Vehicle-mounted fragrance control method, vehicle-mounted controller, system, vehicle and medium

Abstract

The invention discloses a vehicle-mounted fragrance control method, a vehicle-mounted controller, a system, a vehicle and a medium, and the vehicle-mounted fragrance control method comprises the steps: obtaining a face image of a driver; performing emotion recognition on the driver face image by adopting an emotion classification model constructed based on a Transform network, and determining a current emotion type corresponding to the driver face image; querying a fragrance database based on the current emotion type, and determining a target fragrance combination corresponding to the current emotion type; and controlling the fragrance release system to work based on the target fragrance combination. According to the technical scheme, the target fragrance combination corresponding to the current emotion type can be matched in a personalized mode, finally, the fragrance release system is controlled to work based on the target fragrance combination, the target fragrance combination corresponding to the current emotion type is accurately matched, then the emotion requirement of a driver is better met, and the cockpit interaction effect and the driving experience are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of vehicles, and particularly to an in-vehicle fragrance control method, an in-vehicle controller, a system, a vehicle and a medium. Background Art

[0002] With the rapid development of intelligent technologies, the research and application of intelligent driving systems are continuously increasing. The intelligent driving system includes many functions, such as autonomous driving, driving assistance, in-vehicle entertainment, driver monitoring, etc., aiming to improve driving safety, comfort and convenience. In this context, in order to better meet the needs of drivers, emotion recognition and personalized interaction have become increasingly important.

[0003] With the development of technology, the fragrance release system on vehicles has also become more and more intelligent. The existing fragrance release system controls the fragrance release system to release the fragrance corresponding to the recognized emotion by recognizing the driver's emotion. However, due to the low accuracy of the existing system in recognizing the driver's emotion, the released fragrance cannot match the driver's emotional needs. Summary of the Invention

[0004] Embodiments of the present invention provide an in-vehicle fragrance control method, an in-vehicle controller, a system, a vehicle and a medium to solve the problem that the existing system has low accuracy in recognizing the driver's emotion, so that the released fragrance cannot match the driver's emotional needs.

[0005] An in-vehicle fragrance control method includes:

[0006] Obtain a driver's face image;

[0007] Use an emotion classification model constructed based on the Transformer network to perform emotion recognition on the driver's face image, and determine the current emotion type corresponding to the driver's face image;

[0008] Query a fragrance database based on the current emotion type to determine a target fragrance combination corresponding to the current emotion type;

[0009] Control the fragrance release system to work based on the target fragrance combination.

[0010] Further, the step of using an emotion classification model constructed based on the Transformer network to perform emotion recognition on the driver's face image, and determine the current emotion type corresponding to the driver's face image includes:

[0011] Segment the driver's face image to obtain at least two face image blocks;

[0012] Extract features from each of the face image blocks to obtain a face feature vector corresponding to each of the face image blocks;

[0013] Use an emotion classification model constructed based on the Transformer network to perform emotion recognition on the face feature vectors corresponding to at least two of the face image blocks, and determine the current emotion type corresponding to the driver's face image.

[0014] Further, querying the fragrance database based on the current emotion type to determine the target fragrance combination corresponding to the current emotion type includes:

[0015] Query the fragrance database based on the current emotion type to determine whether there is a historical fragrance combination corresponding to the current emotion type;

[0016] If the historical fragrance combination exists, determine the historical fragrance combination as the target fragrance combination;

[0017] If the historical fragrance combination does not exist, use the fuzzy system algorithm to perform matching processing on the current emotion type and the in-stock fragrance labels to determine the target fragrance combination corresponding to the current emotion type.

[0018] Further, using the fuzzy system algorithm to perform matching processing on the current emotion type and the in-stock fragrance combinations to determine the target fragrance combination corresponding to the current emotion type includes:

[0019] Query the emotion-fragrance mapping table based on the current emotion type to determine at least two in-stock fragrance labels corresponding to the current emotion type;

[0020] Based on the fuzzy system algorithm, perform matching processing on the in-stock fragrance combinations corresponding to at least two in-stock fragrance labels to determine the target fragrance combination corresponding to the current emotion type.

[0021] Further, based on the fuzzy system algorithm, performing matching processing on the in-stock fragrance combinations corresponding to at least two in-stock fragrance labels to determine the target fragrance combination corresponding to the current emotion type includes:

[0022] Obtain the emotion matching score, emotional intensity score, and personal preference score corresponding to the current emotion type;

[0023] Based on the target weights, process the emotion matching score, the emotional intensity score, and the personal preference score to determine the target scores corresponding to the in-stock fragrance combinations corresponding to at least two in-stock fragrance labels;

[0024] Determine the in-stock fragrance combination with the maximum target score as the target fragrance combination corresponding to the current emotion type.

[0025] Further, querying the fragrance database based on the current emotion type to determine the target fragrance combination corresponding to the current emotion type includes:

[0026] If the current emotion type is a specific emotion type, obtain the duration corresponding to the specific emotion type;

[0027] If the duration is greater than a preset time, determine the warning fragrance combination as the target fragrance combination corresponding to the current emotion type.

[0028] A vehicle-mounted controller is used to execute the above vehicle-mounted fragrance control method.

[0029] A fragrance control system includes a fragrance release system and the above vehicle-mounted controller.

[0030] A vehicle includes the above fragrance control system.

[0031] A computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the above vehicle-mounted fragrance control method is implemented.

[0032] The above vehicle-mounted fragrance control method, vehicle-mounted controller, system, vehicle and medium obtain the driver's face image, use an emotion classification model constructed based on the Transformer network to perform emotion recognition on the driver's face image, determine the current emotion type corresponding to the driver's face image, to ensure the accuracy of the recognized current emotion type to the greatest extent, and query the fragrance database based on the current emotion type to determine the target fragrance combination corresponding to the current emotion type, to personalize the matching of the target fragrance combination corresponding to the current emotion type, and finally control the fragrance release system to work based on the target fragrance combination, so as to accurately match the target fragrance combination corresponding to the current emotion type, and further better meet the driver's emotional needs and improve the cockpit interaction effect and driving experience. Description of the Drawings

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

[0034] Figure 1 is a flowchart of a vehicle-mounted fragrance control method in an embodiment of the present invention;

[0035] Figure 2 is another flowchart of a vehicle-mounted fragrance control method in an embodiment of the present invention;

[0036] Figure 3 is another flowchart of the vehicle-mounted fragrance control method in an embodiment of the present invention;

[0037] Figure 4 is another flowchart of the vehicle-mounted fragrance control method in an embodiment of the present invention;

[0038] Figure 5 is another flowchart of the vehicle-mounted fragrance control method in an embodiment of the present invention;

[0039] Figure 6 is another flowchart of the vehicle-mounted fragrance control method in an embodiment of the present invention;

[0040] Figure 7 is a schematic diagram of a vehicle-mounted controller in an embodiment of the present invention. Specific Embodiments

[0041] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts belong to the scope of protection of the present invention.

[0042] The vehicle-mounted fragrance control method provided by the embodiment of the present invention can be applied in a vehicle. Specifically, the vehicle-mounted fragrance control method can be applied to the fragrance control system of the vehicle. Exemplarily, the fragrance control system includes a vehicle-mounted controller and a fragrance release system. The vehicle-mounted controller executes the vehicle-mounted fragrance control method, accurately analyzes the driver's emotion, and matches a personalized target fragrance combination to control the operation of the fragrance release system. It can be understood that the fragrance release system can adopt the fragrance release system well-known to those skilled in the art, which will not be elaborated here.

[0043] This embodiment provides a vehicle-mounted fragrance control method, as Figure 1 shown, applied in a vehicle-mounted controller, including:

[0044] S101: Obtain the driver's face image.

[0045] S102: Use an emotion classification model constructed based on the Transformer network to perform emotion recognition on the driver's face image, and determine the current emotion type corresponding to the driver's face image.

[0046] S103: Query the fragrance database based on the current emotion type, and determine the target fragrance combination corresponding to the current emotion type.

[0047] S104: Control the operation of the fragrance release system based on the target fragrance combination.

[0048] Among them, the driver's face image refers to the face image corresponding to the driver in the vehicle cockpit. The Transformer network refers to a neural network that applies the Transformer architecture to computer vision tasks. Specifically, the Transformer network is a Vision Transformer network. The emotion classification model refers to a model for classifying the driver's emotions. The current emotion type refers to the emotion type of the driver at the current moment. Exemplarily, the emotion types include, but are not limited to, happy, angry, anxious, fearful, sad, impatient, tired, and tense. The fragrance database refers to a database for storing fragrance data. Exemplarily, the fragrance data includes at least two in-library fragrance labels and the corresponding in-library fragrance combinations, and each in-library fragrance label and the corresponding in-library fragrance combination are associated with a specific emotion type. Each specific emotion type corresponds to at least two in-library fragrance labels and at least two in-library fragrance combinations corresponding to the at least two in-library fragrance labels. For example, if the emotion type is "happy", the corresponding in-library fragrance labels include "pleasant" or "relaxed", or the in-library fragrance combination formed by "pleasant" and "relaxed". The target fragrance combination refers to a fragrance combination including the fragrance combination corresponding to the current emotion type.

[0049] As an example, in step S101, the vehicle-mounted controller can obtain the driver's face image through the face acquisition device on the vehicle. Exemplarily, the face acquisition device includes a camera. Specifically, the vehicle-mounted controller receives the face information corresponding to the driver output by the face acquisition device, and obtains the driver's face image according to the face information. The face information can be the driver's face image or the driver's face video. When the face information is the driver's face video, the vehicle-mounted controller parses the driver's face video, extracts at least one video frame from the driver's face video, and determines the driver's face image from the at least one video frame. It should be noted that in order to ensure the real-time nature of the driver's emotion analysis, the face acquisition device obtains the driver's face information in real time to ensure that the vehicle-mounted controller can obtain the driver's face image in real time.

[0050] As an example, in step S102, the vehicle-mounted controller uses an emotion classification model constructed based on the Transformer network to perform emotion recognition on the driver's face image and determine the current emotion type corresponding to the driver's face image. Among them, the emotion classification model can be constructed based on the Transformer network and pre-trained to classify the driver's emotions.

[0051] As an example, in step S102, the vehicle-mounted controller constructs an emotion classification model as follows: First, a large number of driver face images are collected in advance as training samples for the emotion classification model. The training samples include driver face images corresponding to different emotion types, and each driver face image is assigned a corresponding emotion label. Then, the collected face images are preprocessed. Exemplarily, the preprocessing process includes adjusting the image size, denoising, and enhancing the contrast, etc. Next, through the Transformer network, emotion features are extracted from the preprocessed driver face images, and the emotion type is determined according to the emotion features. Then, the actual loss value between the emotion type and the emotion label is calculated. When the actual loss value meets the preset condition, an emotion classification model that can accurately classify different emotion types is obtained. The preset condition can be a preset loss threshold. When the actual loss value is less than the loss threshold, it is considered to meet the preset condition. When the preset condition is not met, the above training process is repeated until the actual loss value meets the preset condition.

[0052] In this example, by using an emotion classification model constructed based on the Transformer network to perform emotion recognition on driver face images and determine the current emotion type corresponding to the driver face images, since the Transformer network can capture the associations between different regions of the image through the self-attention mechanism, it can better capture the global information of the image compared with traditional convolutional neural networks and avoid overfitting. Therefore, when performing emotion recognition on driver face images and determining the current emotion type corresponding to the driver face images, the accuracy of the recognized current emotion type can be maximally guaranteed.

[0053] As an example, in step S103, the vehicle-mounted controller queries the fragrance database based on the current emotion type to determine the target fragrance combination corresponding to the current emotion type. In this example, the vehicle-mounted controller accurately determines the current emotion type corresponding to the driver face image through the emotion classification model. Thus, on the premise of accurately judging the driver's current emotion type, based on this current emotion type, the fragrance database is queried. Since the fragrance database stores fragrance data related to specific emotion types, based on this fragrance data and the current emotion type, the in-stock fragrance label corresponding to the current emotion type and the target fragrance combination formed by the in-stock fragrance labels can be determined, so as to accurately match the target fragrance combination corresponding to the current emotion type, and further better meet the driver's emotion needs and improve the cockpit interaction effect and driving experience.

[0054] Specifically, the vehicle-mounted controller can determine the historical fragrance combination corresponding to the current emotion type from the fragrance data stored in the fragrance database according to the current emotion type, and determine the target fragrance combination based on the historical fragrance combination; or it can be to generate a personalized target fragrance combination according to the current emotion type, as well as the in-stock fragrance labels and corresponding in-stock fragrance combinations in the fragrance data, using a preset fragrance combination generation strategy, so as to achieve personalized fragrance matching and better meet the driver's emotional needs. It can be understood that the preset fragrance combination generation strategy is pre-set and can generate a target fragrance combination matching the current emotion type in a personalized manner according to specific rules and fragrance data. The feature rules include the matching degree between the current emotion type and the emotion label, the emotional intensity of the current emotion type, or the driver's personal preferences. Optionally, the preset fragrance combination generation strategy can adopt a fuzzy system algorithm or other algorithms that can perform personalized matching on the emotion type and the in-stock fragrance labels.

[0055] As an example, in step S104, the vehicle-mounted controller controls the fragrance release system to work based on the target fragrance combination. In this example, the vehicle-mounted controller controls the fragrance release system to release the fragrance corresponding to the target fragrance combination based on the target fragrance combination, realizing vehicle-mounted fragrance control.

[0056] Furthermore, the vehicle-mounted controller can also obtain the environmental parameters in the cockpit, determine the fragrance dosage of the fragrance corresponding to the target fragrance combination according to the environmental parameters, and control the fragrance release system to release the fragrance corresponding to the target fragrance combination based on the target fragrance combination and the corresponding fragrance dosage. Exemplarily, the environmental parameters include the temperature in the cockpit, the air flow in the cockpit, and the humidity in the cockpit, etc. It can be understood that the vehicle-mounted controller can also control the fragrance release system to release the fragrance corresponding to the target fragrance combination according to the fragrance dosage defined by the user.

[0057] In this embodiment, by obtaining the driver's face image, using an emotion classification model constructed based on the Transformer network to perform emotion recognition on the driver's face image, determining the current emotion type corresponding to the driver's face image, ensuring the accuracy of the recognized current emotion type to the greatest extent, querying the fragrance database based on the current emotion type, determining the target fragrance combination corresponding to the current emotion type, to personalize the matching of the target fragrance combination corresponding to the current emotion type, and finally controlling the fragrance release system to work based on the target fragrance combination, so as to accurately match the target fragrance combination corresponding to the current emotion type, and further better meet the driver's emotional needs, improving the cockpit interaction effect and driving experience.

[0058] In one embodiment, as Figure 2As shown, in step S102, an emotion classification model based on the Transformer network is used to perform emotion recognition on the driver's face image to determine the current emotion type corresponding to the driver's face image, including:

[0059] S201: Segment the driver's face image to obtain at least two face image patches.

[0060] S202: Extract features from each face image patch to obtain the face feature vector corresponding to each face image patch.

[0061] S203: Use the emotion classification model based on the Transformer network to perform emotion recognition on the face feature vectors corresponding to at least two face image patches to determine the current emotion type corresponding to the driver's face image.

[0062] Among them, a face image patch refers to an image patch obtained by segmenting the driver's face image. A face feature vector refers to a vector obtained by performing vector conversion after feature extraction from a face image patch.

[0063] As an example, in step S201, the vehicle-mounted controller segments the driver's face image through the emotion classification model to obtain at least two face image patches for subsequent accurate feature extraction.

[0064] As an example, in step S202, the vehicle-mounted controller extracts features from each face image patch to obtain the face feature vector corresponding to each face image patch. Specifically, features are extracted from each face image patch to facilitate the extraction of key feature points in the driver's face image, such as the features corresponding to the eyes, eyebrows, mouth, and nose, in order to obtain the driver's emotion.

[0065] As an example, in step S203, the vehicle-mounted controller uses the emotion classification model based on the Transformer network to perform emotion recognition on the face feature vectors corresponding to at least two face image patches to determine the current emotion type corresponding to the driver's face image. Specifically, after obtaining the face feature vector corresponding to each face image patch, the relationship between the face feature vectors is calculated through the Transformer encoder in the emotion classification model, thereby performing emotion recognition to determine the current emotion type corresponding to the driver's face image.

[0066] In this embodiment, the driver's face image is segmented to obtain at least two face image blocks. Feature extraction is performed on each face image block to obtain the face feature vector corresponding to each face image block. An emotion classification model constructed based on the Transformer network is used to perform emotion recognition on the face feature vectors corresponding to at least two face image blocks, and the current emotion type corresponding to the driver's face image is determined, so as to effectively extract the current emotion type in the driver's face image and ensure the accuracy of the current emotion type. It should be noted that the emotion classification model constructed by the Transformer network enables the emotion classification model to segment the driver's face image, obtain at least two face image blocks, and linearly project each face image block into the feature space, enabling the emotion classification model to process a large amount of data in parallel, thereby improving the calculation efficiency of the current emotion type and ensuring the real-time nature of emotion analysis.

[0067] In one embodiment, as Figure 3 shown, in step S103, the fragrance database is queried based on the current emotion type to determine the target fragrance combination corresponding to the current emotion type, including:

[0068] S301: Query the fragrance database based on the current emotion type to determine whether there is a historical fragrance combination corresponding to the current emotion type.

[0069] S302: If there is a historical fragrance combination, determine the historical fragrance combination as the target fragrance combination.

[0070] S303: If there is no historical fragrance combination, use the fuzzy system algorithm to perform matching processing on the current emotion type and the fragrance labels in the library to determine the target fragrance combination corresponding to the current emotion type.

[0071] As an example, in step S301, the vehicle-mounted controller queries the fragrance database based on the current emotion type to determine whether there is a historical fragrance combination corresponding to the current emotion type, and based on the judgment result, executes step S302 or step S303.

[0072] As an example, in step S302, when there is a historical fragrance combination in the fragrance database, the vehicle-mounted controller determines the historical fragrance combination as the target fragrance combination to reduce the amount of computation and help improve the fragrance control efficiency.

[0073] As an example, in step S303, when there is no historical fragrance combination in the in-vehicle fragrance database, the fuzzy system algorithm is used to match the current emotion type with the in-stock fragrance labels to determine the target fragrance combination corresponding to the current emotion type, so as to personalized generate the target fragrance combination corresponding to the current emotion type. Exemplarily, before using the fuzzy system algorithm to match the current emotion type with the in-stock fragrance labels, the input and output variables of the fuzzy system algorithm are set according to actual experience, and a suitable fuzzy rule base is constructed; then, a suitable fuzzy algorithm is selected. Optionally, the fuzzy algorithm can be the Mamdani algorithm and the Sugeno algorithm, to match the current emotion type with the in-stock fragrance labels to determine the target fragrance combination corresponding to the current emotion type, and personalized generate the target fragrance combination corresponding to the current emotion type.

[0074] In this embodiment, the fragrance database is queried based on the current emotion type to determine whether there is a historical fragrance combination corresponding to the current emotion type. If there is a historical fragrance combination, the historical fragrance combination is determined as the target fragrance combination. If there is no historical fragrance combination, the fuzzy system algorithm is used to match the current emotion type with the in-stock fragrance labels to determine the target fragrance combination corresponding to the current emotion type. Thus, on the basis of accurately determining the current emotion type corresponding to the driver's face image through the emotion classification model constructed based on the Transformer network, when there is no historical fragrance combination, the fuzzy system algorithm is used to match the current emotion type with the in-stock fragrance labels to determine the target fragrance combination corresponding to the current emotion type, realizing personalized generation of the target fragrance combination corresponding to the current emotion type, and further better meeting the driver's emotion needs and improving the cockpit interaction effect and driving experience.

[0075] In one embodiment, as Figure 4 shown, in step S303, using the fuzzy system algorithm to match the current emotion type with the in-stock fragrance combination to determine the target fragrance combination corresponding to the current emotion type includes:

[0076] S401: Query the emotion-fragrance mapping table based on the current emotion type to determine at least two in-stock fragrance labels corresponding to the current emotion type.

[0077] S402: Based on the fuzzy system algorithm, match the in-stock fragrance combinations corresponding to at least two in-stock fragrance labels to determine the target fragrance combination corresponding to the current emotion type.

[0078] Among them, the emotion-fragrance mapping table refers to a pre-set data table, including the mapping relationship between the in-stock fragrance labels and the corresponding in-stock fragrance combinations and specific emotion types.

[0079] As an example, the in-vehicle controller queries the emotion fragrance mapping table based on the current emotion type, and determines at least two in-stock fragrance tags corresponding to the current emotion type. As shown in Table 1 below, when the current emotion type is A, at least two in-stock fragrance tags a1, a2,..., an corresponding to the current emotion type are determined. Based on the fuzzy system algorithm, matching processing is performed on the in-stock fragrance combinations corresponding to at least two in-stock fragrance tags, namely 1. (a1\a2); 2. (a1\a3);... n. (a1\a2\...\an), that is, the target fragrance combination that best matches the current emotion type is determined from the in-stock fragrance combinations corresponding to at least two in-stock fragrance tags.

[0080] Table 1 Emotion Fragrance Mapping Table

[0081]

[0082]

[0083] In this embodiment, the emotion fragrance mapping table is queried based on the current emotion type, at least two in-stock fragrance tags corresponding to the current emotion type are determined, and matching processing is performed on the in-stock fragrance combinations corresponding to at least two in-stock fragrance tags based on the fuzzy system algorithm to determine the target fragrance combination corresponding to the current emotion type. Thus, based on the fuzzy system algorithm, personalized generation of the target fragrance combination corresponding to the current emotion type is achieved, and further, the emotional needs of the driver are better met, and the interaction effect and driving experience in the cockpit are improved.

[0084] In one embodiment, as Figure 5 shown, in step S402, based on the fuzzy system algorithm, matching processing is performed on the in-stock fragrance combinations corresponding to at least two in-stock fragrance tags to determine the target fragrance combination corresponding to the current emotion type, including:

[0085] S501: Obtain the emotion matching score, emotional intensity score, and personal preference score corresponding to the current emotion type.

[0086] S502: Based on the target weights, process the emotion matching score, emotional intensity score, and personal preference score to determine the target scores corresponding to the in-stock fragrance combinations corresponding to at least two in-stock fragrance tags.

[0087] S503: Determine the in-stock fragrance combination with the maximum target score as the target fragrance combination corresponding to the current emotion type.

[0088] Among them, the emotion matching score refers to the matching score between the current emotion type and the in-stock fragrance combinations. The emotion intensity score refers to the intensity of the current emotion type. The personal preference score refers to the degree of preference of the driver for a specific in-stock fragrance combination. The target weight is the weight assigned to each in-stock fragrance label by the fuzzy system algorithm according to the emotion matching score, emotion intensity score, and personal preference score corresponding to the current emotion type. The target score refers to the matching score between the in-stock fragrance combination and the current emotion type.

[0089] As an example, in step S501, the vehicle-mounted controller obtains the emotion matching score, emotion intensity score, and personal preference score corresponding to the current emotion type, so that the vehicle-mounted controller can determine the importance of each in-stock fragrance label in meeting the current emotion type based on the fuzzy system algorithm. Exemplarily, the emotion matching score, emotion intensity score, and personal preference score between the current emotion type and each in-stock fragrance label can be obtained through a pre-trained emotion matching model. It can be understood that the emotion matching model can be constructed according to traditional machine learning methods or deep learning methods, which will not be elaborated here.

[0090] As an example, in step S502, based on the target weight, the vehicle-mounted controller processes the emotion matching score, emotion intensity score, and personal preference score to determine the target scores corresponding to the in-stock fragrance combinations corresponding to at least two in-stock fragrance labels. That is, based on the target weight of each in-stock fragrance label in meeting the current emotion type, the target scores corresponding to the in-stock fragrance combinations corresponding to at least two in-stock fragrance labels are determined. Exemplarily, the target weight includes an emotion matching weight, an emotion intensity weight, and a preference weight, and the emotion matching weight, emotion intensity weight, and preference weight are weighted and processed to determine the target scores corresponding to the in-stock fragrance combinations corresponding to at least two in-stock fragrance labels.

[0091] As an example, in step S503, the in-stock fragrance combination with the highest target score is determined as the target fragrance combination corresponding to the current emotion type, so as to accurately generate the target fragrance combination corresponding to the current emotion type, and further better meet the driver's emotional needs, improving the cockpit interaction effect and driving experience.

[0092] In this embodiment, the emotion matching score, emotion intensity score, and personal preference score corresponding to the current emotion type are obtained, and based on the target weight, the emotion matching score, emotion intensity score, and personal preference score are processed to determine the target scores corresponding to the in-stock fragrance combinations corresponding to at least two in-stock fragrance labels. The in-stock fragrance combination with the highest target score is determined as the target fragrance combination corresponding to the current emotion type, so as to accurately generate the target fragrance combination corresponding to the current emotion type, and further better meet the driver's emotional needs, improving the cockpit interaction effect and driving experience.

[0093] In one embodiment, as Figure 6 shown, in step S103, query the fragrance database based on the current emotion type to determine the target fragrance combination corresponding to the current emotion type, including:

[0094] S601: If the current emotion type is a specific emotion type, obtain the duration corresponding to the specific emotion type.

[0095] S602: If the duration is greater than the preset time, determine the warning fragrance combination as the target fragrance combination corresponding to the current emotion type.

[0096] Among them, the specific emotion type is a pre-set emotion type. The preset time is a custom-set time. The warning fragrance combination is a custom-set fragrance combination.

[0097] As an example, when the current emotion type of the vehicle-mounted controller is a specific emotion type, obtain the duration corresponding to the specific emotion type. For example, when the current emotion type of the driver is fatigue or inattentive, obtain the duration corresponding to the specific emotion type; then, compare the duration corresponding to the specific emotion type with the preset time. If the duration is greater than the preset time, it means that the driver is currently in a state of fatigue or inattentiveness and is prone to safety hazards during driving. Then, determine the warning fragrance combination as the target fragrance combination corresponding to the current emotion type. For example, the warning fragrance combination is a fragrance combination that can refresh the mind, such as a fragrance combination with a higher stimulation, so as to make the driver get out of the state of fatigue or inattentiveness and ensure the safety of the driver.

[0098] In this embodiment, if the current emotion type is a specific emotion type, obtain the duration corresponding to the specific emotion type. If the duration is greater than the preset time, determine the warning fragrance combination as the target fragrance combination corresponding to the current emotion type to improve the safety during driving.

[0099] It should be understood that the magnitudes of the sequence numbers of the steps in the above embodiments do not mean 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 to the implementation process of the embodiments of the present invention.

[0100] This embodiment provides a vehicle-mounted controller for executing the vehicle-mounted fragrance control method in the above embodiment.

[0101] This embodiment provides a fragrance control system, including a fragrance release system and the vehicle-mounted controller in the above embodiment.

[0102] This embodiment provides a vehicle, including the fragrance control system in the above embodiment.

[0103] This embodiment provides a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the above-mentioned vehicle-mounted fragrance control method.

[0104] In one embodiment, a vehicle-mounted controller is provided, and its internal structural diagram can be as Figure 7 shown. The vehicle-mounted controller includes a processor, a memory, a network interface, and a database connected through a system bus. Among them, the processor of the vehicle-mounted controller is used to provide computing and control capabilities. The memory of the vehicle-mounted controller includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program, and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the vehicle-mounted controller is used for vehicle-mounted fragrance control. The network interface of the vehicle-mounted controller is used to communicate with an external terminal through a network connection. The computer program, when executed by the processor, implements the above-mentioned vehicle-mounted fragrance control method.

[0105] In one embodiment, a vehicle-mounted controller is provided, including a memory, a processor, and a computer program stored on the memory and executable on the processor. When the processor executes the computer program, it implements the vehicle-mounted fragrance control method in the above embodiment. To avoid repetition, it will not be elaborated here.

[0106] In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, it implements the vehicle-mounted fragrance control method in the above embodiment. To avoid repetition, it will not be elaborated here.

[0107] Those of ordinary skill in the art can understand that all or part of the processes in the methods of the above embodiments can be completed by instructing relevant hardware through a computer program. The computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above methods. Among them, any reference to a memory, storage, database, or other medium used in the embodiments provided in the present application can include non-volatile and / or volatile memories. Non-volatile memories can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memories can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in many forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDR SDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), Rambus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and Rambus dynamic RAM (RDRAM), etc.

[0108] Those skilled in the art can clearly understand that, for the convenience and brevity of description, only the above division of each functional unit and module is used as an example. In actual applications, the above functions can be allocated to different functional units and modules according to needs, 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.

[0109] The above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit them; although the present invention 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 recorded 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 the embodiments of the present invention, and should all be included in the protection scope of the present invention.

Claims

1. A vehicle-mounted fragrance control method, characterized in that, Including: Obtain the driver's face image; Use an emotion classification model constructed based on the Transformer network to perform emotion recognition on the driver's face image, and determine the current emotion type corresponding to the driver's face image; Query the fragrance database based on the current emotion type, and determine the target fragrance combination corresponding to the current emotion type; Based on the target fragrance combination, control the fragrance release system to work.

2. The in-vehicle fragrance control method according to claim 1, wherein, The step of using an emotion classification model constructed based on the Transformer network to perform emotion recognition on the driver's face image and determine the current emotion type corresponding to the driver's face image includes: Segment the driver's face image to obtain at least two face image blocks; Extract features from each of the face image blocks to obtain a face feature vector corresponding to each of the face image blocks; Use an emotion classification model constructed based on the Transformer network to perform emotion recognition on the face feature vectors corresponding to at least two of the face image blocks, and determine the current emotion type corresponding to the driver's face image.

3. The in-vehicle fragrance control method according to claim 1, characterized in that, The step of querying the fragrance database based on the current emotion type and determining the target fragrance combination corresponding to the current emotion type includes: Query the fragrance database based on the current emotion type, and determine whether there is a historical fragrance combination corresponding to the current emotion type; If the historical fragrance combination exists, determine the historical fragrance combination as the target fragrance combination; If the historical fragrance combination does not exist, use the fuzzy system algorithm to perform matching processing on the current emotion type and the in-library fragrance labels, and determine the target fragrance combination corresponding to the current emotion type.

4. The in-vehicle fragrance control method according to claim 3, wherein, The step of using the fuzzy system algorithm to perform matching processing on the current emotion type and the in-library fragrance combinations and determine the target fragrance combination corresponding to the current emotion type includes: Query the emotion-fragrance mapping table based on the current emotion type, and determine at least two in-library fragrance labels corresponding to the current emotion type; Based on the fuzzy system algorithm, perform matching processing on the in-library fragrance combinations corresponding to at least two in-library fragrance labels, and determine the target fragrance combination corresponding to the current emotion type.

5. The vehicle-mounted fragrance control method according to claim 4, characterized in that, The step of using the fuzzy system algorithm to perform matching processing on the in-library fragrance combinations corresponding to at least two in-library fragrance labels and determine the target fragrance combination corresponding to the current emotion type includes: Obtain the emotion matching score, emotional intensity score, and personal preference score corresponding to the current emotion type; Based on the target weights, process the emotion matching score, the emotional intensity score, and the personal preference score to determine the target scores corresponding to the in-library fragrance combinations corresponding to at least two in-library fragrance labels; Determine the in-library fragrance combination with the maximum target score as the target fragrance combination corresponding to the current emotion type.

6. The in-vehicle fragrance control method according to claim 1, wherein, The step of querying the fragrance database based on the current emotion type and determining the target fragrance combination corresponding to the current emotion type includes: If the current emotion type is a specific emotion type, obtain the duration corresponding to the specific emotion type; If the duration is greater than a preset time, the warning fragrance combination is determined as the target fragrance combination corresponding to the current emotion type.

7. A vehicle-mounted controller, characterized in that, For implementing the vehicle-mounted fragrance control method according to any one of claims 1 to 6.

8. An aroma control system, characterized in that, Comprising a fragrance release system and a vehicle-mounted controller according to claim 7.

9. A vehicle, characterized in that, Comprising a fragrance control system according to claim 8.

10. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by a processor, it implements the vehicle-mounted fragrance control method according to any one of claims 1 to 6.

Images