restrictions, the elements defined by the statement "comprise a ..." do not exclude the presence of other identical elements in the process, method, article or equipment including the elements.

[0160] Obviously, the above embodiments of the present disclosure are merely examples for clearly illustrating the present disclosure, and are not limitations on the implementation methods of the present disclosure. For ordinary technicians in this field, other different forms of changes or modifications can be made on the basis of the above description. It is impossible to list all the implementation methods here. All obvious changes or modifications derived from the technical solution of the present disclosure are still within the protection scope of the present disclosure.

Claims

1. A coffee bean roasting device based on a blade roasting method, characterized in that: include, Fluidized bed full hot air roaster, used to roast green coffee beans in full hot air roasting mode and collect real-time roasting temperature; A controller, used to obtain a target baking degree and a target baking temperature; read the real-time baking temperature collected by the baking machine and control the baking machine to be preheated to the target baking temperature, and control the baking temperature of the baking machine based on a PID algorithm, with the real-time baking temperature as a feedback amount of the PID algorithm and the target baking temperature as a given amount of the PID algorithm; a roasting identification component, configured to acquire image data of coffee beans, extract a first average size and a first average density of the coffee beans to identify a real-time roasting degree of the coffee beans and output the data to the controller; The controller is further configured to generate and output a baking completion instruction when the real-time baking degree is equal to the target baking degree.

2. The baking device according to claim 1, characterized in that: The baking recognition component includes an image collector, a feature extraction module, and a recognition module; The image collector is used to collect the initial image of the coffee beans in the roaster; The feature extraction module is used to extract a first average size and a first average density of coffee beans in the initial image; The recognition module is used to input the first average size and the first average density of the coffee beans into a trained machine learning model to obtain the real-time roasting degree of the coffee beans.

3. The baking device according to claim 2, characterized in that: The feature extraction module includes a conversion module, a shape extraction module, a contour recognition module, a density calculation module, and a size calculation module; The conversion module is used to convert the initial image into a grayscale image; The shape extraction module is used to perform binarization processing on the grayscale image, extract the shape features of the coffee beans based on a preset threshold, and obtain a binary image containing the shape features of the coffee beans; The contour recognition module is used to identify all connected areas in the binary image based on an edge detection algorithm, and extract contour areas of multiple coffee beans; The density calculation module is used to calculate the grayscale mean of each identified connected area to obtain the calculated density of each coffee bean; and calculate the first average density of the coffee beans according to the calculated density; The size calculation module is used to extract the contour area of ​​each coffee bean, calculate the area and perimeter of each coffee bean respectively, calculate the average area of ​​the coffee beans according to the area of ​​each coffee bean, and calculate the average perimeter of the coffee beans according to the perimeter of each coffee bean; and calculate the first average size of the coffee beans based on the average area and the average perimeter.

4. The baking device according to claim 3, characterized in that: Calculating the first average size of the coffee beans based on the average area and the average perimeter comprises: According to the formula Calculate the average major axis length of the coffee beans and the average minor axis length of coffee beans in, is the average circumference of the coffee beans; is the average area of ​​the coffee beans; According to the formula Calculate a first average size of the coffee beans 5. The baking device according to claim 3, characterized in that: The feature extraction module also includes an alarm module; The size calculation module is further used to output the first average size of the coffee beans to the alarm module; The alarm module is used to determine whether the first average size of the coffee beans is within the preset size range according to the first average size of the coffee beans, and output an alarm signal if it is not within the preset size range; The preset size range is from 13 mesh to 18 mesh.

6. The baking device according to claim 2, characterized in that: The baking recognition component also includes a training module; The training module is used to obtain a plurality of images of coffee beans at different roasting stages to obtain model training images; the model training images are configured to be labeled with roasting degrees of the coffee beans; The feature extraction module is further used to extract the second average size and the second average density of the coffee beans in each of the model training images to obtain a first data set; Dividing the first data set into a training set and a test set; Build a machine learning model and input the training set into the machine learning model training to obtain a trained machine learning model.

7. The baking device according to claim 6, characterized in that: The machine learning model is a neural network model.

8. The baking device according to claim 1, characterized in that: The target baking degree is very light baking, and the target baking temperature is 200° C. to 250° C.

9. A roasting process for coffee beans based on a knife edge roasting method, characterized in that: include, Roast the green coffee beans using a fluidized bed full hot air roaster and collect the real-time roasting temperature; Obtaining a target baking degree and a target baking temperature through a controller; reading the real-time baking temperature collected by the baking machine and controlling the baking machine to be preheated to the target baking temperature, and controlling the baking temperature of the baking machine based on a PID algorithm, with the real-time baking temperature as a feedback amount of the PID algorithm and the target baking temperature as a given amount of the PID algorithm; Acquiring image data of coffee beans through a roasting identification component, extracting a first average size and a first average density of the coffee beans to identify a real-time roasting degree of the coffee beans and outputting the data to the controller; When the real-time baking degree is equal to the target baking degree, the controller generates and outputs a baking completion instruction.

10. The baking process according to claim 9, characterized in that: The method of acquiring image data of coffee beans through a roasting identification component, extracting a first average size and a first average density of the coffee beans to identify the real-time roasting degree of the coffee beans and outputting the data to the controller includes: By means of the roasting identification component, an initial image of coffee beans in the roasting machine is collected; extracting a first average size and a first average density of coffee beans in the initial image; The first average size and the first average density of the coffee beans are input into the trained machine learning model to obtain the real-time roasting degree of the coffee beans.

Images