Method for detecting water content of thalli and method for grading thalli

Abstract

The invention provides a thallus water content detection method and grading method, which comprises the following steps: irradiating a light source comprising a plurality of spectral components with different wavelengths to a unit area region of thallus, so that the thallus transmits reflected light back to a probe, and the probe transmits the reflected light to a spectrograph to generate a reflected light spectrum; on the basis of a spectrum-water content mapping model, the water content of the thallus is determined according to the intensity characteristics of the sensitive wave band in the reflected light spectrum, and in the spectrum-water content mapping model, when the water content of the thallus is different, the intensity of single-wavelength reflected light in the sensitive wave band of the spectrum is different; when the water content of the thalli is certain, the intensities of reflected light with different wavelengths in the sensitive wave band of the spectrum are different, and the intensity difference value of the reflected light with different wavelengths is related to the water content of the thalli and the wavelength difference value of the reflected light. The water content of the thalli can be accurately and rapidly measured, and the classification accuracy of the thalli can be improved.

Description

Technical Field

[0001] The present invention belongs to the field of intelligent grading, and in particular relates to a bacterial water content detection method and a grading method. Background Art

[0002] Morels, named for the honeycomb-like grooves on their cap that resemble a sheep's belly, are a rare and valuable edible and medicinal fungus. They contain a variety of amino acids and vitamins beneficial to the human body, helping maintain normal metabolism and physiological functions and promoting muscle growth and repair. These nutrients contribute to the unique edible and medicinal properties of morels.

[0003] In the morel processing industry, the grading of morels is of vital importance. The traditional grading method of morels mainly relies on manual work. During manual grading, workers rely on experience and visual observation to identify and classify morels one by one according to their size, shape, color and other characteristics. This method is inefficient and inconsistent, so people have invented automatic grading devices for morels. However, the grading devices currently under research mostly distinguish morels by a single dimension (length, volume). This method ignores the influence of other factors of morels (weight, water content, internal defects) on grading, and is prone to inaccurate grading. Summary of the Invention

[0004] The present invention provides a bacterial water content detection method and a grading method to solve the current problems of being unable to detect bacterial water content and bacterial grading relying on a single dimension and having low grading accuracy.

[0005] According to a first aspect of an embodiment of the present invention, a method for detecting moisture content of bacterial cells is provided, comprising:

[0006] Step S100: irradiating a unit area of a bacterial cell with a light source having multiple spectral components of different wavelengths, so that the bacterial cell transmits reflected light back to a probe, which transmits the reflected light to a spectrometer to generate a reflected light spectrum;

[0007] Step S200: Determine the water content of the bacteria based on the intensity characteristics of the sensitive band in the reflected light spectrum based on a spectrum-water content mapping model. In the spectrum-water content mapping model, the intensity of the reflected light of a single wavelength in the sensitive band of the spectrum varies when the water content of the bacteria is different; when the water content of the bacteria is constant, the intensity of the reflected light of different wavelengths in the sensitive band of the spectrum varies, and the difference in the intensity of the reflected light of different wavelengths is related to the water content of the bacteria and the difference in the wavelength of the reflected light.

[0008] Optionally, the sensitive wavelength is 1350-1500 nm, and after step S200, the method further includes: cross-verifying the determined bacterial water content by combining the polysaccharide absorption peak appearing in the 1200 nm wavelength;

[0009] In step S100, the probe detects the reflected light returned from each unit area of the bacterial cell and transmits the reflected light from each unit area to the spectrometer to generate a reflected light spectrum corresponding to each unit area.

[0010] Before step S200, the method further includes: superimposing and averaging the reflected light spectra corresponding to the respective unit area regions to obtain the reflected light spectrum;

[0011] In step S200 , a plurality of reflected light spectra with known water contents are obtained, and a spectrum-water content mapping model is established using partial least squares regression.

[0012] Optionally, the defects on the bacterial body are different, the reflectivity of the bacterial body in the corresponding defect monitoring band is different, and the intensity abnormality degree in the corresponding defect monitoring band is different; the method also includes: step S300, determining the defect of the bacterial body according to the intensity abnormality degree of the corresponding defect monitoring band in the reflected light spectrum.

[0013] Optionally, before step S300, the method further includes: performing deep learning training on the reflected light spectra of healthy bacteria and moldy bacteria, and extracting spectral features of the moldy bacteria using a convolutional neural network;

[0014] The step S300 specifically includes: based on the spectral characteristics of the moldy bacteria, determining whether the bacteria are moldy according to the abnormality of the intensity of the corresponding defect monitoring band in the reflected light spectrum.

[0015] Optionally, the defect monitoring band includes the 1650nm band and all bands, and the first intensity abnormality appearing in all bands is much greater than the second intensity abnormality appearing in the 1650nm band; when the first intensity abnormality appears in all bands, it indicates that there are holes or hollows on the bacteria, and when the second intensity abnormality appears in the 1650nm band, it indicates that there is mold on the bacteria.

[0016] According to a second aspect of an embodiment of the present invention, a bacterial cell classification method is provided, comprising:

[0017] Step S10: Obtaining the volume of the bacterial body based on the collected three-dimensional data of the bacterial body;

[0018] Step S20: Detecting the water content of the bacteria using the method for detecting water content of the bacteria according to any one of claims 1 to 5;

[0019] Step S30, correcting the volume according to the detected water content;

[0020] Step S40: Classify the bacteria according to the detected water content and the corrected volume.

[0021] Optionally, before step S30, the process further includes establishing a shrinkage rate model in the following manner:

[0022] Measure the volume of multiple fresh bacterial samples;

[0023] Drying the fresh bacterial cell sample according to the set drying process to obtain a dried bacterial cell sample;

[0024] The volume and moisture content of the dried bacterial sample were measured, and the shrinkage rate was determined for each dried bacterial sample with different moisture content. The shrinkage rate was equal to the volume of the dried bacterial sample divided by the volume of the fresh bacterial sample before drying. The shrinkage rate was correlated with the moisture content of the dried bacterial sample to obtain the shrinkage rate model.

[0025] Optionally, the step S30 specifically includes: according to the detected water content, finding the corresponding shrinkage rate from the shrinkage rate model, correcting the volume according to the following formula to obtain the corrected volume: corrected volume = the volume / (1-shrinkage rate).

[0026] Optionally, before step S40, the method further includes: detecting defects of the bacterial cells using the bacterial cell moisture content detection method according to any one of claims 1 to 5; and measuring the weight of the bacterial cells;

[0027] The step S40 specifically includes: grading the bacteria according to the detected water content, the corrected volume, the defects of the bacteria and the weight.

[0028] Optionally, the controller controls the feeding mechanism to sequentially transfer the bacterial cells to a conveyor belt, controls a three-dimensional data acquisition mechanism above the conveyor belt to acquire three-dimensional data of the bacterial cells, and controls a moisture content detection mechanism below the conveyor belt to detect the moisture content of the bacterial cells.

[0029] The controller obtains the volume of the bacteria based on the collected three-dimensional data of the bacteria, corrects the volume according to the detected water content, and grades the bacteria based on the detected water content and the corrected volume. Thereafter, the controller controls the sorting mechanism according to the grading results to sort the bacteria into the collection frame of the corresponding level.

[0030] The beneficial effects of the present invention are:

[0031] 1. The present invention illuminates the bacteria with a light source containing multiple spectral components of different wavelengths. The water content of the bacteria is determined based on the spectrum of reflected light with multiple wavelengths returned by the bacteria. That is, the water content of the bacteria is determined from two dimensions: wavelength and spectral intensity. This can achieve water content measurement and ensure measurement accuracy. The present invention establishes a spectrum-water content mapping model. Based on this spectrum-water content mapping model, the water content of the bacteria is determined based on the intensity characteristics of sensitive bands in the reflected light spectrum, which can improve the speed of water content detection.

[0032] 2. After detecting the moisture content of the bacteria, the present invention also combines the polysaccharide absorption peak appearing in the 1200nm band to cross-validate the determined moisture content of the bacteria, thereby further improving the accuracy of the detection of moisture content of the bacteria. The present invention can more comprehensively reflect the reflected light spectrum of the entire bacteria by superimposing and averaging the reflected light spectra corresponding to each unit area of the bacteria. The moisture content of the bacteria is determined based on the superimposed and averaged reflected light spectrum, which can further ensure the accuracy of the detection of moisture content of the bacteria.

[0033] 3. The present invention can determine the defect of the bacterial cell based on the degree of intensity abnormality in the corresponding defect monitoring band in the reflected light spectrum. The present invention establishes different defect monitoring bands according to the characteristics of different bacterial cell defects, and determines the defect of the bacterial cell based on the different intensity abnormalities in the corresponding defect monitoring band of the reflected light spectrum, which can ensure the accuracy of bacterial defect determination.

[0034] 4. The present invention classifies the cells based on the moisture content and volume of the dried cells, relying on multiple dimensions. This makes the classification results more in line with actual value requirements and realizes the dual-dimensional classification of "size-quality". Before the cells are classified, the present invention corrects the volume of the dried cells, which can improve classification accuracy, enhance processing consistency, and increase added value.

[0035] 5. The present invention classifies bacteria not only based on their water content and corrected volume, but also on their defects and weight. This allows identification of irregular bacteria (e.g., bacteria with damaged caps), allowing for multi-dimensional classification, thereby further improving the accuracy of bacterial classification.

[0036] 6. The present invention uses a controller to control the loading, transportation, three-dimensional data collection, and moisture content detection of the bacteria. After completing the bacteria grading based on the collected three-dimensional data and moisture content, the controller controls the sorting mechanism to sort the bacteria into collection frames of corresponding levels, thereby realizing automatic grading and sorting of the bacteria. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] Figure 1 This is a flow chart of an embodiment of the method for detecting water content of bacterial cells of the present invention;

[0038] Figure 2 This is a flow chart of an embodiment of the bacterial cell classification method of the present invention;

[0039] Figure 3 This is a schematic structural diagram of an embodiment of a bacterial grading device of the present invention;

[0040] Figure 4 It is a front perspective view of an embodiment of the feeding mechanism of the present invention;

[0041] Figure 5 This is a front view of an embodiment of the feeding mechanism of the present invention;

[0042] Figure 6 It is a side view of an embodiment of the feeding mechanism of the present invention;

[0043] Figure 7 It is a top view of an embodiment of the feeding mechanism of the present invention;

[0044] Figure 8 This is a three-dimensional diagram of an embodiment of the feeding mechanism of the present invention;

[0045] Figure 9 It is a schematic structural diagram of an embodiment of the bacterial cell moisture content detection mechanism of the present invention. DETAILED DESCRIPTION

[0046] In order to enable those skilled in the art to better understand the technical solutions in the embodiments of the present invention and to make the above-mentioned purposes, features and advantages of the embodiments of the present invention more obvious and easy to understand, the technical solutions in the embodiments of the present invention are further described in detail below with reference to the accompanying drawings.

[0047] In the description of the present invention, unless otherwise specified and limited, it should be noted that the term "connection" should be understood in a broad sense. For example, it can be a mechanical connection or an electrical connection, or it can be the internal connection between two elements. It can be a direct connection or an indirect connection through an intermediate medium. For ordinary technicians in this field, the specific meaning of the above terms can be understood according to the specific circumstances.

[0048] See also Figure 1 , is a flow chart of an embodiment of a method for detecting moisture content in bacterial cells according to the present invention. The method for detecting moisture content in bacterial cells may include:

[0049] Step S100: irradiate a unit area of a bacterial cell with a light source including a plurality of spectral components of different wavelengths, so that the bacterial cell transmits reflected light back to a probe, and the probe transmits the reflected light to a spectrometer to generate a reflected light spectrum.

[0050] In this embodiment, the bacteria can be transported on a transparent light-transmitting conveyor belt, and the light source generator and the probe can be located directly below the conveyor belt. The light source generator can irradiate the light source onto each unit area of the bacteria respectively, and the light source can be near-infrared light, irradiating the bottom surface of the fungus's umbrella; the probe can be a fiber-optic coupled near-infrared probe, which can be arranged in a matrix form at intervals directly below the conveyor belt (for example, at intervals of 15 mm), and connected to the spectrometer via optical fiber.

[0051] Research has found that the higher the water content in the bacteria, the lower its reflectivity to near-infrared light. The larger the wavelength of the light source, the lower the bacteria's reflectivity to near-infrared light. The lower the reflectivity, the lower the intensity of the reflected light. This shows that the bacteria's reflectivity to the light source is not only related to the water content of the bacteria, but also to the wavelength of the light source. If a single-wavelength light source is used to illuminate the bacteria, the cell's reflectivity, and thus the cell's water content, is determined based on the intensity of the reflected light returned by the bacteria. This only reflects the cell's water content from a single dimension, resulting in low cell water content measurement accuracy. To this end, the present invention illuminates the bacteria with a light source containing multiple spectral components of different wavelengths. The cell water content is determined based on the spectrum of the reflected light with multiple wavelengths returned by the bacteria. This two-dimensional determination of cell water content from both wavelength and spectral intensity allows for accurate cell water content measurement.

[0052] Step S200: Determine the moisture content of the bacteria based on the intensity characteristics of the sensitive band in the reflected light spectrum based on a spectrum-moisture content mapping model. In the spectrum-moisture content mapping model, when the moisture content of the bacteria is different, the reflectivity of the bacteria is correspondingly different, and the intensity of the reflected light of a single wavelength in the sensitive band of the spectrum is different. When the moisture content of the bacteria is constant, the intensity of the reflected light of different wavelengths in the sensitive band of the spectrum is different, and the difference in the intensity of the reflected light of different wavelengths is related to the moisture content of the bacteria and the difference in the wavelength of the reflected light.

[0053] In the present embodiment, it has been found through research that the sensitive band of the bacteria can be the 1350-1500nm band (for morels, the sensitive band can be the 1450nm band). After the step S200, the present invention may further include: combining the polysaccharide absorption peak that appears in the 1200nm band to cross-validate the determined bacteria water content. The present invention establishes a spectrum-water content mapping model, and based on the spectrum-water content mapping model, determines the bacteria water content according to the intensity characteristics of the sensitive band in the reflected light spectrum, which can improve the bacteria water content detection speed. After detecting the bacteria water content, the present invention also combines the polysaccharide absorption peak that appears in the 1200nm band to cross-validate the determined bacteria water content, thereby further improving the bacteria water content detection accuracy.

[0054] In step S100, the probe can detect reflected light from each unit area of the bacterial cell and transmit the reflected light from each unit area to the spectrometer to generate a reflected light spectrum corresponding to each unit area. Before step S200, the present invention can further include superimposing and averaging the reflected light spectra corresponding to each unit area to obtain a reflected light spectrum. By superimposing and averaging the reflected light spectra corresponding to each unit area of the bacterial cell, the present invention can more comprehensively reflect the reflected light spectrum of the entire bacterial cell. The moisture content of the bacterial cell can be determined based on the superimposed and averaged reflected light spectrum, thereby further ensuring the accuracy of the bacterial moisture content detection.

[0055] In step S200 , a plurality of reflected light spectra with known water contents may be acquired and partial least squares regression may be used to establish the spectrum-water content mapping model.

[0056] In addition, the classification of the bacterial cells should not only consider the moisture content, but also the presence of defects on the bacterial cells, such as mold, cavities, or hollows. Different defects on the bacterial cells will result in different reflectivities and different levels of intensity abnormality in the corresponding defect monitoring bands. The method may further include: step S300, determining the defect of the bacterial cell based on the level of intensity abnormality in the corresponding defect monitoring band in the reflected light spectrum.

[0057] Before step S300, the method may further include: performing deep learning training on the reflected light spectra of healthy bacteria and moldy bacteria, and extracting the spectral features of the moldy bacteria using a convolutional neural network; step S300 specifically includes: based on the spectral features of the moldy bacteria, determining whether the bacteria is moldy according to the abnormal degree of intensity of the corresponding defect monitoring band in the reflected light spectrum.

[0058] Research has found that moldy bacteria will have an abnormal absorption peak in the 1650nm band (NH bond vibration of aflatoxin). The hollow or hollow bacteria will scatter light multiple times inside, resulting in a significant increase in its reflectivity compared to the reflectivity of healthy bacteria. That is, compared to healthy bacteria, the intensity of the reflected light spectrum corresponding to the hollow or hollow bacteria will be significantly increased, and compared to moldy bacteria, the increase in the intensity of the reflected light spectrum corresponding to the hollow or hollow bacteria is higher. Therefore, the defect monitoring band can include the 1650nm band and all bands. The first intensity abnormality that appears in all bands is much greater than the second intensity abnormality that appears in the 1650nm band. When the first intensity abnormality appears in all bands, it indicates that there are cavities or hollows on the bacteria. When the second intensity abnormality appears in the 1650nm band, it indicates that the bacteria is moldy. The present invention can determine the defects of the bacteria based on the intensity abnormality of the corresponding defect monitoring band in the reflected light spectrum; the present invention establishes different defect monitoring bands according to the characteristics of the bacteria defects for different bacteria defects, and determines the defects of the bacteria based on the different intensity abnormalities of the reflected light spectrum in the corresponding defect monitoring band, which can ensure the accuracy of the determination of the bacteria defects.

[0059] As can be seen from the above embodiments, the present invention irradiates the bacteria with a light source including multiple spectral components of different wavelengths, and determines the water content of the bacteria based on the spectrum of reflected light with multiple different wavelengths returned by the bacteria, that is, the water content of the bacteria is determined from two dimensions of wavelength and spectral intensity, which can realize the measurement of the water content of the bacteria and ensure the measurement accuracy; the present invention establishes a spectrum-water content mapping model, and based on the spectrum-water content mapping model, determines the water content of the bacteria according to the intensity characteristics of the sensitive bands in the reflected light spectrum, which can improve the speed of bacteria water content detection.

[0060] See also Figure 2 , is a flow chart of an embodiment of the bacterial cell classification method of the present invention. The bacterial cell classification method may include:

[0061] Step S10: Obtain the volume of the bacterial body based on the collected three-dimensional data of the bacterial body.

[0062] In this embodiment, a binocular camera and a structured light projector can be used to collect 3D data of the bacteria. The binocular camera can be mounted on a 30° angled bracket to the side of the bacteria conveyor belt, with a working distance of 500mm. The structured light projector can be installed 400mm above the bacteria conveyor belt. After the bacteria are conveyed to the designated location, the binocular camera and structured light projector begin to work, scanning the bacteria's 3D outline and obtaining data on its major axis, minor axis, and volume.

[0063] Step S20: using the above-mentioned bacterial cell water content detection method to detect the water content of the bacterial cell.

[0064] Step S30: Correct the volume according to the detected water content.

[0065] During the drying process of the bacteria, water loss will cause the cell structure to shrink and the volume to decrease significantly (shrinkage rate is about 10-20%). If the grading is directly based on the volume after drying, problems such as misjudgment and distortion of quality correlation may occur. Among them, the risk of misjudgment means that in the same batch, bacteria with different dryness may be incorrectly graded due to different shrinkage degrees (such as high-dryness bacteria are downgraded due to their small volume). Weakened quality correlation means that the volume is originally positively correlated with the maturity and thickness of the bacteria, but the volume after shrinkage cannot directly reflect the original quality. In addition, existing grading standards (such as cap length) are usually formulated based on the bacteria after drying, but if the drying process fluctuates during processing (such as differences in water content between different batches), the volume parameters need to be dynamically calibrated to maintain standard consistency. To this end, the present invention proposes to correct the volume of the dried bacteria according to the detected water content.

[0066] Before step S30, the present invention may further include establishing a shrinkage model in the following manner:

[0067] Measure the volume of multiple fresh bacterial samples;

[0068] Drying the fresh bacterial cell sample according to the set drying process to obtain a dried bacterial cell sample;

[0069] The volume and moisture content of the dried bacterial sample were measured, and the shrinkage rate was determined for each dried bacterial sample with different moisture content. The shrinkage rate was equal to the volume of the dried bacterial sample divided by the volume of the fresh bacterial sample before drying. The shrinkage rate was correlated with the moisture content of the dried bacterial sample to obtain the shrinkage rate model.

[0070] The step S30 may specifically include: according to the detected water content, finding the corresponding shrinkage rate from the shrinkage rate model, and correcting the volume according to the following formula to obtain the corrected volume: Corrected volume = the volume / (1-shrinkage rate). For example, the measured volume of the dried bacteria is 2.5 cm 3 , the water content is 8%. From the shrinkage rate model, the shrinkage rate corresponding to the water content is found to be 15%. The corrected volume is 2.5 / (1-0.15)≈2.94cm 3. The present invention corrects the volume of the dried bacteria according to the moisture content of the dried bacteria, which can improve grading accuracy, enhance processing consistency, and increase added value. Improving grading accuracy means correcting the influence of drying shrinkage and restoring the volume of the bacteria in the "theoretical drying state" (such as uniformly calculating at a moisture content of 12%) to ensure that the grading results truly reflect their original quality; enhancing processing consistency means adapting to fluctuations in different drying processes (such as a higher natural drying shrinkage rate) to ensure that the same grading standard applies to all batches and reduce manual intervention; improving added value means that the corrected volume is more correlated with nutritional components (such as polysaccharide content), supporting the precise sorting of high-value products and increasing the selling price.

[0071] Step S40: Classify the bacteria according to the detected water content and the corrected volume.

[0072] As can be seen from the above embodiments, the present invention relies on multiple dimensions to grade the bacteria based on the moisture content and volume of the dried bacteria, which makes the grading results of the bacteria more in line with the actual value requirements and realizes the "size-quality" two-dimensional grading; and before grading the bacteria, the present invention corrects the volume of the dried bacteria, which can improve the grading accuracy, enhance the processing consistency, and increase the added value.

[0073] In addition, before step S40, the bacterial grading method may further include: detecting defects in the bacterial body using the above-mentioned bacterial body moisture content detection method; and measuring the weight of the bacterial body. Step S40 may specifically include: grading the bacterial body based on the detected moisture content, the corrected volume, the bacterial body defects, and the weight. The present invention not only grades bacterial bodies based on the moisture content and corrected volume of the bacterial body, but also based on the bacterial body defects and weight. This allows identification of irregular bacterial bodies (e.g., bacterial bodies with damaged caps), and grading of bacterial bodies from multiple dimensions, thereby further improving the accuracy of bacterial grading. A piezoelectric sensor may be used to measure the weight of the bacterial body. For example, if the weight exceeds a preset threshold (e.g., <5g or >20g), the bacterial body is marked as defective.

[0074] Prior to step S40, the bacterial grading method may further include extracting the long axis, short axis, and surface wrinkling of the bacterial cells from the collected three-dimensional bacterial cell data. Step S40 may specifically include grading the bacterial cells based on the detected moisture content, corrected volume, bacterial defects, weight, long axis, short axis, and surface wrinkling. Prior to step S40, a grading model may be established. Multiple parameters are input into the grading model when grading the bacterial cells, and a comprehensive grade (special grade / first grade / second grade / defective) is output. Dynamic thresholds may also be set to adjust sorting parameters in real time based on user-preset standards (e.g., export grade: long axis ≥ 3 cm, moisture ≤ 10%, no mold).

[0075] See also Figure 3 , is a schematic structural diagram of an embodiment of a bacterial cell grading device according to the present invention. The bacterial cell grading device may include a controller, a loading mechanism connected to the controller, a conveyor belt, a moisture content detection mechanism, a three-dimensional data acquisition mechanism, a weight sensor, and a sorting mechanism. The controller controls the loading mechanism to sequentially transfer bacterial cells to the conveyor belt, which may be a transparent and light-transmitting conveyor belt. The three-dimensional data acquisition mechanism above the conveyor belt is controlled to collect three-dimensional data of the bacterial cells, the moisture content detection mechanism below the conveyor belt is controlled to detect the moisture content of the bacterial cells, and the weight sensor is controlled to detect the weight of the bacterial cells. The controller obtains the volume of the bacterial cells based on the collected three-dimensional data, corrects the volume based on the detected moisture content, and grades the bacterial cells based on the detected moisture content and the corrected volume. The sorting mechanism is then controlled based on the grading results to sort the bacterial cells into collection bins of corresponding grades. The sorting mechanism may be a two-degree-of-freedom servo gimbal, which may be located below the unloading guide rail of the conveyor belt. The present invention controls the feeding, transmission, three-dimensional data collection, and water content detection of the bacteria by the controller. After the bacteria are graded according to the collected three-dimensional data and water content, the controller controls the sorting mechanism to sort the bacteria into the collection frame of the corresponding level, thereby realizing the automatic grading and sorting of the bacteria. In this embodiment, the feeding mechanism can be a fully automatic vibrating plate feeder, combined with Figures 4 to 8 As shown, the fully automatic vibrating plate feeder may include a hopper 1, a movable frame 2, a spring sheet 3, an electromagnet 4, an outer cover 5 and a base 6. The hopper 1 may be a spiral trough, the movable frame 2 is fixedly connected to the hopper 1, the outer cover 5 is covered on the base 6, the electromagnet 4 is fixed on the base 6 and is located inside the outer cover 5, and a plurality of first fixed blocks 61 located inside the outer cover 5 are also fixed on the base 6. A plurality of second fixed blocks 11 are fixed to the bottom side of the hopper 1, and each first fixed block 61 cooperates with the corresponding second fixed block 11 to fix the corresponding spring sheet 3. Each spring sheet 3 is tilted around the vertical axis of the hopper 1 at the same angle, wherein the tilted state of each spring sheet 3 is formed after its bottom end is tangent to the same circle, and the tangent point is used as the rotation point, and it is rotated outward by a set angle, and the center of the circle is on the vertical axis of the hopper 1. When the electromagnet 4 is powered off, each spring piece 3 is in a natural state. Under the support of each spring piece 3, part of the movable frame 2 is located outside the outer cover 5, and the hopper 1 is spaced apart from the outer cover 5; when the electromagnet 4 is powered on, the movable frame 2 moves toward the direction of the electromagnet 4 under the electromagnetic force of the electromagnet 4, and each spring piece 3 gradually enters a compressed state, thereby ensuring that the hopper 1 moves downward smoothly; when the electromagnet 4 is powered off again, each spring piece 3 recovers from the compressed state to the natural state. During the recovery process, the bacteria in the hopper 1 move from the inside to the outside of the hopper 1, and also move toward the discharge port of the hopper 1, thereby realizing the directional spiral motion of the bacteria.

[0076] The feeding mechanism relies on a pulsed electromagnet under the hopper, which continuously switches on and off under the action of an electric current, causing the hopper to vibrate vertically. The tilt of the spring sheet connected to the hopper causes the hopper to torsional swing in one direction. This vibration causes the material inside the hopper to continuously rise along the spiral trough. During this process, the material constantly adjusts its position and evenly disperses, being transported along a directional track. During operation, bacterial cells are poured into the hopper and transported in a directional manner via a vibrating plate. A timer-operated baffle can be installed at the discharge port to control the distance between the cells, making detection more accurate.

[0077] The movable frame is the base of the vibration system and is usually fixedly connected to the electromagnet. It is responsible for transmitting the driving force generated by the electromagnet to the hopper. The movable frame is easy to disassemble and maintain, and can adapt to hoppers of different specifications and vibration requirements. The spring leaf connects the movable frame to the base, storing and releasing energy through elastic deformation, forming a directional vibration trajectory (such as spiral motion). The spring leaf can absorb high-frequency vibration impacts and reduce interference with the base and surrounding equipment. It also has adjustability. The vibration amplitude or frequency can be changed by adjusting the number or stiffness of the spring leaf. The electromagnet is the power source of the vibration system. It generates an alternating magnetic field by periodically turning the power on and off, driving the movable frame and hopper to vibrate. The electromagnet can achieve rapid start and stop of dozens to hundreds of times per second, which is suitable for high-speed feeding requirements. It can be precisely controlled and the vibration amplitude and speed can be flexibly controlled by adjusting the current size or pulse frequency. In addition, the feeding mechanism can also include a base plate 7 and a vibration-damping rubber foot pad 8. The bottom side of the base 6 can be fixed to the base plate 7 via the vibration-damping rubber foot pad 8. The feeding mechanism of the present invention can realize automatic feeding and has a simple structure.

[0078] See also Figure 9 , is a schematic structural diagram of an embodiment of a bacterial cell moisture content detection mechanism of the present invention. The bacterial cell moisture content detection mechanism may include a processor, a light source generator, a probe, and a spectrometer connected to the processor. The probe is connected to the spectrometer. The light source generator irradiates a light source comprising multiple spectral components of different wavelengths onto a unit area of the bacterial cell. The probe detects the reflected light reflected back from the bacterial cell and transmits the reflected light to the spectrometer, which generates a reflected light spectrum and transmits the reflected light spectrum to the processor. The processor determines the moisture content of the bacterial cell based on the intensity characteristics of the sensitive band in the reflected light spectrum based on a spectrum-water content mapping model. In this spectrum-water content mapping model, the intensity of a single wavelength of reflected light within the sensitive band of the spectrum varies with the moisture content of the bacterial cell. When the moisture content of the bacterial cell is constant, the intensity of reflected light of different wavelengths within the sensitive band of the spectrum varies, and the difference in the intensity of reflected light of different wavelengths is related to the moisture content of the bacterial cell and the difference in the wavelength of the reflected light. It should be noted that the bacterial cell mentioned above may be a morel.

[0079] Other embodiments of the present invention will readily occur to those skilled in the art after considering the specification and practicing the invention disclosed herein. This application is intended to cover any variations, uses, or adaptations of the present invention that follow the general principles of the invention and include common knowledge or customary techniques in the art not disclosed herein. The description and examples are to be considered as exemplary only, with the true scope and spirit of the invention being indicated by the following claims.

[0080] It will be appreciated that the present invention is not limited to the precise construction that has been described above and shown in the accompanying drawings, and that various modifications and variations can be made without departing from its scope, which is governed solely by the appended claims.

Claims

1. A method for detecting water content of bacterial cells, characterized in that: include: Step S100: irradiating a unit area of a bacterial cell with a light source having multiple spectral components of different wavelengths, so that the bacterial cell transmits reflected light back to a probe, which transmits the reflected light to a spectrometer to generate a reflected light spectrum; Step S200: Determine the water content of the bacteria based on the intensity characteristics of the sensitive band in the reflected light spectrum based on a spectrum-water content mapping model. In the spectrum-water content mapping model, the intensity of the reflected light of a single wavelength in the sensitive band of the spectrum varies when the water content of the bacteria is different; when the water content of the bacteria is constant, the intensity of the reflected light of different wavelengths in the sensitive band of the spectrum varies, and the difference in the intensity of the reflected light of different wavelengths is related to the water content of the bacteria and the difference in the wavelength of the reflected light.

2. The method for detecting water content of bacterial cells according to claim 1, wherein: The sensitive wavelength range is 1350-1500 nm. After step S200, the method further includes: combining the polysaccharide absorption peak appearing in the 1200 nm wavelength range to cross-validate the determined bacterial water content; In step S100, the probe detects the reflected light returned from each unit area of the bacterial cell and transmits the reflected light from each unit area to the spectrometer to generate a reflected light spectrum corresponding to each unit area. Before step S200, the method further includes: superimposing and averaging the reflected light spectra corresponding to the respective unit area regions to obtain the reflected light spectrum; In step S200 , a plurality of reflected light spectra with known water contents are obtained, and a spectrum-water content mapping model is established using partial least squares regression.

3. The method for detecting water content of bacterial cells according to claim 1, wherein: The defects on the bacterial body are different, the reflectivity of the bacterial body in the corresponding defect monitoring band is different, and the intensity abnormality degree in the corresponding defect monitoring band is different; the method also includes: step S300, determining the defects of the bacterial body according to the intensity abnormality degree of the corresponding defect monitoring band in the reflected light spectrum.

4. The method for detecting water content of bacterial cells according to claim 3, wherein: Before step S300, the method further includes: performing deep learning training on the reflected light spectra of healthy bacteria and moldy bacteria, and extracting spectral features of the moldy bacteria using a convolutional neural network; The step S300 specifically includes: based on the spectral characteristics of the moldy bacteria, determining whether the bacteria are moldy according to the abnormality of the intensity of the corresponding defect monitoring band in the reflected light spectrum.

5. The method for detecting water content of bacterial cells according to claim 3 or 4, wherein: The defect monitoring band includes the 1650nm band and all bands. The first intensity abnormality appearing in all bands is much greater than the second intensity abnormality appearing in the 1650nm band. When the first intensity abnormality appears in all bands, it indicates that there are holes or hollows on the bacterial body. When the second intensity abnormality appears in the 1650nm band, it indicates that there is mold on the bacterial body.

6. A bacterial cell classification method, characterized in that: include: Step S10: Obtaining the volume of the bacterial body based on the collected three-dimensional data of the bacterial body; Step S20: Detecting the water content of the bacteria using the method for detecting water content of the bacteria according to any one of claims 1 to 5; Step S30, correcting the volume according to the detected water content; Step S40: Classify the bacteria according to the detected water content and the corrected volume.

7. The bacterial cell classification method according to claim 6, characterized in that: Before step S30, the method further includes establishing a shrinkage rate model in the following manner: Measure the volume of multiple fresh bacterial samples; Drying the fresh bacterial cell sample according to the set drying process to obtain a dried bacterial cell sample; The volume and moisture content of the dried bacterial sample were measured, and the shrinkage rate was determined for each dried bacterial sample with different moisture content. The shrinkage rate was equal to the volume of the dried bacterial sample divided by the volume of the fresh bacterial sample before drying. The shrinkage rate was correlated with the moisture content of the dried bacterial sample to obtain the shrinkage rate model.

8. The bacterial cell classification method according to claim 7, characterized in that: The step S30 specifically includes: according to the detected water content, finding the corresponding shrinkage rate from the shrinkage rate model, and correcting the volume according to the following formula to obtain the corrected volume: corrected volume = the volume / (1-shrinkage rate).

9. The bacterial cell classification method according to claim 6, characterized in that: Before step S40, the method further includes: detecting defects of the bacterial cells using the bacterial cell moisture content detection method according to any one of claims 1 to 5; and measuring the weight of the bacterial cells; The step S40 specifically includes: grading the bacteria according to the detected water content, the corrected volume, the defects of the bacteria and the weight.

10. The bacterial cell classification method according to claim 6, characterized in that: The controller controls the feeding mechanism to sequentially transfer the bacterial cells to the conveyor belt, controls the three-dimensional data acquisition mechanism above the conveyor belt to acquire three-dimensional data of the bacterial cells, and controls the moisture content detection mechanism below the conveyor belt to detect the moisture content of the bacterial cells. The controller obtains the volume of the bacteria based on the collected three-dimensional data of the bacteria, corrects the volume according to the detected water content, and grades the bacteria based on the detected water content and the corrected volume. Thereafter, the controller controls the sorting mechanism according to the grading results to sort the bacteria into the collection frame of the corresponding level.

Images