Liquid dispensing method, apparatus, device and medium

Abstract

This invention relates to the field of biomedical technology and discloses a liquid dispensing method: An initialization operation is performed to purge gas from tubing and containers via an injection container; the target solution is transferred from the product container to the injection container; after dispensing the target solution into dispensing containers via a control valve through the injection container, bubble recognition is performed on each dispensing container using a bubble visual recognition model to obtain bubble recognition results; based on all bubble recognition results, bubble extraction is performed on the dispensing containers, and the dispensing containers are heat-sealed to obtain sealed dispensing containers. This invention, through the initialization operation, injection container, and control valve, achieves the removal of gas from tubing and containers, realizes the dispensing of the target solution, and achieves precise control of dispensing volume. The bubble visual recognition model avoids cell damage caused by bubbles in the dispensing containers, thereby ensuring the consistency of dispensing volume and improving dispensing efficiency.

Description

Technical Field

[0001] This invention relates to the field of biomedical technology, and in particular to a liquid dispensing method, apparatus, equipment, and medium. Background Technology

[0002] Cell therapy is a key area of ​​development at the forefront of international medicine, and liquid dispensing technology is one of the crucial steps in the process of cell therapy.

[0003] Currently, liquid dispensing primarily relies on manual operation under aseptic conditions. This manual process demands high skill from operators, leading to a high error rate. Furthermore, dispensing liquid into multiple containers makes it difficult to ensure consistent volume across all containers, resulting in low dispensing efficiency. Therefore, a liquid dispensing technology is urgently needed to address these issues. Summary of the Invention

[0004] This invention provides a liquid dispensing method, apparatus, equipment, and medium to solve the problems in the prior art that cannot guarantee the consistency of volume in each container and that the high error rate in multiple dispensing operations leads to low dispensing efficiency.

[0005] A liquid dispensing method, comprising: Perform initialization to purge gas from the delivery lines and all dispensing containers through the injection container; The target solution in the product container is transferred to the injection container. After the target solution is dispensed into the dispensing container through the injection container via the control valve, the bubble visual recognition model is used to identify bubbles in each of the dispensing containers to obtain bubble recognition results that correspond one-to-one with each of the dispensing containers. Based on all the bubble identification results, bubble extraction is performed on each of the sub-containers, and each of the sub-containers is heat-sealed to obtain the sealed sub-containers.

[0006] A liquid dispensing device, comprising: The gas discharge module is used to perform initialization operations to discharge gas from the delivery line and all dispensing containers through the injection container; The liquid dispensing module is used to transfer the target solution in the product container to the injection container. After the target solution is dispensed into the dispensing container through the injection container via a control valve, a bubble visual recognition model is used to identify bubbles in each dispensing container to obtain bubble recognition results that correspond one-to-one with each dispensing container. The bubble extraction module is used to extract bubbles from each of the sub-containers based on all the bubble identification results, and to heat-seal each of the sub-containers to obtain the sealed sub-containers.

[0007] A liquid dispensing device includes a controller, a product container, a dispensing container, an injection container, and a control valve, wherein the controller is used to execute the liquid dispensing method.

[0008] A computer-readable storage medium storing a computer program that, when executed by a processor, implements the above-described liquid dispensing method.

[0009] The liquid dispensing method, apparatus, equipment, and medium described above, in this invention's liquid dispensing method, achieves the purging of gas in the delivery pipeline and all dispensing containers by performing an initialization operation. Through the injection container and control valve, the target solution is dispensed, and precise control of the dispensing volume is achieved, ensuring the accuracy of the liquid volume in the dispensing containers. Through a bubble visual recognition model and bubble extraction processing, bubbles in the dispensing containers are detected, ensuring that no bubbles are present in the dispensing containers and preventing gas from damaging the cells in the target solution during freezing. Furthermore, it reduces manual intervention, lowers labor costs and dispensing error rates, and ensures the consistency of dispensing volume, thereby improving dispensing efficiency. Attached Figure Description

[0010] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the description of the embodiments of the present invention will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0011] Figure 1 This is a flowchart of a liquid dispensing method in one embodiment of the present invention; Figure 2 This is a schematic diagram of a liquid dispensing device according to an embodiment of the present invention; Figure 3 This is a connection diagram of a liquid dispensing device in one embodiment of the present invention; The reference numerals in the accompanying drawings are as follows: 1. Product container; 2. Product control valve; 3. Dispensing control valve; 4. Injection container; 5. Aseptic sleeve; 6. Motor; 7. Dispensing container; 8. Rotary three-way valve. Detailed Implementation

[0012] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0013] In one embodiment, such as Figure 1 and Figure 3 As shown, a liquid dispensing method is provided, comprising the following steps: S10: Perform initialization to purge gas from the delivery line and all dispensing containers through the injection container.

[0014] In essence, delivery tubing refers to the tubing connecting the injection container and various dispensing containers, i.e., used to deliver the solution. Dispensing containers are containers used to store the solution, such as product bags or cryopreservation bags. Injection containers are containers used to dispense the solution, such as syringes or injection pumps.

[0015] Specifically, an initialization operation is performed to discharge gas from the delivery pipeline and all dispensing containers 7 through injection container 4. This involves closing product control valve 2 connecting product container 1 and injection container 4, and opening dispensing control valve 3 connecting dispensing containers 7 and injection container 4. Then, motor 6 pulls (moves downward) the push rod in injection container 4 to extract gas from the delivery pipeline and all dispensing containers 7 into injection container 4, and the dispensing control valve connecting dispensing containers 7 and injection container 4 is closed. Next, product control valve 2 connecting product container 1 and injection container 4 is opened to discharge gas from injection container 4 into product container 1. The initialization operation can be repeated to ensure complete discharge of gas from the delivery pipeline and all dispensing containers 7. For example, a set number of initialization operations can be set; when this number is reached, it is determined that the gas in the delivery pipeline and all dispensing containers 7 has been emptied.

[0016] S20: The target solution in the product container is transferred to the injection container. After the target solution is dispensed into the dispensing container through the injection container via the control valve, bubble recognition is performed on each of the dispensing containers using a bubble visual recognition model to obtain bubble recognition results corresponding to each of the dispensing containers.

[0017] Understandably, the bubble identification result is used to characterize whether bubbles exist in the dispensing container, corresponding to the presence of bubbles in the dispensing container in steps S204 and S205. The product container refers to a container used for storing solutions, such as a product bag, centrifuge cup, centrifuge jar, or centrifuge bottle. The target solution refers to the liquid that needs to be dispensed and frozen into the dispensing container, such as immune cell fluid. The control valve is a valve used to control the opening of the pipeline. The control valve can be in contact with the periphery of the pipeline, but not embedded, thus avoiding direct contact and susceptibility to external contamination.

[0018] Specifically, the target solution in product container 1 is transferred to injection container 4. This is achieved by opening product control valve 2, which connects product container 1 and injection container 4, and injecting the target solution from product container 1 into injection container 4. This is done by motor 6 pulling down the push rod of injection container 4 to a preset position, which indicates the volume of target solution to be extracted from product container 1. The target solution is extracted from product container 1 as needed; for example, if there are 10 dispensing containers, 12 ml is extracted. If the volume of target solution extracted in one operation is insufficient to fill all dispensing containers 7, the target solution in product container 1 is extracted again to complete the dispensing process for all dispensing containers 7. In another embodiment, the push rod of the injection container 4 is first pulled downwards by the motor 6 to move it to a preset position. Then, the product control valve 2, which connects the product container 1 and the injection container 4, is opened to inject the target solution from the product container 1 into the injection container 4. When the target solution in the injection container 4 reaches a preset level, the product control valve 2 is closed, and the volume of the target solution in the injection container 4 is the required volume when the liquid in the pipeline flows to the injection container 4. When the pipeline between the product control valve 2 and the injection container 4 is short, the volume of the target solution remaining in the pipeline can be ignored.

[0019] Furthermore, the dispensing control valve 3 connecting the dispensing container 7 and the injection container 4 is opened, as well as the control valve corresponding to the dispensing container 7. Then, the pusher rod in the injection container 4 is pushed upward by the motor 6, allowing the target solution to enter the dispensing container 7 through the delivery pipeline and the control valve corresponding to the dispensing container 7, thereby completing the delivery of the preset volume of target solution into the dispensing container 7. To ensure that the volume entering the dispensing container 7 is the preset volume, a weighing sensor can be set to weigh the dispensing container 7 to ensure that the volume of the target solution is the preset volume.

[0020] Furthermore, bubble recognition is performed on each dispensing container 7 using a bubble visual recognition model. Specifically, after dispensing the target solution of a preset volume into the dispensing container 7, an image of the dispensing container 7 is acquired and input into the bubble visual recognition model. The model then performs bubble recognition on the image of the dispensing container 7, using the recognition capabilities learned during training to detect the presence of arc-shaped boundary features in the image of the dispensing container 7. This yields a bubble recognition result corresponding to each dispensing container 7. Here, arc-shaped boundary features refer to all gray values ​​in the first acquired image that differ from surrounding gray values ​​appearing as arcs or circles. In one embodiment, bubble recognition can be performed on each dispensing container 7 individually after all dispensing is completed.

[0021] S30: Based on all the bubble identification results, bubble extraction is performed on each of the sub-containers, and each of the sub-containers is heat-sealed to obtain the sealed sub-containers.

[0022] Understandably, a packaged refill container refers to a refill container that holds a solution.

[0023] Specifically, based on all bubble recognition results, bubble extraction processing is performed on each dispensing container 7. That is, if the bubble recognition result corresponding to the dispensing container 7 indicates the presence of an arc-shaped boundary feature corresponding to a bubble in the image of the dispensing container 7, then bubble extraction processing is performed on the dispensing container 7. This involves using a syringe 4 to extract the target solution from the dispensing container 7 to remove the bubbles. Then, it is determined again whether bubbles exist in the dispensing container 7. If not, it is determined whether the liquid level in the dispensing pipeline of the dispensing container 7 meets the preset requirements (e.g., a liquid level requirement in the dispensing pipeline). If it meets the preset requirements, the control valve corresponding to the dispensing container 7 is closed. If bubbles are again detected in the dispensing container, the target solution in the dispensing container 7 is repeatedly extracted using the syringe 4 to remove air bubbles and ensure that the liquid level in the dispensing pipeline of the dispensing container 7 meets the preset requirements. Next, the packaging container 7 is sealed using a heat-sealing device. This involves sealing the connection between the packaging pipeline and the delivery pipeline of the packaging container 7, resulting in the sealed packaging container 7. This heat-sealing device can be integrated into a control valve, with the valve port having a heat-sealing structure for integration. Alternatively, after all packaging containers 7 have been packaged, all containers 7 can be subjected to bubble extraction and sealing simultaneously to obtain the sealed packaging container 7. Or, the sealing can be performed uniformly after all packaging containers 7 have undergone bubble extraction.

[0024] In the liquid dispensing device of this invention, an initialization operation is performed to purge gas from the delivery pipeline and all dispensing containers. The injection container and control valve enable the dispensing of the target solution and precise control of the dispensing volume, ensuring the accuracy of the liquid volume in the dispensing containers. A bubble visual recognition model and bubble extraction processing are used to detect bubbles in the dispensing containers, ensuring that no bubbles are present and preventing gas from damaging cells in the target solution during freezing. Furthermore, manual intervention is reduced, lowering labor costs and dispensing error rates, while ensuring consistency in dispensing volume, thereby improving dispensing efficiency.

[0025] In one embodiment, step S10, namely performing an initialization operation to discharge gas from the delivery pipeline and all dispensing containers through the injection container, includes: S101, draw gas from the delivery pipeline and all the dispensing containers into the injection container.

[0026] S102, close the dispensing control valve connected to the dispensing container, and discharge the gas in the injection container to the outside or the product container.

[0027] Understandably, the dispensing control valve is used to connect the injection container and all dispensing containers.

[0028] Specifically, the gas is drawn from the delivery pipeline and all dispensing containers 7 into the injection container 4. That is, the dispensing control valve 3, which connects the injection container 4 to all dispensing containers 7, is first opened. Then, the motor 6 is controlled to move the push rod downwards in the injection container 4, thereby drawing gas from the delivery pipeline and all dispensing containers 7. When the push rod reaches the bottom of the injection container 4, i.e., when the push rod has moved a preset distance, the dispensing control valve 3 is closed, and the product control valve 1 is opened, then the gas in the injection container 4 is pushed into the product container 1. In another embodiment, after closing the dispensing control valve 3, the exhaust valve located in the pipeline is opened. Figure 3 (Not shown) The gas in injection container 4 is discharged to the outside. The number of extractions can be set to completely empty the gas from the delivery pipeline and all dispensing containers 7.

[0029] In another embodiment, by first opening the dispensing control valve 3, and then by using the peristaltic pump ( Figure 3 (Not marked) Extract the gas from the delivery pipeline and all dispensing containers 7. After the peristaltic pump has been running for a preset time, turn off the peristaltic pump and confirm that the gas in the delivery pipeline and all dispensing containers 7 has been completely emptied.

[0030] In this embodiment, by drawing gas into the injection container, the gas in the delivery pipeline and all the dispensing containers is emptied, thereby releasing the gas to the outside or into the product container, and preventing the gas from re-entering the management and dispensing containers during subsequent dispensing.

[0031] In one embodiment, the control valve is a rotary three-way valve, and one rotary three-way valve corresponds to one dispensing container; In step S20, that is, dispensing the target solution into a dispensing container via the injection container and a control valve, the following steps are included: S201, Open the rotary three-way valve corresponding to the dispensing container, and extract the target volume of the target solution from the injection container and deliver it to the dispensing container; the target volume includes the residual volume in the delivery pipeline and the preset volume in the dispensing container.

[0032] Understandably, the target volume includes the residual volume in the delivery tubing and the preset volume in the dispensing containers. The residual volume refers to the capacity of the delivery tubing between the injection container and each dispensing container. The preset volume refers to the volume to be dispensed into each dispensing container in advance, for example, 1 ml or 5 ml. A rotary three-way valve is a valve device that connects three channels, and the channels are opened and closed by rotation.

[0033] Specifically, after transferring the target solution from product container 1 to injection container 4, the product control valve 2 connecting product container 1 and injection container 4 is closed, and the dispensing control valve 3 connecting injection container 4 and all dispensing containers 7 is opened, along with the rotary three-way valve 8 corresponding to the dispensing container. Then, the target volume of target solution is extracted from injection container 4 and delivered to dispensing container 7. Specifically, the control motor 6 pushes the push rod in injection container 4 upwards, causing the target solution in injection container 4 to flow into dispensing container 7 via the delivery pipeline. The volume of the target solution is controlled by controlling the distance the push rod moves, ensuring that the target volume of target solution enters the delivery pipeline and is injected into dispensing container 7. The volume of solution in dispensing container 7 is monitored by a weighing sensor to ensure that the preset volume of target solution enters dispensing container 7. After the preset volume of target solution enters dispensing container 7, the corresponding rotary three-way valve 8 is rotated to close dispensing container 7, preventing further flow of target solution into the dispensing container 7 corresponding to the rotary three-way valve 8. To prevent air bubbles from affecting the volume entering the dispensing container 7, the change in the weighing sensor can be monitored while the target solution is pushed by the pusher. When the weighing sensor reaches the weight of the preset volume, the pushing of the target solution is stopped.

[0034] In another embodiment, when multiple dispensing containers exist, when the rotary three-way valve 8 corresponding to the first dispensing container 7 is closed, the rotary three-way valve 8 corresponding to the second dispensing container 7 is opened. The solution volume in the tubing between the two rotary three-way valves 8 and the preset volume corresponding to the dispensing container 7 (i.e., the volume of the target solution remaining in the tubing is the capacity of the tubing between the injection container and the first dispensing container) are pushed out from the injection container 4. The preset volume of target solution is then injected into the second dispensing container. The weight of the target solution in the second dispensing container is monitored by a weighing sensor. When the weight corresponding to the preset volume is reached, the corresponding rotary three-way valve 8 is rotated to close the dispensing container 7, preventing the target solution from flowing into the dispensing container 7 corresponding to that rotary three-way valve 8. If the dispensing container is the target container, the residual target solution in the delivery tubing is pushed into the target container by a peristaltic pump. The target container refers to the last dispensing container 7, which can be a quality inspection bag.

[0035] In this embodiment, the target solution is dispensed by rotating a three-way valve, thereby achieving precise control of the dispensing volume, ensuring the accuracy of the target solution volume in the dispensing container, and ensuring the consistency of the dispensing volume in all dispensing containers.

[0036] In one embodiment, step S20, namely, performing bubble recognition on each of the dispensing containers using a bubble visual recognition model to obtain bubble recognition results corresponding one-to-one with each of the dispensing containers, includes: S202, Acquire the first image corresponding to the dispensing container.

[0037] S203, bubble recognition is performed on the first acquired image using the bubble visual recognition model to obtain bubble recognition results.

[0038] S204, when the bubble recognition result indicates that there are arc-shaped boundary features in the first acquired image, it is determined that the dispensing container contains bubbles.

[0039] S205, when the bubble recognition result indicates that there are no arc-shaped boundary features in the first acquired image, it is determined that the dispensing container does not contain bubbles.

[0040] Understandably, the first acquired image refers to an image of one or more dispensing containers currently specified or in a state to be acquired, and their surroundings. The surrounding images include the dispensing pipelines corresponding to the dispensing containers. "Currently specified" means specified / selected during the current operation / program execution. "Dispensing containers in a state to be acquired" refers to dispensing containers marked as pending processing during program execution or execution. The first acquired image can be a single image or multiple images (image set). The bubble recognition result is used to characterize whether there are arc-shaped boundary features in the first acquired image. The arc-shaped boundary features are arc-shaped edge lines resembling bubbles at the inner wall edge of the dispensing container.

[0041] Specifically, after the packaging container 7 is filled, an image of the packaging container 7 is captured using a camera positioned next to it, i.e., a first captured image corresponding to the packaging container 7 is acquired. Then, a bubble visual recognition model is obtained, and the first captured image is input into the bubble visual recognition model. The bubble visual recognition model performs bubble recognition on the first captured image; that is, it uses the bubble recognition capability learned during training to perform region recognition on the first captured image. Specifically, it first identifies the boundary of the container image corresponding to each packaging container 7, divides the container image of each packaging container 7 according to the identified boundary, extracts features from each container image, and classifies the extracted features to obtain the bubble recognition result corresponding to the packaging container 7. When the bubble recognition result indicates the presence of arc-shaped boundary features in the first captured image, it is determined that the packaging container 7 contains bubbles. When the bubble recognition result indicates the absence of bubbles with arc-shaped boundary features in the first captured image, it is determined that the packaging container 7 does not contain bubbles.

[0042] In this embodiment, a bubble visual recognition model is used to identify bubbles in the first acquired image, thereby enabling the identification of whether the image contains arc-shaped boundary features. This allows for the detection of whether bubbles exist in the dispensing container, preventing cell damage during cryopreservation due to bubbles in the dispensing container.

[0043] In one embodiment, step S203, namely, performing bubble recognition on the first acquired image using the bubble visual recognition model to obtain a bubble recognition result, includes: S2031, the bubble visual recognition model is used to perform region recognition on the first acquired image to obtain a container image corresponding to each of the dispensing containers.

[0044] S2032, the container image is feature extracted using the bubble visual recognition model to obtain image features corresponding to each container image.

[0045] S2033, The image features are classified and identified using the bubble visual recognition model to obtain the bubble recognition result corresponding to the packaging container.

[0046] Understandably, a container image refers to an image containing only a single packaging container. Image features refer to the features extracted from the arcuate boundaries of bubbles in an image of a packaging container.

[0047] Specifically, after inputting the first acquired image into the bubble visual recognition model, the bubble visual recognition model performs region recognition on the first acquired image, that is, detects whether the first acquired image contains multiple packaging containers 7. If multiple packaging containers 7 exist, the boundary shape of each packaging container 7 is recognized. Then, the images corresponding to each packaging container 7 are divided according to the recognized boundary shape results, thus obtaining the container image corresponding to each packaging container 7. Then, the bubble visual recognition model extracts features from the container images, that is, it extracts features from each container image using the feature extraction capabilities learned during training, thereby obtaining the image features corresponding to each container image. Then, the bubble visual recognition model classifies the image features, that is, it classifies each image feature using the classification capabilities learned during training. That is, if the image has arc-shaped boundary features, the classification result is determined to be that there are bubbles in the packaging container 7; or if the image does not have arc-shaped boundary features, the classification result is determined to be that there are no bubbles in the packaging container 7. The classification result corresponding to each packaging container 7 is determined as the bubble recognition result corresponding to each packaging container 7.

[0048] In one specific embodiment, when the bubble recognition model is a neural network model based on the YOLO (You Only Look Once, object detection model) network architecture, the bubble visual recognition model first performs region recognition on the first acquired image, that is, divides the container image corresponding to each packaging container. Then, the backbone network in the bubble visual recognition model extracts features from each container image, that is, it extracts features of different scales of the container image through the backbone network, thereby obtaining multi-scale features. Then, the neck network in the bubble visual recognition model performs feature fusion on the multi-scale features, that is, it upsamples features of different scales to make them in the same dimension as the previous scale features and fuses the features, thereby obtaining multiple image features. Finally, the output network in the bubble visual recognition model predicts and classifies all image features to obtain the bubble recognition result corresponding to each packaging container 7.

[0049] In this embodiment, a bubble visual recognition model is used for region identification, which enables the segmentation of the first acquired image and subsequently the identification of the container image of each packaging container. Feature extraction is performed using the bubble visual recognition model to extract image features, thereby enabling the classification of image features and ultimately obtaining the bubble recognition results, thus detecting the presence of bubbles in the packaging containers.

[0050] In one embodiment, step S30, namely, performing bubble extraction processing on each of the dispensing containers based on all the bubble identification results, includes: S301, perform a preset number of liquid extraction operations to remove air bubbles adhering to the inner wall of the dispensing container from the dispensing container.

[0051] S302, perform secondary bubble identification on each of the dispensing containers after the liquid extraction operation until each of the dispensing containers is identified as bubble-free.

[0052] Understandably, surface tension refers to the tension along any boundary line on the surface of a liquid caused by the uneven molecular attraction of the liquid surface layer.

[0053] Specifically, when the bubble identification result indicates the presence of bubbles in the dispensing container 7, due to surface tension, bubbles tend to adhere to the inner wall of the dispensing container 7. Therefore, the liquid extraction operation is repeated. That is, after the first liquid extraction operation is completed, the target solution in the dispensing pipeline of the dispensing container 7 is controlled to flow back to the dispensing container 7, and then the motor 6 is controlled to pull down the push rod in the injection container 4 to extract the solution in the dispensing container 7. The number of times the operation is performed is recorded as two. After repeating the operation a preset number of times, the push rod is stopped, and the rotary three-way valve 8 corresponding to the dispensing container 7 is closed. In this way, the liquid extraction operation is performed on all dispensing containers 7 one by one, so that all the bubbles in the dispensing containers 7 are expelled. Furthermore, a secondary bubble identification process is performed on each dispensing container 7 after the liquid extraction operation. This involves photographing the dispensing container 7 again and re-identifying the image using a bubble visual recognition model to determine whether the container 7 contains bubbles. If the container 7 does not contain bubbles, it is determined that the bubbles adhering to the inner wall of the container 7 due to surface tension will be expelled from the container 7. If bubbles are present in the dispensing container 7, the liquid extraction operation is performed again a preset number of times. After the preset number of extractions, bubble identification is performed on the dispensing container 7 again until it is identified as bubble-free. The secondary bubble identification process can refer to the specific bubble identification process in steps S202 to S205, and will not be elaborated further here.

[0054] In this embodiment, by repeatedly performing the liquid extraction operation and using the bubble visual recognition model, the gas adhering to the inner wall of the dispensing container due to surface tension is discharged from the dispensing container, thus realizing the discharge of all bubbles in the dispensing container and avoiding damage to cells by the gas after freezing.

[0055] In one embodiment, step S302, namely, performing secondary bubble identification on each of the dispensing containers after the extraction operation until each of the dispensing containers is identified as bubble-free, includes: S3021, Acquire a second image of the dispensing container after the liquid extraction operation is performed.

[0056] S3022, Based on the second acquired image, bubble recognition is performed on the dispensing container to obtain the image recognition result.

[0057] S3023, Based on the second acquired image, liquid level identification is performed on the dispensing pipeline connected to the dispensing container to obtain the liquid level identification result.

[0058] S3024, if the image recognition result indicates that there are air bubbles in the dispensing container, or the liquid level recognition result indicates that the position of the solid line of the liquid level in the dispensing pipeline does not meet the preset requirements, then the liquid extraction operation continues and a new second acquisition image is acquired.

[0059] S3025, if the image recognition result indicates that there are no air bubbles in the dispensing container, and the liquid level recognition result indicates that the position of the solid line of the liquid level in the dispensing pipeline meets the preset requirements, then the air bubble extraction process is completed.

[0060] Understandably, the second image acquisition refers to the acquisition of images of one or more dispensing containers currently specified or in a pending acquisition state, and their surroundings. The surrounding images include the dispensing tubing corresponding to the dispensing container. "Currently specified" means specified / selected during the current run / program execution. "Dispensing containers in a pending acquisition state" refers to dispensing containers marked as pending processing during program execution, i.e., images of the dispensing containers taken after the liquid extraction operation. The liquid level recognition result refers to whether the position of the solid line of the liquid level in the dispensing tubing meets preset requirements. The characteristics of the solid line of the liquid level are solid line stripes, and / or dynamic changes in the solid line (which can be understood as changes over time, the solid line is not fixed), and / or horizontal liquid level characteristics, and / or the absence of surrounding bubble-shaped interference (which can be understood as the solid line being free of bubble-shaped interference lines around it). The dispensing tubing refers to a section of tubing extending from the container opening of the dispensing container. The position of the solid line of the liquid level refers to the distance of the solid line of the liquid level from the container opening of the dispensing container. The preset requirement refers to the minimum liquid level in the dispensing pipeline that meets the preset requirement, and does not exceed the maximum preset requirement, i.e., it does not exceed the dispensing pipeline limit. The image recognition result refers to the re-identification of whether air bubbles exist in the dispensing container after the liquid extraction operation.

[0061] Specifically, after performing a preset number of liquid extraction operations, the dispensing container 7 is photographed using a camera positioned next to it, thus obtaining a second acquired image. Based on this second acquired image, bubble identification is performed on the dispensing container 7. Specifically, the first recognition module in the image recognition model performs bubble identification on the second acquired image, which involves converting the dispensing container 7 in the second acquired image to grayscale, resulting in a grayscale image corresponding to the dispensing container in the second acquired image. Then, the border of the dispensing container 7 is removed, i.e., the edges of the grayscale image are cut. Next, grayscale value analysis is performed on the grayscale image. When an arc-shaped boundary feature appears in the grayscale image, an image recognition result indicating the presence of bubbles in the dispensing container 7 in the first acquired image is obtained. When no arc-shaped boundary feature appears in the grayscale image, an image recognition result indicating the absence of bubbles in the dispensing container 7 in the first acquired image is obtained.

[0062] Further, liquid level identification is performed on the dispensing pipeline connected to the dispensing container 7 based on the second acquired image. Specifically, the second recognition module in the image recognition model performs liquid level identification on the dispensing pipeline in the second acquired image. This involves using the capabilities learned during the training of the second module to identify whether a solid liquid level line appears in the dispensing pipeline in the second acquired image. When the second module identifies a solid liquid level line in the dispensing pipeline in the second acquired image, it determines the position of the solid liquid level line from the liquid level line to the container opening of the dispensing container 7 and obtains a preset requirement. This preset requirement is compared with the position of the solid liquid level line. If the position of the solid liquid level line is greater than or equal to the preset requirement, a liquid level identification result indicating that the position of the solid liquid level line in the dispensing pipeline meets the preset requirement is obtained. If the position of the solid liquid level line is less than the preset requirement, a liquid level identification result indicating that the position of the solid liquid level line in the dispensing pipeline does not meet the preset requirement is obtained. If air bubbles are present in the dispensing pipeline 7, multiple solid liquid level lines will be identified; the solid liquid level line closest to the container opening of the dispensing container is then determined as the position of the solid liquid level line. If the image recognition result indicates the presence of air bubbles in the dispensing container 7, or if the liquid level recognition result indicates that the position of the solid line on the liquid level in the dispensing pipeline does not meet the preset requirements, the liquid extraction operation continues, and a new second image is acquired. If the image recognition result indicates that there are no air bubbles in the dispensing container 7, and the liquid level recognition result indicates that the position of the solid line on the liquid level in the dispensing pipeline meets the preset requirements, the air bubble extraction process is completed.

[0063] In this embodiment, by acquiring a second image of the dispensing container after the liquid extraction operation, the air bubbles in the dispensing container are detected again, and the liquid level in the dispensing pipeline is identified. This enables the acquisition of image recognition results and liquid level recognition results, and thus determines whether the extraction process is completed or the liquid extraction operation is to continue.

[0064] It should be understood that the sequence number of each step in the above embodiments does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0065] In one embodiment, a liquid dispensing device is provided, which corresponds one-to-one with the liquid dispensing method described in the above embodiments. For example... Figure 2 As shown, the liquid dispensing device includes a gas discharge module 10, a liquid dispensing module 20, and a bubble extraction module 30. Detailed descriptions of each functional module are as follows: Gas discharge module 10 is used to perform initialization operations to discharge gas from the delivery pipeline and all dispensing containers through the injection container; The liquid dispensing module 20 is used to transfer the target solution in the product container to the injection container. After the target solution is dispensed into the dispensing container through the injection container via a control valve, the bubble visual recognition model is used to identify bubbles in each dispensing container to obtain bubble recognition results that correspond one-to-one with each dispensing container. The bubble extraction module 30 is used to perform bubble extraction processing on each of the sub-containers based on all the bubble identification results, and to perform heat sealing on each of the sub-containers to obtain the sealed sub-containers.

[0066] In one embodiment, the gas exhaust module 10 includes: A gas extraction unit is used to extract gas from the delivery pipeline and all the dispensing containers into the injection container; A gas emission unit is used to close the dispensing control valve connected to the dispensing container, thereby discharging the gas in the injection container to the outside or into the product container.

[0067] In one embodiment, the control valve is a rotary three-way valve, and one rotary three-way valve corresponds to one dispensing container; the liquid dispensing module 20 includes: The solution dispensing unit is used to open the rotary three-way valve corresponding to the dispensing container to extract a target volume of solution from the injection container and deliver it to the dispensing container; the target volume includes the residual volume in the delivery pipeline and the preset volume in the dispensing container.

[0068] In one embodiment, the liquid dispensing module 20 further includes: The first image acquisition unit is used to acquire a first image corresponding to the dispensing container; The bubble recognition result unit is used to perform bubble recognition on the first acquired image through the bubble visual recognition model to obtain the bubble recognition result. The container includes a bubble unit, used to determine that the dispensing container contains bubbles when the bubble recognition result indicates that there are arc-shaped boundary features in the first acquired image; The container does not contain bubble units, which is used to determine that the dispensing container does not contain bubbles when the bubble recognition result indicates that there are no arc-shaped boundary features in the first acquired image.

[0069] In one embodiment, the bubble recognition result unit includes: The region identification subunit is used to perform region identification on the first acquired image through the bubble visual recognition model to obtain a container image corresponding to each of the dispensing containers; The feature extraction subunit is used to extract features from the container image through the bubble visual recognition model to obtain image features corresponding to each container image. The classification and recognition subunit is used to classify and recognize the image features through the bubble visual recognition model to obtain the bubble recognition result corresponding to the packaging container.

[0070] In one embodiment, the bubble extraction module 30 includes: A liquid extraction unit is used to perform a preset number of liquid extraction operations to remove air bubbles adhering to the inner wall of the dispensing container from the dispensing container. A secondary bubble recognition unit is used to perform secondary bubble recognition on each of the dispensing containers after the liquid extraction operation until each of the dispensing containers is recognized as bubble-free.

[0071] In one embodiment, the secondary bubble recognition unit includes: The second image acquisition subunit is used to acquire a second image of the dispensing container after the liquid extraction operation is performed. The image recognition result subunit is used to perform bubble recognition on the dispensing container based on the second acquired image to obtain the image recognition result; The liquid level recognition result subunit is used to perform liquid level recognition on the dispensing pipeline connected to the dispensing container based on the second acquired image, and obtain the liquid level recognition result. The acquisition subunit continues to execute, and if the image recognition result indicates that there are air bubbles in the dispensing container, or the liquid level recognition result indicates that the position of the solid line of the liquid level in the dispensing pipeline does not meet the preset requirements, then the liquid extraction operation continues and a new second acquisition image is acquired. The extraction processing completion subunit is used to complete the bubble extraction process if the image recognition result indicates that there are no bubbles in the dispensing container and the liquid level recognition result indicates that the position of the solid line of the liquid level in the dispensing pipeline meets the preset requirements.

[0072] A liquid dispensing device includes a controller, a product container, a dispensing container, an injection container, and a control valve, wherein the controller is used to perform a liquid dispensing procedure.

[0073] Specific limitations regarding the liquid dispensing equipment, controller, and their individual units and modules can be found in the limitations of the liquid dispensing method described above, and will not be repeated here. Each module in the aforementioned controller can be implemented entirely or partially through software, hardware, or a combination thereof. Understandably, the controller includes a processor, memory, network interface, and database connected via a device bus. Each module of the controller can be embedded in hardware or independent of the processor, or stored in memory as software, so that the processor can call and execute the corresponding operations of each module. The processor provides computational and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores operating devices, computer programs, and a database. The internal memory provides an environment for the operation of the operating devices and computer programs in the non-volatile storage media. The database stores data used in the liquid dispensing method of the above embodiments. The network interface is used for communication with external terminals via a network connection. When the computer program is executed by the processor, it implements a liquid dispensing method.

[0074] In one embodiment, a computer-readable storage medium is provided, the computer-readable storage medium storing a computer program that, when executed by a processor, implements the above-described liquid dispensing method.

[0075] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium. When executed, the computer program can include the processes of the embodiments of the above methods. Any references to memory, storage, databases, or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in a variety of forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), RAMbus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0076] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the above-described division of functional units and modules is used as an example. In practical applications, the above functions can be assigned to different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above.

[0077] The above-described embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be included within the protection scope of the present invention.

Claims

1. A liquid dispensing method, characterized in that, include: Perform initialization to purge gas from the delivery lines and all dispensing containers through the injection container; The target solution in the product container is transferred to the injection container. After the target solution is dispensed into the dispensing container through the injection container via the control valve, the bubble visual recognition model is used to identify bubbles in each of the dispensing containers to obtain bubble recognition results that correspond one-to-one with each of the dispensing containers. Based on all the bubble identification results, bubble extraction is performed on each of the sub-containers, and each of the sub-containers is heat-sealed to obtain the sealed sub-containers.

2. The liquid dispensing method as described in claim 1, characterized in that, The initialization operation, which involves purging gas from the delivery lines and all dispensing containers through the injection container, includes: Gas is drawn from the delivery pipeline and all the dispensing containers into the injection container; Close the dispensing control valve connected to the dispensing container to release the gas in the injection container to the outside or into the product container.

3. The liquid dispensing method as described in claim 1, characterized in that, The control valve is a rotary three-way valve, and one rotary three-way valve corresponds to one dispensing container; The process of dispensing the target solution into a dispensing container via the injection container and a control valve includes: Open the rotary three-way valve corresponding to the dispensing container to extract the target volume of solution from the injection container and deliver it to the dispensing container; the target volume includes the residual volume in the delivery pipeline and the preset volume in the dispensing container.

4. The liquid dispensing method as described in claim 1, characterized in that, The step of identifying bubbles in each of the dispensing containers using a bubble visual recognition model to obtain bubble recognition results corresponding to each of the dispensing containers includes: Acquire a first image corresponding to the dispensing container; The bubble recognition model is used to identify bubbles in the first acquired image to obtain bubble recognition results. When the bubble recognition result indicates the presence of arc-shaped boundary features in the first acquired image, it is determined that the dispensing container contains bubbles; When the bubble recognition result indicates that there are no arc-shaped boundary features in the first acquired image, it is determined that the dispensing container does not contain bubbles.

5. The liquid dispensing method as described in claim 4, characterized in that, The step of performing bubble recognition on the first acquired image using the bubble visual recognition model to obtain bubble recognition results includes: The bubble visual recognition model is used to identify regions in the first acquired image to obtain container images corresponding to each of the dispensing containers. The bubble visual recognition model is used to extract features from the container images to obtain image features corresponding to each container image. The bubble visual recognition model is used to classify and recognize the image features to obtain the bubble recognition result corresponding to the packaging container.

6. The liquid dispensing method as described in claim 1, characterized in that, The step of performing bubble extraction processing on each of the dispensing containers based on all the bubble identification results includes: Perform a preset number of liquid extraction operations to remove air bubbles adhering to the inner wall of the dispensing container from the dispensing container; After the liquid extraction operation, each of the dispensing containers is subjected to secondary bubble identification until each of the dispensing containers is identified as bubble-free.

7. The liquid dispensing method as described in claim 6, characterized in that, The step of performing secondary bubble detection on each of the dispensing containers after the extraction operation until each of the dispensing containers is identified as bubble-free includes: A second image of the dispensing container after the liquid extraction operation is performed; Based on the second acquired image, bubble recognition is performed on the dispensing container to obtain the image recognition result; Based on the second acquired image, the liquid level of the dispensing pipeline connected to the dispensing container is identified to obtain the liquid level identification result. If the image recognition result indicates that there are air bubbles in the dispensing container, or if the liquid level recognition result indicates that the position of the solid line of the liquid level in the dispensing pipeline does not meet the preset requirements, then the liquid extraction operation continues and a new second image is acquired. If the image recognition result indicates that there are no air bubbles in the dispensing container, and the liquid level recognition result indicates that the position of the solid line of the liquid level in the dispensing pipeline meets the preset requirements, then the air bubble extraction process is completed.

8. A liquid dispensing device, characterized in that, include: The gas discharge module is used to perform initialization operations to discharge gas from the delivery line and all dispensing containers through the injection container; The liquid dispensing module is used to transfer the target solution in the product container to the injection container. After the target solution is dispensed into the dispensing container through the injection container via a control valve, a bubble visual recognition model is used to identify bubbles in each dispensing container to obtain bubble recognition results that correspond one-to-one with each dispensing container. The bubble extraction module is used to extract bubbles from each of the sub-containers based on all the bubble identification results, and to heat-seal each of the sub-containers to obtain the sealed sub-containers.

9. A liquid dispensing device, characterized in that, It includes a controller, a product container, a dispensing container, an injection container, and a control valve, wherein the controller is used to perform the liquid dispensing method as described in any one of claims 1 to 7.

10. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by the processor, it implements the liquid dispensing method as described in any one of claims 1 to 7.

Images