A liquid recognition method based on the fusion of visual and olfactory information and its related device

Abstract

The present application discloses a liquid recognition method and related device based on the fusion of visual and olfactory information. The method includes: collecting a liquid image of the liquid to be recognized when irradiated by blue-green laser light; obtaining the olfactory information of the liquid to be recognized; extracting visual features from the liquid image of the liquid to be recognized through an image recognition model, and performing recognition based on the visual features to obtain a first recognition result of the liquid to be recognized; extracting odor features from the olfactory information of the liquid to be recognized through an odor recognition model, and performing recognition based on the odor features to obtain a second recognition result of the liquid to be recognized; fusing the first recognition result and the second recognition result of the liquid to be recognized to obtain a final recognition result of the liquid to be recognized, which improves the technical problem in the prior art that relies on visual information for recognition. Since liquids with similar colors have few visual features and single visual information is used for recognition, the physical and chemical properties of the liquid cannot be comprehensively reflected, resulting in a low recognition rate.

Description

Technical Field

[0001] The present application relates to the technical field of liquid recognition, and in particular, to a liquid recognition method based on the fusion of visual and olfactory information and related devices thereof. Background Art

[0002] At present, in the field of intelligent perception, the technical requirements for machines to perceive environmental objects show a trend of short supply in industries such as industrial automation and intelligent home entry, and the recognition and detection of liquids are the research and application hotspots of current perception technologies. Existing liquid recognition methods mainly rely on visual information. However, liquids with similar colors have few visual features, resulting in low recognition rates of existing recognition methods. Moreover, using a single visual information for recognition cannot comprehensively reflect the physical and chemical properties of liquids, and there is a problem of low recognition rate. Summary of the Invention

[0003] The present application provides a liquid recognition method based on the fusion of visual and olfactory information and related devices thereof, which is used to improve the technical problem that existing technologies rely on visual information for recognition, and due to the few visual features of liquids with similar colors and the use of a single visual information for recognition, the physical and chemical properties of liquids cannot be comprehensively reflected, resulting in a low recognition rate.

[0004] In view of this, in the first aspect of the present application, a liquid recognition method based on the fusion of visual and olfactory information is provided, including:

[0005] Collecting a liquid image of the liquid to be recognized when irradiated by blue-green laser light;

[0006] Obtaining the olfactory information of the liquid to be recognized;

[0007] Extracting visual features from the liquid image of the liquid to be recognized through an image recognition model, and performing recognition based on the visual features to obtain a first recognition result of the liquid to be recognized;

[0008] Extracting odor features from the olfactory information of the liquid to be recognized through an odor recognition model, and performing recognition based on the odor features to obtain a second recognition result of the liquid to be recognized;

[0009] Fusing the first recognition result and the second recognition result of the liquid to be recognized to obtain a final recognition result of the liquid to be recognized.

[0010] Optionally, the training processes of the image recognition model and the odor recognition model are as follows:

[0011] Obtaining liquid images and olfactory information of a number of liquid samples with known categories and concentrations when irradiated by blue-green light;

[0012] Input the liquid image of the liquid sample into an image recognition network for recognition, and input the olfactory information of the liquid sample into an odor recognition network for recognition;

[0013] Fuse the recognition results of the image recognition network and the recognition results of the odor recognition network to obtain the final recognition result of the liquid sample;

[0014] Calculate the loss value based on the final recognition result of the liquid sample and the true label of the liquid sample, and update the parameters of the image recognition network and the parameters of the odor recognition network through the loss value until the image recognition network and the odor recognition network converge, obtaining a trained image recognition model and a trained odor recognition model.

[0015] Optionally, the image recognition model is EfficientNetV2.

[0016] Optionally, the odor recognition model is an odor calibration convolutional autoencoder network.

[0017] The second aspect of this application provides a liquid recognition system based on the fusion of visual and olfactory information, including:

[0018] A visual information acquisition module for acquiring the liquid image of the liquid to be recognized when irradiated by blue-green laser;

[0019] An olfactory information acquisition module for obtaining the olfactory information of the liquid to be recognized;

[0020] A central controller for extracting visual features from the liquid image of the liquid to be recognized through an image recognition model, and performing recognition based on the visual features to obtain the first recognition result of the liquid to be recognized; extracting odor features from the olfactory information of the liquid to be recognized through an odor recognition model, and performing recognition based on the odor features to obtain the second recognition result of the liquid to be recognized; fusing the first recognition result and the second recognition result of the liquid to be recognized to obtain the final recognition result of the liquid to be recognized.

[0021] Optionally, the visual information acquisition module includes: a camera and two lasers;

[0022] One laser is used to emit blue laser to irradiate the liquid to be recognized, and the other laser is used to emit green laser to irradiate the liquid to be recognized;

[0023] The camera is used to acquire the liquid image of the liquid to be recognized when irradiated by blue-green laser.

[0024] Optionally, the olfactory information acquisition module includes a metal oxide gas sensor array, a gas chamber, and an analog-to-digital conversion module.

[0025] Optionally, the image recognition model is EfficientNetV2, and the odor recognition model is an odor calibration convolutional autoencoder network.

[0026] The third aspect of the present application provides a liquid recognition device, which includes a processor and a memory;

[0027] The memory is used to store program code and transmit the program code to the processor;

[0028] The processor is used to execute any one of the liquid recognition methods based on visual and olfactory information fusion described in the first aspect according to the instructions in the program code.

[0029] The fourth aspect of the present application provides a computer-readable storage medium, which is used to store program code, and when the program code is executed by a processor, it implements any one of the liquid recognition methods based on visual and olfactory information fusion described in the first aspect.

[0030] It can be seen from the above technical solutions that the present application has the following advantages:

[0031] The present application provides a liquid recognition method based on visual and olfactory information fusion, including: collecting a liquid image of a liquid to be recognized when irradiated by blue-green laser; obtaining the olfactory information of the liquid to be recognized; extracting visual features from the liquid image of the liquid to be recognized through an image recognition model, and performing recognition based on the visual features to obtain a first recognition result of the liquid to be recognized; extracting odor features from the olfactory information of the liquid to be recognized through an odor recognition model, and performing recognition based on the odor features to obtain a second recognition result of the liquid to be recognized; fusing the first recognition result and the second recognition result of the liquid to be recognized to obtain a final recognition result of the liquid to be recognized.

[0032] In the present application, a liquid image of a liquid to be recognized when irradiated by blue-green laser is collected, and the propagation characteristics of lasers with different wavelengths in different medium liquids are utilized to increase the depth information of visual perception, so as to obtain highly distinguishable information visually; and the olfactory information of the liquid to be recognized is obtained, and different recognition models are used to respectively extract features and perform liquid recognition on the extracted liquid image and olfactory information, and finally these two recognition results are fused to obtain a final recognition result, making full use of visual information and olfactory information for liquid recognition, considering the multi-dimensional information of the liquid, and being able to comprehensively reflect the physical and chemical properties of the liquid, thereby improving the technical problem in the prior art that relies on visual information for recognition, where liquids with similar colors have few visual features and single visual information is used for recognition, and the physical and chemical properties of the liquid cannot be comprehensively reflected, resulting in a low recognition rate. Description of the Drawings

[0033] To more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the accompanying drawings required for the description of the embodiments or the prior art. Obviously, the accompanying drawings in the following description are only some embodiments of the present application. For those of ordinary skill in the art, without creative efforts, other accompanying drawings can also be obtained based on these drawings.

[0034] Figure 1 FIG. 4 is a schematic flow chart of a liquid recognition method based on visual and olfactory information fusion provided by an embodiment of the present application;

[0035] Figure 2 FIG. 5 is a schematic structural diagram of a liquid recognition system based on visual and olfactory information fusion provided by an embodiment of the present application. Detailed implementation manners

[0036] In order to enable those skilled in the art to better understand the solutions of the present application, the following will clearly and completely describe the technical solutions in the embodiments of the present application with reference to the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments of the present application, all other embodiments obtained by those of ordinary skill in the art without creative efforts belong to the scope of protection of the present application.

[0037] For ease of understanding, please refer to Figure 1 , an embodiment of the present application provides a liquid recognition method based on visual and olfactory information fusion, including:

[0038] Step 101: Collect a liquid image of the liquid to be recognized when irradiated by blue-green laser.

[0039] Two lasers can be used to emit blue laser and green laser respectively to irradiate the liquid to be recognized, and then an industrial camera is used to collect the visual image of the liquid to be recognized under the blue-green laser irradiation to obtain the liquid image of the liquid to be recognized. The present application utilizes the propagation characteristics of lasers with different wavelengths in different medium liquids to increase the depth information of visual perception, and then collects the visual image of the liquid to be recognized when irradiated by blue-green laser. Through experiments, it is found that the liquid recognition accuracy of using the visual image corresponding to the simultaneous irradiation of the liquid to be recognized by blue laser and green laser for recognition is higher than that of using the visual image corresponding to the irradiation of a single green laser, blue laser or red laser.

[0040] Step 102: Obtain the olfactory information of the liquid to be recognized.

[0041] The response signal of the volatile odor substances of the liquid to be recognized can be obtained through a metal oxide gas sensor array and a gas chamber as the olfactory information of the liquid to be recognized. The metal oxide gas sensor array can convert the gas signal into an electrical signal, and then the analog-to-digital conversion module eliminates the noise signal in the electrical signal and amplifies the useful signal to improve stability. The finally output digital response signal is the olfactory information, and the olfactory information is a numerical matrix.

[0042] Step 103: Extract the visual features of the liquid image of the liquid to be recognized through an image recognition model, and perform recognition based on the visual features to obtain the first recognition result of the liquid to be recognized.

[0043] After the liquid image of the liquid to be recognized is extracted, it can be preprocessed, and then the pre-trained image recognition model is used to extract the visual features of the liquid image of the liquid to be recognized, and the liquid is classified and its concentration is detected based on the extracted visual features, so as to output the first recognition result of the liquid to be recognized.

[0044] In the embodiment of the present application, EfficientNetV2 can be used as the image recognition model.

[0045] Step 104: Extract the odor features of the olfactory information of the liquid to be recognized through an odor recognition model, and perform recognition based on the odor features to obtain the second recognition result of the liquid to be recognized.

[0046] After the olfactory information of the liquid to be recognized is extracted, it can be preprocessed, and then the pre-trained odor recognition model is used to extract the odor features of the olfactory information of the liquid to be recognized, and the liquid to be recognized is classified and its concentration is detected based on the extracted odor features, so as to output the second recognition result of the liquid to be recognized.

[0047] In the embodiment of the present application, an odor calibration convolutional autoencoder network can be used as the odor recognition model.

[0048] Step 105: Fuse the first recognition result and the second recognition result of the liquid to be recognized to obtain the final recognition result of the liquid to be recognized.

[0049] After obtaining the first recognition result in terms of vision and the second recognition result in terms of smell, these two recognition results are fused, that is, the visual perception result and the olfactory perception result are fused to obtain the final recognition result of the liquid to be recognized. In this way, when certain similar liquids cannot be recognized through visual information, the second recognition result corresponding to the olfactory information can be combined to determine the final recognition result, and the recognition result includes the liquid category and the liquid concentration.

[0050] Furthermore, the training processes of the image recognition model and the odor recognition model are as follows:

[0051] Obtain the liquid images and olfactory information of several liquid samples with known categories and concentrations when irradiated by blue-green light;

[0052] Input the liquid images of the liquid samples into an image recognition network for recognition, and input the olfactory information of the liquid samples into an odor recognition network for recognition;

[0053] Fuse the recognition results of the image recognition network and the odor recognition network to obtain the final recognition result of the liquid sample;

[0054] Calculate the loss value according to the final recognition result of the liquid sample and the true label of the liquid sample, and update the parameters of the image recognition network and the odor recognition network through the loss value until the image recognition network and the odor recognition network converge, obtaining a trained image recognition model and a trained odor recognition model.

[0055] Before training the model, a data set needs to be constructed. According to the need to identify liquids, establish a data set for model network training from the visual information (i.e., liquid images when irradiated by blue-green laser) and olfactory information of several collected liquid samples. Each liquid sample in the data set has a corresponding true label, and the label information includes the liquid category and liquid concentration. Input the liquid images of the liquid samples in the data set into an image recognition network for recognition, input the olfactory information of the liquid sample into an odor recognition network for recognition, then fuse the recognition results of these two recognition networks, and calculate the loss value in combination with the true label of the liquid sample. Update the network parameters of these two networks through back-iteration of the loss value. These two networks are jointly trained by superimposing the loss values. When the training error is lower than the preset error threshold, or the number of iterations during training reaches the preset maximum number of iterations, it is considered that the model converges, obtaining a trained image recognition model and a trained odor recognition model.

[0056] This application proposes a joint framework deep neural network perception model for image recognition models and odor recognition models to achieve the discrimination and concentration detection of liquids. The embodiments of this application consider that the collected liquid images and olfactory information are two types of modal data. There are differences in the data dimensions of these two types of modal data and the sensors used for collection. If these two types of modal data are directly fused and input into a network for recognition, the recognition effect cannot be maximally improved. Therefore, the embodiments of this application use two different models to perform feature learning and recognition on these two different types of modal data, and finally fuse the recognition results corresponding to these two types of modal data, making full use of visual information and olfactory information for liquid recognition. It not only considers the multi-dimensional information of liquids but also can avoid the problem of mutual influence between data caused by directly fusing different modal data, and can effectively improve the accuracy and robustness of liquid recognition.

[0057] In the embodiments of the present application, a liquid image of a liquid to be recognized when irradiated by blue-green laser is collected. By utilizing the propagation characteristics of lasers with different wavelengths in different medium liquids, the visual perception depth information is increased, thereby obtaining visually highly distinguishable information. And the olfactory information of the liquid to be recognized is obtained, and the extracted liquid image and olfactory information are respectively subjected to feature extraction and liquid recognition through different recognition models. Finally, the two recognition results are fused to obtain the final recognition result, making full use of visual information and olfactory information for liquid recognition, considering the multi-dimensional information of the liquid, and comprehensively reflecting the physical and chemical properties of the liquid. Thus, it improves the technical problem in the prior art that relies on visual information for recognition, where the visual features of liquids with similar colors are few and the recognition is carried out using a single visual information, unable to comprehensively reflect the physical and chemical properties of the liquid and resulting in a low recognition rate.

[0058] The above is an embodiment of a liquid recognition method based on the fusion of visual and olfactory information provided by the present application. The following is an embodiment of a liquid recognition system based on the fusion of visual and olfactory information provided by the present application.

[0059] Please refer to Figure 2 , a liquid recognition system based on the fusion of visual and olfactory information provided by the embodiments of the present application, includes:

[0060] A visual information acquisition module, configured to collect a liquid image of a liquid to be recognized when irradiated by blue-green laser;

[0061] An olfactory information acquisition module, configured to obtain the olfactory information of the liquid to be recognized;

[0062] A central controller, configured to perform visual feature extraction on the liquid image of the liquid to be recognized through an image recognition model, and perform recognition based on the visual features to obtain a first recognition result of the liquid to be recognized; perform odor feature extraction on the olfactory information of the liquid to be recognized through an odor recognition model, and perform recognition based on the odor features to obtain a second recognition result of the liquid to be recognized; fuse the first recognition result and the second recognition result of the liquid to be recognized to obtain the final recognition result of the liquid to be recognized.

[0063] Further, the visual information acquisition module includes: a camera and two lasers;

[0064] One laser is used to emit blue laser to irradiate the liquid to be recognized, and the other laser is used to emit green laser to irradiate the liquid to be recognized;

[0065] The camera is configured to collect a liquid image of the liquid to be recognized when irradiated by blue-green laser.

[0066] Further, the olfactory information acquisition module includes a metal oxide gas sensor array, a gas chamber, and an analog-to-digital conversion module (ADC).

[0067] Further, the image recognition model is EfficientNetV2, and the odor recognition model is an odor calibration convolutional autoencoder network.

[0068] Further, the liquid recognition system based on visual and olfactory information fusion in the embodiments of the present application further includes a communication module. The visual information acquisition module and the olfactory information acquisition module transmit the information collected by each of them to the central controller through the communication module, so that the central controller performs liquid recognition according to the received visual image and olfactory information.

[0069] Further, the liquid recognition system based on visual and olfactory information fusion in the embodiments of the present application further includes a power supply module for supplying power to the central controller.

[0070] In the embodiments of the present application, a liquid image of the liquid to be recognized when irradiated by blue-green laser is collected. By using the propagation characteristics of lasers with different wavelengths in different medium liquids, the visual perception depth information is increased, so as to obtain visually highly distinguishable information; and the olfactory information of the liquid to be recognized is obtained, and different recognition models are used to extract features and perform liquid recognition on the extracted liquid image and olfactory information respectively. Finally, the two recognition results are fused to obtain the final recognition result, making full use of visual information and olfactory information for liquid recognition, considering the multi-dimensional information of the liquid, and comprehensively reflecting the physical and chemical properties of the liquid, thereby improving the prior art that relies on visual information for recognition. Since liquids with similar colors have few visual features and single visual information is used for recognition, the physical and chemical properties of the liquid cannot be comprehensively reflected, and there is a technical problem of low recognition rate.

[0071] The embodiments of the present application also provide a liquid recognition device, which includes a processor and a memory;

[0072] The memory is used to store program codes and transmit the program codes to the processor;

[0073] The processor is used to execute the liquid recognition method based on visual and olfactory information fusion in the foregoing method embodiments according to the instructions in the program codes.

[0074] The embodiments of the present application also provide a computer-readable storage medium, which is used to store program codes. When the program codes are executed by a processor, the liquid recognition method based on visual and olfactory information fusion in the foregoing method embodiments is implemented.

[0075] Those skilled in the art can clearly understand that for the convenience and conciseness of description, the specific working processes of the systems and units described above can refer to the corresponding processes in the foregoing method embodiments, and will not be elaborated here.

[0076] In the description of this application and the above-mentioned accompanying drawings, terms such as "first", "second", "third", "fourth", etc. (if any) are used to distinguish similar objects and do not necessarily describe a specific order or sequence. It should be understood that the data used in this way can be interchanged under appropriate circumstances so that the embodiments of this application described herein can be implemented in an order different from those illustrated or described herein. In addition, the terms "comprising" and "having" and any variations thereof are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or device that comprises a series of steps or units does not necessarily have to be limited to those steps or units clearly listed, but may include other steps or units not clearly listed or inherent to these processes, methods, products, or devices.

[0077] It should be understood that in this application, "at least one (item)" means one or more, and "a plurality" means two or more. "And / or" is used to describe the association relationship of associated objects and indicates that there can be three relationships. For example, "A and / or B" can mean: only A exists, only B exists, and both A and B exist simultaneously. Here, A and B can be singular or plural. The character " / " generally indicates that the associated objects before and after are in an "or" relationship. "At least one (one) of the following" or its similar expressions refer to any combination of these items, including any combination of single items (ones) or plural items (ones). For example, at least one (one) of a, b, or c can mean: a, b, c, "a and b", "a and c", "b and c", or "a and b and c", where a, b, and c can be single or multiple.

[0078] In several embodiments provided in this application, it should be understood that the disclosed devices and methods can be implemented in other ways. For example, the device embodiments described above are merely illustrative. For example, the division of the units is only a logical function division, and there can be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the displayed or discussed coupling or direct coupling or communication connection to each other can be through some interfaces, and the indirect coupling or communication connection of the devices or units can be in electrical, mechanical, or other forms.

[0079] The units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they can be located in one place or distributed to multiple network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0080] In addition, in each embodiment of the present application, each functional unit can be integrated into a processing unit, or each unit can exist physically alone, or two or more units can be integrated into one unit. The above-mentioned integrated unit can be implemented in the form of hardware or in the form of a software functional unit.

[0081] If the above-mentioned integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on such an understanding, the technical solution of the present application, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in each embodiment of the present application. The foregoing storage medium includes: various media that can store program codes, such as USB flash drives, mobile hard disks, read-only memories (English full name: Read-Only Memory, English abbreviation: ROM), random access memories (English full name: Random Access Memory, English abbreviation: RAM), magnetic disks, or optical discs.

[0082] As described above, the above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them; although the present application has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions described in the foregoing embodiments, or perform equivalent replacements on some of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of each embodiment of the present application.

Claims

1. A liquid recognition method based on the fusion of visual and olfactory information, characterized in that, Including: Collecting a liquid image of a liquid to be recognized when irradiated by blue laser and green laser; Obtaining olfactory information of the liquid to be recognized; Performing visual feature extraction on the liquid image of the liquid to be recognized through an image recognition model, and performing recognition based on the visual features to obtain a first recognition result of the liquid to be recognized; Performing odor feature extraction on the olfactory information of the liquid to be recognized through an odor recognition model, and performing recognition based on the odor features to obtain a second recognition result of the liquid to be recognized; Fusing the first recognition result and the second recognition result of the liquid to be recognized to obtain a final recognition result of the liquid to be recognized; The training processes of the image recognition model and the odor recognition model are as follows: Obtaining liquid images and olfactory information of a number of liquid samples with known categories and concentrations when irradiated by blue laser and green laser; Inputting the liquid images of the liquid samples into an image recognition network for recognition, and inputting the olfactory information of the liquid samples into an odor recognition network for recognition; Fusing the recognition results of the image recognition network and the recognition results of the odor recognition network to obtain a final recognition result of the liquid samples; Calculating a loss value according to the final recognition result of the liquid samples and the true labels of the liquid samples, and updating the parameters of the image recognition network and the parameters of the odor recognition network through the loss value until the image recognition network and the odor recognition network converge, obtaining a trained image recognition model and a trained odor recognition model.

2. The liquid recognition method based on visual and olfactory information fusion according to claim 1, characterized in that, The image recognition model is EfficientNet V2.

3. The liquid recognition method based on visual and olfactory information fusion according to claim 1, wherein The odor recognition model is an odor calibration convolutional autoencoder network.

4. A liquid recognition system based on the fusion of visual and olfactory information, characterized in that, Including: A visual information collection module for collecting a liquid image of a liquid to be recognized when irradiated by blue laser and green laser; An olfactory information collection module for obtaining olfactory information of the liquid to be recognized; A central controller for performing visual feature extraction on the liquid image of the liquid to be recognized through an image recognition model, and performing recognition based on the visual features to obtain a first recognition result of the liquid to be recognized; Performing odor feature extraction on the olfactory information of the liquid to be recognized through an odor recognition model, and performing recognition based on the odor features to obtain a second recognition result of the liquid to be recognized; Fusing the first recognition result and the second recognition result of the liquid to be recognized to obtain a final recognition result of the liquid to be recognized; The training processes of the image recognition model and the odor recognition model are as follows: Obtaining liquid images and olfactory information of a number of liquid samples with known categories and concentrations when irradiated by blue laser and green laser; Inputting the liquid images of the liquid samples into an image recognition network for recognition, and inputting the olfactory information of the liquid samples into an odor recognition network for recognition; Fusing the recognition results of the image recognition network and the recognition results of the odor recognition network to obtain a final recognition result of the liquid samples; Calculate a loss value based on the final recognition result of the liquid sample and the true label of the liquid sample, and update the parameters of the image recognition network and the parameters of the odor recognition network through the loss value until the image recognition network and the odor recognition network converge, obtaining a trained image recognition model and a trained odor recognition model.

5. The liquid recognition system based on visual and olfactory information fusion according to claim 4, characterized in that The visual information acquisition module includes: a camera and two lasers; One laser is used to emit blue laser to irradiate the liquid to be recognized, and the other laser is used to emit green laser to irradiate the liquid to be recognized; The camera is used to acquire the liquid image of the liquid to be recognized when irradiated by the blue laser and the green laser.

6. The liquid recognition system based on visual and olfactory information fusion according to claim 4, characterized in that, The olfactory information acquisition module includes a metal oxide gas sensor array, a gas chamber, and an analog-to-digital conversion module.

7. The liquid recognition system based on visual and olfactory information fusion according to claim 4, characterized in that, The image recognition model is EfficientNet V2, and the odor recognition model is an odor calibration convolutional autoencoder network.

8. A liquid identification device, characterized in that, The device includes a processor and a memory; The memory is used to store program code and transmit the program code to the processor; The processor is used to execute the liquid recognition method based on visual and olfactory information fusion according to any one of claims 1-3 according to the instructions in the program code.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium is used to store program code, and when the program code is executed by a processor, it implements the liquid recognition method based on visual and olfactory information fusion according to any one of claims 1-3.

Images