An intelligent wardrobe system based on multi-modal recognition and adaptive care

Abstract

The application relates to the technical field of smart home, and discloses an intelligent wardrobe system based on multi-modal recognition and adaptive nursing, which comprises a wardrobe body, a multi-modal recognition module, a central control module, an adaptive nursing module, a storage module and a man-machine interaction module; the wardrobe body is internally provided with a mounting chamber, the multi-modal recognition module, the central control module, the adaptive nursing module and the storage module are fixedly installed in the mounting chamber, the man-machine interaction module is embedded on the outer surface of the wardrobe body, the modules are connected through wired or wireless communication lines, and multi-dimensional recognition and adaptive nursing of clothes are realized in cooperation. By installing the recognition units of the multi-modal recognition module in the wardrobe body according to a specific layout and combining the integrated structure design of the central control module, a multi-dimensional cooperative recognition structure is formed, and compared with a traditional single recognition mode, multi-source characteristic information such as the appearance, material, label and smell of clothes can be comprehensively collected.

Description

Technical Field

[0001] This invention relates to the field of home kitchen technology, specifically to an intelligent wardrobe system based on multimodal recognition and adaptive care. Background Technology

[0002] With the improvement of living standards and the increasing variety of clothing, people's needs for clothing storage, identification, and care are constantly increasing. As the core carrier of clothing storage and management, the functional upgrade of wardrobes has become an important direction for industry development. At present, wardrobe products on the market are mainly divided into two categories: traditional ordinary wardrobes and simple smart wardrobes. Traditional ordinary wardrobes only have basic clothing storage functions, with simple structures and no identification or care capabilities. The classification, identification, and care of clothing all need to be done manually, which is not only time-consuming and laborious, but also difficult to achieve targeted care based on differences in clothing materials and styles. This can easily lead to problems such as shrinkage, wrinkles, mildew, and fiber damage. Especially for delicate fabrics such as wool and silk, improper manual care can easily cause irreversible damage.

[0003] While existing simple smart wardrobes have incorporated some smart elements, such as basic temperature and humidity control and clothing barcode scanning, they still suffer from several technical shortcomings: First, the recognition methods are relatively simple, often relying solely on visual or tag recognition. This fails to comprehensively collect multi-source characteristic information such as clothing material, odor, and wear, resulting in low recognition accuracy, susceptibility to errors, and insufficient multimodal fusion depth, making it difficult to form a complete clothing feature profile and provide a comprehensive and accurate basis for subsequent care. Second, the care functions are relatively basic, mostly providing uniform care with fixed parameters, without effective linkage with the recognition function. This prevents adaptive care based on the specific characteristics of the clothing, resulting in poor targeted care and difficulty in meeting the personalized care needs of clothing of different materials and conditions, still requiring manual intervention to set care parameters. Third, the system structure design is unreasonable, with each functional module being independently set up, leading to poor connection stability, low data transmission efficiency, and a lack of reasonable installation layout and integration design. This results in a large system size, inconvenient installation and maintenance, a simple storage structure, and susceptibility to data loss and privacy leaks. The human-computer interaction methods are also relatively simple, lacking ease of use.

[0004] Furthermore, current garment care faces numerous industry pain points. High-end garments often present the dilemma of being unwashable and undryable, traditional care methods easily damage the fabric, while professional care institutions are costly and time-consuming. Simultaneously, the wide variety of garments leads to cumbersome categorization and organization, with users frequently encountering difficulties in finding and organizing their clothes. Existing wardrobe systems struggle to effectively address these pain points. Therefore, developing an intelligent wardrobe system capable of accurate multimodal garment recognition, adaptive personalized care, and with a rational structure and user-friendly operation has become an urgent technical challenge. Summary of the Invention

[0005] To address the shortcomings of existing technologies, this invention provides an intelligent wardrobe system based on multimodal recognition and adaptive care, which solves the problems mentioned in the background section.

[0006] To achieve the above objectives, the present invention provides the following technical solution: an intelligent wardrobe system based on multimodal recognition and adaptive care, comprising a wardrobe body, a multimodal recognition module, a central control module, an adaptive care module, a storage module, and a human-computer interaction module; the wardrobe body has an internal installation chamber, in which the multimodal recognition module, central control module, adaptive care module, and storage module are all fixedly installed; the human-computer interaction module is embedded in the outer surface of the wardrobe body; the modules are connected via wired or wireless communication lines to collaboratively achieve multi-dimensional recognition and adaptive care of clothing; the multimodal recognition module is used to collect and process multi-source feature information of clothing; the central control module is used to receive the recognition results from the multimodal recognition module, generate adaptive care instructions, and send them to the adaptive care module; the adaptive care module performs targeted care operations on the clothing according to the care instructions; the storage module is used to store recognition data, care parameters, and user preference information; and the human-computer interaction module is used to realize information interaction between the user and the system.

[0007] Preferably, the multimodal recognition module includes a visual recognition unit, a material recognition unit, a label recognition unit, and an odor recognition unit, each of which is electrically connected to the central control module. The visual recognition unit includes a high-definition camera and a recognition circuit board. The high-definition camera is embedded in the top and sides of the wardrobe body, and the recognition circuit board is fixed in the installation cavity. The high-definition camera is electrically connected to the recognition circuit board and is used to capture images of the appearance of the clothing and transmit them to the recognition circuit board for recognition processing, identifying the color, style, pattern, wear level, and wrinkle level of the clothing. The material recognition unit includes an infrared spectral sensor, which is evenly distributed below the hanging rod inside the wardrobe body and electrically connected to the central control module. By detecting the infrared absorption spectrum of the clothing fibers, it identifies the material type of the clothing, including cotton, linen, silk, wool, chemical fibers, and blended materials.

[0008] Preferably, the tag identification unit includes an RFID reader and an image recognition component. The RFID reader is installed at the end of the hanging rod inside the wardrobe body. The image recognition component shares a data acquisition end with the high-definition camera of the visual recognition unit. Both the RFID reader and the image recognition component are electrically connected to the central control module, forming a dual identification structure. The RFID reader is used to read the clothing information stored in the clothing tag, including clothing size, washing instructions, care contraindications, and manufacturing information. The image recognition component is used to identify the printed label on the surface of the clothing. When the RFID tag cannot be read, the text information of the label is extracted through image recognition and processed by OCR recognition to ensure the complete acquisition of the label information. The odor identification unit includes a gas sensor array, which is installed at the bottom inside the wardrobe body and electrically connected to the central control module. It is used to detect the odor components and concentration on the surface of the clothing, identify the odor type of the clothing, and quantify the odor level, providing a basis for care operations.

[0009] Preferably, the central control module includes a main control circuit board, a data processing unit, and an instruction generation unit. The main control circuit board is fixedly installed in the middle of the installation chamber, and the data processing unit and the instruction generation unit are integrated on the main control circuit board and electrically connected. The data processing unit is used to fuse multi-source feature information collected by the multimodal recognition module, and fuses visual, material, label, and odor features through an attention mechanism, removes redundant information, and generates a comprehensive clothing feature profile. The instruction generation unit generates personalized adaptive nursing instructions based on the comprehensive clothing feature profile, historical nursing data in the storage module, and user preferences, combined with preset nursing rules. The nursing instructions include nursing mode, nursing parameters, and nursing duration.

[0010] Preferably, the adaptive care module includes a temperature and humidity control unit, a sterilization and deodorization unit, a wrinkle removal and shaping unit, and a dust protection unit. Each unit is fixedly installed inside the wardrobe body and electrically connected to the central control module, and can operate independently or in conjunction with each other. The temperature and humidity control unit includes a constant temperature and humidity sensor and an intelligent air conditioning component. The constant temperature and humidity sensor is evenly distributed in various areas inside the wardrobe body, and the intelligent air conditioning component is installed on one side of the installation chamber. The constant temperature and humidity sensor is electrically connected to the intelligent air conditioning component to detect the temperature and humidity inside the wardrobe body in real time and adjust the temperature and humidity to a preset range according to the material of the clothing and storage requirements.

[0011] Preferably, the sterilization and deodorization unit includes an ultraviolet germicidal lamp and an activated carbon filter assembly. The ultraviolet germicidal lamp is installed at the top inside the wardrobe body and is evenly arranged along the length of the hanging rod. The activated carbon filter assembly is installed at the ventilation opening of the wardrobe body and is integrated with the ventilation structure. The wrinkle removal and shaping unit includes a low-temperature dry steam generating assembly and an intelligent ironing mechanism. The low-temperature dry steam generating assembly is installed in the installation chamber. The intelligent ironing mechanism is slidably installed on the top guide rail inside the wardrobe body. The low-temperature dry steam generating assembly and the intelligent ironing mechanism are connected by pipes. The intelligent ironing mechanism can automatically adjust the steam volume, steam temperature and ironing speed according to the wrinkle level and material of the clothing to avoid high temperature damage to the clothing fibers.

[0012] Preferably, the storage module includes a local storage unit and a cloud storage unit. The local storage unit is fixedly installed on one side of the main control circuit board and electrically connected to the central control module. The cloud storage unit establishes a communication connection with the local storage unit through a network module, which is integrated on the main control circuit board. The local storage unit is used to store real-time identification data, immediate nursing parameters, and frequently used user preferences. The cloud storage unit is used to store historical identification data, historical nursing records, clothing files, and system upgrade packages, supports data backup and recovery, and has data encryption functionality.

[0013] Preferably, the human-computer interaction module includes a touch screen, a voice interaction unit, and a mobile terminal APP. The touch screen is embedded in the outer surface of the wardrobe body and electrically connected to the central control module. It is used to display clothing recognition results, care status, temperature and humidity information, and operation menus, and supports users to manually set care parameters and modify user preferences. The voice interaction unit includes a noise-canceling microphone and a voice processing module. The noise-canceling microphone is embedded in one side of the touch screen, and the voice processing module is integrated on the main control circuit board. The two are electrically connected and are used to receive user voice commands and transmit them to the central control module. The mobile terminal APP establishes a connection with the system through a wireless network to realize remote viewing, remote control, and parameter setting.

[0014] Preferably, the system also includes a clothing positioning module and an intelligent sorting module. Both modules are installed inside the wardrobe body and are electrically connected to the central control module. The clothing positioning module includes a UWB positioning transmitter and a UWB positioning receiver. Each UWB transmitter corresponds to a piece of clothing, and the UWB positioning receiver is installed in a corner inside the wardrobe body to locate the position of the clothing in the wardrobe body in real time. Users can query the position of clothing and control the wardrobe to push the corresponding clothing through the human-computer interaction module. The intelligent sorting module includes a robotic arm and a sorting guide rail. The sorting guide rail is set along the width direction inside the wardrobe body, and the robotic arm can be slidably installed on the sorting guide rail. Based on the recognition results of the multimodal recognition module, the robotic arm automatically sorts and stores the clothing according to material, style, season, or wearing frequency.

[0015] Preferably, the central control module has self-learning capabilities, optimizing nursing parameters and recognition models by analyzing user operating habits, clothing care feedback, and environmental change data, and adaptively adjusting nursing strategies; the system supports firmware upgrades, automatically downloading upgrade packages through the cloud storage unit to achieve iterative updates of functions; the wardrobe body is made of moisture-proof, insect-proof, and fire-resistant materials in one piece, with removable partitions and retractable hanging rods inside, both of which are connected to the wardrobe body through a snap-fit ​​structure to adapt to the storage needs of different types of clothing; the wardrobe body door panel is equipped with a sealing strip to achieve a sealed and dustproof effect.

[0016] This invention provides an intelligent wardrobe system based on multimodal recognition and adaptive care. It has the following beneficial effects:

[0017] 1. This invention installs the various identification units of the multimodal recognition module in a specific layout inside the wardrobe body, combined with the integrated structural design of the central control module, to form a multi-dimensional collaborative recognition structure. Compared with the traditional single recognition method, it can comprehensively collect multi-source feature information such as clothing appearance, material, label and smell. Moreover, each module is stably connected through wired or wireless communication lines to ensure the accuracy and timeliness of recognition data transmission. At the same time, it simplifies the internal structural layout of the system, reduces the difficulty of module installation and maintenance, and improves the overall stability and reliability of the system.

[0018] 2. This invention forms a linked structure between the various nursing units of the adaptive nursing module, the multimodal recognition module, and the central control module. Combined with the internal partitions, retractable hanging rods, and sealing structure design of the wardrobe body, it can perform targeted nursing operations such as temperature and humidity adjustment, sterilization and deodorization, and wrinkle removal and shaping based on the clothing feature information obtained by the recognition module. This achieves intelligent and personalized clothing care, while optimizing the utilization rate of the wardrobe's internal space. It can adapt to the storage and care needs of different types of clothing, reduce manual intervention in the clothing care process, and improve the convenience of clothing care.

[0019] 3. This invention achieves secure storage and convenient retrieval of identification data, nursing parameters, and user preference information through a dual storage structure integrating local and cloud storage modules, combined with a multi-mode interactive design of the human-computer interaction module. Users can control the system and query information through touch, voice, and mobile terminal APP. At the same time, the self-learning function of the central control module and the system firmware upgrade design enable the system to adaptively adjust nursing strategies, achieve functional iteration, extend the system's service life, improve the system's adaptability and practicality, and meet the personalized usage needs of different users. Attached Figure Description

[0020] Figure 1 This is a flowchart of the present invention. Detailed Implementation

[0021] The following will refer to the appendix to the specification of this invention. Figure 1 The technical solutions in the embodiments of the present invention are clearly and completely described herein. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0022] Please see the appendix Figure 1 This invention provides an intelligent wardrobe system based on multimodal recognition and adaptive care, including a wardrobe body, a multimodal recognition module, a central control module, an adaptive care module, a storage module, a human-computer interaction module, a clothing positioning module, and an intelligent sorting module. The modules work together to realize multimodal recognition, adaptive care, intelligent classification, and location search of clothing. The overall structure is reasonably designed, easy to install, and stable in operation.

[0023] Specifically, the wardrobe body is made of a single piece of moisture-proof, insect-proof, and fire-retardant material, with an overall rectangular structure. Its front features an openable door panel with a sealing strip along the inner edge. When closed, the door panel fits tightly against the wardrobe body, achieving a dust-proof seal and preventing external dust and moisture from entering and damaging clothing. The wardrobe body also has an independent installation chamber located on the inner back side, isolated from the clothing storage area. This prevents heat and noise generated during module operation from affecting clothing storage and facilitates module installation, inspection, and maintenance. The wardrobe's interior storage area features removable partitions and a retractable hanging rod. The partitions connect to the wardrobe's inner wall via snap-fit ​​mechanisms, allowing for flexible adjustment of their position to accommodate different garment heights and sizes, creating storage spaces of varying sizes. The retractable hanging rod, made of stainless steel, connects to the wardrobe's inner wall at both ends via sliding connectors. It can be adjusted to accommodate different garment types, such as long coats and short tops, based on the quantity and length of clothing. The rod's surface features anti-slip textures to prevent garments from slipping.

[0024] The multimodal recognition module is installed inside the wardrobe body and electrically connected to the central control module. It is used to collect multi-source feature information of clothing and perform preliminary recognition processing. It includes a visual recognition unit, a material recognition unit, a label recognition unit, and an odor recognition unit. Each recognition unit is installed in a specific layout to ensure that the recognition range covers the entire clothing storage area without any blind spots. Among them, the visual recognition unit consists of high-definition cameras and recognition circuit boards. A total of four high-definition cameras are set, which are respectively embedded in the top center and the middle of the inner walls on both sides of the wardrobe body. The camera lenses face the clothing storage area and can collect images of the appearance of the clothing from all directions. The high-definition cameras are electrically connected to the recognition circuit board fixedly installed in the installation cavity through wires. The recognition circuit board integrates an image processing chip to receive the image data transmitted by the cameras and perform preliminary recognition of the color, style, pattern, wear degree, and wrinkle level of the clothing. After the recognition is completed, the data is transmitted to the central control module. The material identification unit uses infrared spectral sensors, with a total of 6 sensors evenly distributed below the hanging rods inside the wardrobe. Each pair of sensors corresponds to a group of hanging rod areas. The sensor's detection end faces the clothes hanging on the hanging rod. By emitting infrared light and receiving the infrared absorption spectrum of the clothing fibers, the composition of the clothing fibers is analyzed, thereby identifying the material type of the clothing, including cotton, linen, silk, wool, chemical fibers, and mixed materials. The identification data is transmitted to the central control module in real time.

[0025] The tag identification unit consists of an RFID reader and an image recognition component. The RFID reader is installed at both ends of the hanging rod inside the wardrobe body, with one RFID reader for each hanging rod. Its reading range covers the entire hanging rod area and is used to read the clothing information stored in the clothing tag, including clothing size, washing instructions, care precautions, and manufacturing information. The image recognition component shares the same acquisition end as the high-definition camera of the vision recognition unit, eliminating the need for additional acquisition equipment. The image recognition component is integrated on the recognition circuit board and is used to identify printed labels on the surface of the clothing. When the RFID tag on the clothing is damaged, detached, or cannot be read, the high-definition camera captures an image of the printed label on the surface of the clothing. The image recognition component extracts the text information on the label and performs OCR recognition processing, converting the recognized text information into a digital signal and transmitting it to the central control module. This ensures the complete acquisition of the tag information, forming a dual recognition structure and improving the reliability of tag identification. The odor recognition unit uses a gas sensor array, which is installed in the bottom center of the wardrobe body. The sensor array consists of 8 different types of gas sensors, which can detect the components and concentrations of different odors such as sweat, mildew, and cooking odor, and quantify the odor level (divided into 1-5 levels, with level 1 being no odor and level 5 being a severe odor). The detection data is transmitted to the central control module to provide a basis for subsequent care operations.

[0026] The central control module, serving as the core control unit of the system, is fixedly installed in the center of the installation chamber. It includes a main control circuit board, a data processing unit, and an instruction generation unit. The data processing unit and instruction generation unit are integrated on the main control circuit board and electrically connected via circuitry. The main control circuit board is connected to the inner wall of the installation chamber via a fixed bracket to ensure stable installation and prevent vibrations during system operation from affecting module operation. The data processing unit receives recognition data transmitted from various units of the multimodal recognition module. It uses an attention mechanism to fuse visual, material, label, and odor features, eliminating redundant information and integrating them to form a complete comprehensive garment feature profile. Each garment corresponds to a unique feature profile containing all characteristic information such as appearance, material, label, and odor. The instruction generation unit generates personalized adaptive care instructions based on the comprehensive garment feature profile, historical care data stored in the module, and user preferences, combined with preset care rules. These instructions explicitly include the care mode (e.g., daily care, deep care, emergency deodorization), care parameters (e.g., temperature and humidity values, steam temperature, sterilization time), and care duration. After generation, the instructions are transmitted to the adaptive care module via a wired connection.

[0027] The adaptive care module is installed inside the wardrobe body and electrically connected to the central control module. It can perform targeted care operations on clothing according to care instructions sent by the central control module. It includes a temperature and humidity control unit, a sterilization and deodorization unit, a wrinkle removal and shaping unit, and a dust protection unit. Each unit operates independently but can also work collaboratively, covering all aspects of clothing care. The temperature and humidity control unit consists of constant temperature and humidity sensors and an intelligent air conditioning component. Four constant temperature and humidity sensors are installed in the four corners inside the wardrobe body, respectively, to detect temperature and humidity data in different areas of the wardrobe body in real time. The detected data is transmitted to the intelligent air conditioning component, which is installed on one side of the installation chamber and connected to the ventilation opening of the wardrobe body. Based on the detected temperature and humidity data and the preset temperature and humidity range corresponding to the clothing material, the intelligent air conditioning component automatically adjusts the cooling, heating, or dehumidifying mode to ensure that the temperature and humidity inside the wardrobe body are maintained within a suitable range, adapting to the storage needs of different clothing materials.

[0028] The sterilization and deodorization unit consists of ultraviolet germicidal lamps and activated carbon filters. The ultraviolet germicidal lamps are installed at the top inside the wardrobe body, evenly arranged along the length of the hanging rods, with 2-3 lamps corresponding to each rod. The lamps are directed towards the clothes, providing comprehensive coverage of the clothing storage area. Upon receiving a care instruction, the ultraviolet germicidal lamps activate to kill bacteria and mold on the surface of the clothes. The activated carbon filter is installed at the ventilation opening of the wardrobe body, integrated with the ventilation structure. The ventilation structure consists of a fan and ventilation ducts. The fan is installed in the installation chamber and connected to the activated carbon filter. When the fan is activated, it circulates the air inside the wardrobe. As the air passes through the activated carbon filter, odor molecules and dust are adsorbed, achieving deodorization and dust prevention. At the same time, the activated carbon filter can be automatically regenerated periodically to extend its service life. The regeneration process is automatically controlled by the central control module, requiring no manual intervention. The wrinkle-removing and shaping unit consists of a low-temperature dry steam generator and an intelligent steaming mechanism. The low-temperature dry steam generator is installed in the mounting chamber and connected to the intelligent steaming mechanism via pipes. The intelligent steaming mechanism is slidably mounted on the top guide rail inside the wardrobe body via a sliding component. The guide rail is set along the length of the hanging rod, allowing the intelligent steaming mechanism to move along the guide rail and cover the entire hanging rod area. Upon receiving a wrinkle-removing command, the low-temperature dry steam generator produces low-temperature dry steam, which is delivered to the intelligent steaming mechanism through pipes. The intelligent steaming mechanism automatically adjusts the steam volume, steam temperature, and steaming speed according to the wrinkle level and material of the garment, moving slowly to iron the garment, avoiding high-temperature damage to the clothing fibers, and achieving wrinkle removal and shaping. The dustproof protection unit works in conjunction with the wardrobe door panel and sealing strip. After the door panel is closed, the sealing strip seals the gap between the door panel and the wardrobe body, while the dustproof protection unit activates, using a fan to adsorb dust inside the wardrobe to the activated carbon filter component, reducing dust adhesion to the surface of the clothing and achieving dust protection.

[0029] The storage module includes a local storage unit and a cloud storage unit. The local storage unit is fixedly installed on one side of the main control circuit board and electrically connected to the central control module. It adopts an embedded storage structure, is small in size, and easy to install. It stores real-time identification data, immediate care parameters, and frequently used user preferences, ensuring rapid data retrieval. The cloud storage unit establishes a communication connection with the local storage unit through a network module integrated on the main control circuit board. The network module supports both wired and wireless network connections, which can be flexibly switched according to the usage environment. The cloud storage unit stores historical identification data, historical care records, comprehensive clothing feature files, and system upgrade packages. It supports automatic data backup and recovery and also has data encryption capabilities. Encryption algorithms are used to encrypt user privacy information and clothing data to prevent data leakage. The local storage unit and the cloud storage unit synchronize data in real time to ensure data consistency. When the local storage unit fails, data can be recovered through the cloud storage unit to avoid data loss.

[0030] The human-computer interaction module is used to realize information interaction between the user and the system. It includes a touch screen display, a voice interaction unit, and a mobile terminal APP. The touch screen display is embedded in the outer surface of the wardrobe body, located on one side of the door panel, and electrically connected to the central control module. The display screen is a high-definition touch screen with a waterproof and dustproof coating. It can display clothing recognition results, care status, temperature and humidity information, and operation menus. Users can manually set care parameters, modify user preferences, query clothing information, and control the start and stop of care operations through the touch screen display. The voice interaction unit consists of a noise-canceling microphone and a voice processing module. The noise-canceling microphone is embedded in one side of the touch screen display and is used to receive user voice commands. The noise-canceling microphone has noise reduction capabilities. The noise reduction function filters out ambient noise, improving the accuracy of voice recognition. The voice processing module is integrated on the main control circuit board and electrically connected to the noise-canceling microphone. It converts received voice commands into digital signals and transmits them to the central control module for voice control. Users can control the system to start or stop care, check the location of clothing, and perform other operations via voice commands. The mobile terminal APP can be downloaded and installed on smartphones, tablets, and other mobile devices. It connects to the system via a wireless network, allowing users to remotely view the temperature and humidity inside the wardrobe, the clothing care status, and the clothing recognition results. Users can also remotely control the start and stop of care operations, set user preferences, and view historical care records for remote management.

[0031] Both the clothing positioning module and the intelligent sorting module are installed inside the wardrobe body and are electrically connected to the central control module, further enhancing the system's intelligence. The clothing positioning module consists of a UWB positioning transmitter and a UWB positioning receiver. The UWB positioning transmitter is a small tag structure that can be fixed to the collar or corner of the clothing, corresponding one-to-one with each garment, and each transmitter has a unique positioning code. The UWB positioning receivers are installed in the four corners inside the wardrobe body and are electrically connected to the central control module. By receiving the positioning signals sent by the UWB positioning transmitters, they can obtain the specific location of the clothing in the wardrobe body in real time. The positioning data is transmitted to the central control module. Users can query the location of the clothing through a touch screen, voice, or mobile terminal APP. The central control module controls the pushing mechanism inside the wardrobe (the pushing mechanism consists of a small motor and a pushing plate, installed in each partition storage area) based on the positioning data to push the corresponding clothing to an easily accessible location, solving the problem of inconvenient clothing retrieval. The intelligent sorting module consists of a robotic arm and a sorting guide rail. The sorting guide rail is set along the width of the wardrobe body and installed at the top of the clothing storage area. The robotic arm can be slidably installed on the sorting guide rail via a sliding connector. The end of the robotic arm is equipped with a flexible clamping structure to avoid damaging the clothing during clamping. The central control module controls the robotic arm to move along the sorting guide rail according to the recognition results of the multimodal recognition module, clamping the clothing and transporting it to the corresponding classification area, realizing automatic classification and storage of clothing and reducing the workload of manual sorting.

[0032] In this embodiment, the central control module also has self-learning capabilities. By analyzing user operating habits, clothing care feedback, and environmental change data, it continuously optimizes nursing parameters and recognition models, and adaptively adjusts nursing strategies to make the nursing effect more in line with user needs. At the same time, the system supports firmware upgrades. The central control module periodically checks for new upgrade packages through the cloud storage unit. If a new upgrade package is available, it automatically downloads and completes the upgrade, realizing iterative updates of system functions without the need for manual upgrades.

[0033] The working process of this invention is as follows: The user places clothing inside the wardrobe body, hangs the clothing on a hanging rod or places it on a partition, and the UWB positioning transmitter on the clothing establishes a connection with the UWB positioning receiver to complete the clothing positioning initialization; the multimodal recognition module is activated, the visual recognition unit acquires an image of the clothing appearance, the material recognition unit detects the clothing material, the label recognition unit reads the clothing label information, and the odor recognition unit detects the clothing odor. Each recognition unit transmits the recognition data to the central control module; the data processing unit of the central control module fuses the multi-source recognition data to generate a comprehensive clothing feature profile, which is stored in the storage module; the instruction generation unit, based on the comprehensive clothing feature profile, historical care data, and user... Based on user preferences, adaptive care instructions are generated and sent to the adaptive care module. The adaptive care module then activates the corresponding care unit to perform care operations such as temperature and humidity adjustment, sterilization and deodorization, and wrinkle removal and shaping on the clothing. During the care process, the human-machine interface module displays the care status in real time, and users can view and control the care operations via touch, voice, or a mobile app. After care is completed, the central control module stores the care record in the storage module, while the intelligent sorting module automatically categorizes and stores the clothing based on the recognition results. When a user needs to retrieve clothing, they send a query command through the human-machine interface module. The central control module then controls the push mechanism to push the clothing based on the location data from the clothing positioning module, making it convenient for the user to retrieve.

[0034] The above description is merely a preferred embodiment of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. An intelligent wardrobe system based on multimodal recognition and adaptive care, characterized in that, The system includes a wardrobe body, a multimodal recognition module, a central control module, an adaptive care module, a storage module, and a human-computer interaction module. The wardrobe body has an internal installation chamber where the multimodal recognition module, central control module, adaptive care module, and storage module are fixedly installed. The human-computer interaction module is embedded in the outer surface of the wardrobe body. All modules are connected via wired or wireless communication lines to collaboratively achieve multi-dimensional recognition and adaptive care of clothing. The multimodal recognition module collects and processes multi-source feature information from the clothing. The central control module receives the recognition results from the multimodal recognition module, generates adaptive care instructions, and sends them to the adaptive care module. The adaptive care module performs targeted care operations on the clothing according to the instructions. The storage module stores recognition data, care parameters, and user preference information. The human-computer interaction module enables information interaction between the user and the system.

2. The intelligent wardrobe system according to claim 1, characterized in that, The multimodal recognition module includes a visual recognition unit, a material recognition unit, a label recognition unit, and an odor recognition unit, each of which is electrically connected to the central control module. The visual recognition unit includes a high-definition camera and a recognition circuit board. The high-definition camera is embedded in the top and sides of the wardrobe body, and the recognition circuit board is fixed in the installation cavity. The high-definition camera is electrically connected to the recognition circuit board and is used to capture images of the appearance of the clothing and transmit them to the recognition circuit board for recognition processing, identifying the color, style, pattern, wear level, and wrinkle level of the clothing. The material recognition unit includes an infrared spectral sensor, which is evenly distributed below the hanging rod inside the wardrobe body and electrically connected to the central control module. By detecting the infrared absorption spectrum of the clothing fibers, it identifies the material type of the clothing, including cotton, linen, silk, wool, chemical fibers, and blended materials.

3. The intelligent wardrobe system according to claim 2, characterized in that, The tag identification unit includes an RFID reader and an image recognition component. The RFID reader is installed at the end of the hanging rod inside the wardrobe body. The image recognition component shares a data acquisition end with the high-definition camera of the visual recognition unit. Both the RFID reader and the image recognition component are electrically connected to the central control module, forming a dual identification structure. The RFID reader is used to read the clothing information stored in the clothing tag, including clothing size, washing instructions, care contraindications, and manufacturing information. The image recognition component is used to identify the printed label on the surface of the clothing. When the RFID tag cannot be read, the text information of the label is extracted through image recognition and processed by OCR recognition to ensure complete acquisition of the label information. The odor identification unit includes a gas sensor array, which is installed at the bottom inside the wardrobe body and electrically connected to the central control module. It is used to detect the odor components and concentration on the surface of the clothing, identify the odor type of the clothing, and quantify the odor level, providing a basis for care operations.

4. The intelligent wardrobe system according to claim 1, characterized in that, The central control module includes a main control circuit board, a data processing unit, and an instruction generation unit. The main control circuit board is fixedly installed in the middle of the installation chamber. The data processing unit and the instruction generation unit are integrated on the main control circuit board and are electrically connected. The data processing unit is used to fuse multi-source feature information collected by the multimodal recognition module. It fuses visual, material, label, and odor features through an attention mechanism, removes redundant information, and generates a comprehensive feature profile of the clothing. The instruction generation unit generates personalized adaptive nursing instructions based on the comprehensive feature profile of the clothing, historical nursing data in the storage module, and user preferences, combined with preset nursing rules. The nursing instructions include nursing mode, nursing parameters, and nursing duration.

5. The intelligent wardrobe system according to claim 1, characterized in that, The adaptive care module includes a temperature and humidity control unit, a sterilization and deodorization unit, a wrinkle removal and shaping unit, and a dust protection unit. Each unit is fixedly installed inside the wardrobe body and is electrically connected to the central control module, and can operate independently or in conjunction with each other. The temperature and humidity control unit includes a constant temperature and humidity sensor and an intelligent air conditioning component. The constant temperature and humidity sensor is evenly distributed in various areas inside the wardrobe body, and the intelligent air conditioning component is installed on one side of the installation chamber. The constant temperature and humidity sensor is electrically connected to the intelligent air conditioning component to detect the temperature and humidity inside the wardrobe body in real time and adjust the temperature and humidity to a preset range according to the material of the clothing and storage requirements.

6. The intelligent wardrobe system according to claim 5, characterized in that, The sterilization and deodorization unit includes an ultraviolet germicidal lamp and an activated carbon filter assembly. The ultraviolet germicidal lamp is installed at the top inside the wardrobe body and is evenly arranged along the length of the hanging rod. The activated carbon filter assembly is installed at the ventilation opening of the wardrobe body and is integrated with the ventilation structure. The wrinkle removal and shaping unit includes a low-temperature dry steam generator assembly and an intelligent ironing mechanism. The low-temperature dry steam generator assembly is installed in the installation chamber. The intelligent ironing mechanism is slidably installed on the top guide rail inside the wardrobe body. The low-temperature dry steam generator assembly and the intelligent ironing mechanism are connected by pipes. The intelligent ironing mechanism can automatically adjust the steam volume, steam temperature and ironing speed according to the wrinkle level and material of the clothing to avoid high temperature damage to the clothing fibers.

7. The intelligent wardrobe system according to claim 1, characterized in that, The storage module includes a local storage unit and a cloud storage unit. The local storage unit is fixedly installed on one side of the main control circuit board and electrically connected to the central control module. The cloud storage unit establishes a communication connection with the local storage unit through a network module integrated on the main control circuit board. The local storage unit is used to store real-time identification data, immediate nursing parameters, and frequently used user preferences. The cloud storage unit is used to store historical identification data, historical nursing records, clothing files, and system upgrade packages, supports data backup and recovery, and has data encryption functionality.

8. The intelligent wardrobe system according to claim 1, characterized in that, The human-computer interaction module includes a touch screen, a voice interaction unit, and a mobile terminal APP. The touch screen is embedded in the outer surface of the wardrobe body and electrically connected to the central control module. It is used to display clothing recognition results, care status, temperature and humidity information, and operation menus, and supports users to manually set care parameters and modify user preferences. The voice interaction unit includes a noise-canceling microphone and a voice processing module. The noise-canceling microphone is embedded in one side of the touch screen, and the voice processing module is integrated on the main control circuit board. The two are electrically connected and are used to receive user voice commands and transmit them to the central control module. The mobile terminal APP establishes a connection with the system through a wireless network to realize remote viewing, remote control, and parameter setting.

9. The intelligent wardrobe system according to claim 1, characterized in that, It also includes a clothing positioning module and an intelligent sorting module, both installed inside the wardrobe body and electrically connected to the central control module. The clothing positioning module includes a UWB positioning transmitter and a UWB positioning receiver. The UWB positioning transmitter corresponds to a single garment, and the UWB positioning receiver is installed in a corner inside the wardrobe body to locate the position of the garment in real time. Users can query the position of the garment and control the wardrobe to push the corresponding garment through the human-computer interaction module. The intelligent sorting module includes a robotic arm and a sorting guide rail. The sorting guide rail is set along the width direction inside the wardrobe body, and the robotic arm can be slidably installed on the sorting guide rail. Based on the recognition results of the multimodal recognition module, the clothing is automatically classified and stored according to material, style, season, or wearing frequency.

10. The intelligent wardrobe system according to any one of claims 1-9, characterized in that, The central control module has self-learning capabilities. By analyzing user operating habits, clothing care feedback, and environmental change data, it optimizes care parameters and recognition models, and adaptively adjusts care strategies. The system supports firmware upgrades, automatically downloading upgrade packages through the cloud storage unit to achieve iterative updates. The wardrobe body is made of moisture-proof, insect-proof, and fire-resistant materials in one piece. It has removable partitions and retractable hanging rods inside, which are connected to the wardrobe body through a snap-fit ​​structure to accommodate the storage needs of different types of clothing. The wardrobe door panel has a sealing strip to achieve a sealed and dustproof environment.

Images