A multi-dimensional perception method for laundry care and a laundry care system

Abstract

This invention provides a multi-dimensional sensing method and system for clothing care. The method includes: S1, acquiring first-dimensional information; S2, acquiring second-dimensional information; S3, generating sensing data within the care space based on the first-dimensional information and the second-dimensional information; S4, generating at least one strategy among environmental control strategy, motion control strategy, care medium control strategy, and prompting strategy based on the sensing data. This multi-dimensional sensing method and system for clothing care, by acquiring image information and physical measurement information respectively, and combining information from different dimensions, supplements and corrects the sensing information within the care space, thereby obtaining more accurate sensing information. This provides support for intelligent adjustment of the care mode, improving the intelligence and precision of the care process.

Description

Technical Field

[0001] This invention relates to the field of household appliance technology, and in particular to a multi-dimensional sensing method and clothing care system for clothing care. Background Technology

[0002] With the improvement of living standards and the increase in personalized needs, consumers have higher and higher requirements for clothing care. Garment care machines, represented by garment steamers, are common small household appliances that use high-pressure steam generated by internal heating to continuously contact and soften clothing fibers, thereby smoothing out wrinkles and restoring the garment to its smooth state.

[0003] To enhance user experience, garment care machines often offer multiple care modes. Different garments can be cared for using the appropriate mode based on their specific needs.

[0004] In implementing the existing technology, the inventors discovered that in order to achieve intelligent care, the garment care machine needs to sense the interior of the care chamber in order to adjust the care mode accordingly.

[0005] Existing technologies often rely on image acquisition of the nursing compartment for perception, but this technology has poor perception accuracy and is easily affected by noise in the acquired signal, resulting in a low level of intelligence and precision in the nursing process. Summary of the Invention

[0006] In view of this, the present invention aims to propose a multi-dimensional sensing method and system for clothing care, in order to solve the problems of poor sensing accuracy and susceptibility to noise interference in the existing clothing care technology, resulting in low intelligence and precision in the care process.

[0007] To achieve the above objectives, the technical solution of the present invention is implemented as follows:

[0008] A multi-dimensional sensing method for clothing care includes: S1, acquiring first-dimensional information; S2, acquiring second-dimensional information; S3, generating sensing data within a care space based on the first-dimensional information and the second-dimensional information; S4, generating at least one of an environmental control strategy, a motion control strategy, a care medium control strategy, and a prompting strategy based on the sensing data.

[0009] Furthermore, the first dimension information is image information within the nursing space, and the second dimension information is physical measurement information within the nursing space, including at least one of temperature information, humidity information, size information, motion information, location information, and object recognition information.

[0010] Further, step S3 includes: S31, identifying the first attribute information of at least one of the environmental objects, nursing objects, and execution objects in the nursing space based on the first dimension information; S32, identifying the second attribute information of at least one of the environmental objects, nursing objects, and execution objects in the nursing space based on the second dimension information; S33, combining the first attribute information and the second attribute information to perform information correction and obtain the perception data in the nursing space.

[0011] Furthermore, the environmental object is the care space of the clothing care system; the care object is the item to be cared for; and the execution object is the care arm.

[0012] Further, step S33 includes any one of methods one to four, or step S33 includes any one of methods one to four combined with at least one of the others; wherein, method one is: using the first attribute information as the basic information, obtaining the attention range, using the second attribute information to perform size measurement within the attention range, correcting the first attribute information, and obtaining the perception data within the nursing space; method two is: using the second attribute information as the basic information, using the first attribute information to determine whether the data in the second attribute information is within a reasonable range; if so, then the data in the second attribute information is not corrected, and the data in the second attribute information is used as the corresponding perception data, if If not, then the corresponding data in the first attribute information is used for correction, and the corrected data is used as the corresponding perceived data; Method 3 is: statistically analyze the first attribute information and the corresponding second attribute information to obtain the average error statistics of the attribute information, and fit the correction algorithm according to the average error statistics; substitute the first attribute information and the second attribute information obtained in the actual clothing care process into the correction algorithm to obtain the perceived data in the care space; Scheme 4 is: acquire multiple sets of attribute information for training and obtaining an error prediction model; import the first attribute information and / or the second attribute information obtained in the actual clothing care process into the error prediction model to obtain the perceived data in the care space.

[0013] Furthermore, the perceived data includes at least information on clothing material, clothing size, and foreign objects in the environment.

[0014] Furthermore, step S4 includes: generating an environmental control strategy for controlling environmental parameters within the nursing space based at least on clothing material information; generating a motion control strategy for controlling the stroke of the nursing arm based at least on clothing size information; generating a nursing media control strategy for controlling nursing media parameters based at least on clothing size information and / or material information; and generating a prompting strategy based at least on foreign object information in the environment.

[0015] Furthermore, the environmental parameters include humidity and / or temperature; the nursing arm stroke includes the clamping stroke and ironing stroke of the nursing arm; the nursing medium parameters include the supply volume of the nursing medium per unit time, the temperature of the nursing medium, and the supply pressure of the nursing medium, wherein the nursing medium includes steam and hot air.

[0016] A garment care system employs the aforementioned multi-dimensional sensing method for garment care.

[0017] Compared with existing technologies, the multi-dimensional sensing method and clothing care system for clothing care described in this invention have the following advantages:

[0018] The present invention discloses a multi-dimensional sensing method and system for clothing care. By acquiring image information and physical measurement information respectively, and combining information from different dimensions, the sensing information in the care space is supplemented and corrected to obtain more accurate sensing information, thereby providing support for intelligent adjustment of the care mode and improving the intelligence and precision of the care process. Detailed Implementation

[0019] The inventive concepts of this disclosure will be described below using terminology commonly used by those skilled in the art to communicate the essence of their work to others skilled in the art. However, these inventive concepts may be embodied in many different forms and should not be construed as limited to the embodiments described herein.

[0020] It should be noted that, unless otherwise specified, the embodiments and features described in the present invention can be combined with each other.

[0022] To address the issues of poor sensing accuracy and susceptibility to noise interference in existing garment care technologies, resulting in low levels of intelligence and precision in the care process, this embodiment proposes a multi-dimensional sensing method for garment care, including:

[0023] S1. Obtain the first dimension information;

[0024] The first dimension information is image information within the nursing space, which can be acquired by the first acquisition device; preferably, the first acquisition device is a camera.

[0025] The camera is installed inside the garment care system's care space and can capture images of the environment, the object being cared for, the object being handled, and even foreign objects. Of course, multiple cameras can be installed in different locations to supplement the information captured by the images.

[0026] It should be noted that the environmental object in this application is the care space of the clothing care system; the care object is the item to be cared for, such as clothing, garments, fabrics, and other items that people need to care for in their daily lives; the execution object is the care arm (such as an ironing arm, a clamping arm, etc.) used to perform care actions; foreign objects can be understood as items that do not belong to the care space, as well as items that are in an abnormal spatial position in the care space, such as pets, infants, dangerous objects, clothing that has fallen in the care space, etc.

[0027] S2, Obtain the second dimension information;

[0028] The second dimension information is physical measurement information within the nursing space, which can be acquired through a second acquisition device; preferably, the second acquisition device is a sensor used to collect physical measurement information, which includes, but is not limited to, at least one of temperature information, humidity information, size information, motion information, location information, object recognition information, etc.

[0029] Preferably, the sensor is mounted on the care arm and can perform detection tasks as the care arm moves; alternatively, the sensor is mounted inside the care space of the garment care system and can acquire information such as temperature, humidity, and object recognition information within the care space. Of course, depending on specific implementation requirements, multiple sensors in different locations or of different types can also be used.

[0030] S3. Generate perception data within the nursing space based on the first dimension information and the second dimension information;

[0031] S4. Based on the perceived data, generate at least one of the following strategies: environmental control strategy, motion control strategy, nursing media control strategy, and prompting strategy.

[0032] Therefore, this application supplements and corrects the perceived information in the nursing space by acquiring image information and physical measurement information separately and combining information from different dimensions, thereby obtaining more accurate perceived information, providing support for intelligent adjustment of nursing modes, and improving the intelligence and precision of the nursing process.

[0033] Step S3 includes:

[0034] S31. Based on the first dimension information, identify the first attribute information of at least one of the environmental objects, nursing objects, and execution objects within the nursing space;

[0035] S32. Based on the second dimension information, identify the second attribute information of at least one of the environmental objects, nursing objects, and execution objects within the nursing space;

[0036] S33. Combining the first attribute information and the second attribute information, information correction is performed to obtain the perception data within the nursing space.

[0037] The first attribute information includes, but is not limited to: the positional relationship between the environmental object and the nursing object, the length, width, thickness, material, and wrinkle information of the nursing object, the action and position information of the performing object, and whether there are foreign objects in the environmental object, such as fallen clothing, dangerous objects, pets, or infants.

[0038] The second attribute information includes, but is not limited to: detecting temperature and / or humidity information in the care space through temperature / humidity sensors; detecting the length, width, and thickness information of the care object through sensors such as microswitches, pressure sensors, capacitive sensors, infrared sensors, and laser sensors; detecting the motion and position information of the object being cared for through limit switches, microswitches, pressure sensors, infrared sensors, posture sensors, or gyroscopes; and detecting the presence of foreign objects in the environment, such as fallen clothing, dangerous objects, pets, or young children, through acoustic radar, acoustic sensors, and X-ray sensors.

[0039] Of course, perceived data can be regarded as attribute information after information correction, including at least clothing material information, clothing size information, and foreign object information in the environment; in addition, the content of perceived data can also include the records of the first attribute information and the second attribute information mentioned above, which will not be elaborated further.

[0040] Step S4 includes, but is not limited to:

[0041] At least based on clothing material information, an environmental control strategy is generated to control environmental parameters within the care space.

[0042] Based at least on clothing size information, a motion control strategy for controlling the movement of the nursing arm is generated;

[0043] Based at least on clothing size information and / or material information, generate a care media control strategy to control care media parameters;

[0044] A prompting strategy is generated based at least on foreign object information in the environment.

[0045] Among them, environmental parameters include humidity and / or temperature; nursing arm stroke includes the clamping stroke and ironing stroke of the corresponding nursing arm, such as the clamping stroke of the clamping arm and the ironing stroke of the ironing arm; nursing medium includes steam, hot air, etc., and correspondingly, nursing medium parameters include the supply volume of nursing medium per unit time, the temperature of nursing medium, the supply pressure of nursing medium, etc., such as steam volume, steam temperature, steam pressure, hot air volume, etc.

[0046] In practical applications, the visual size of the image typically has an error of about 2cm, which is enough to affect the accuracy of clamping or ironing, leading to situations where the garment is not clamped properly or not avoided. Although the sensors on the nursing arm have high-precision recognition, they require corresponding motion support to complete the detection and are limited to local detection. The information collected cannot completely and accurately represent the size of the clothing, and is easily affected by the cutouts, gaps, and obstructions of irregular clothing.

[0047] Therefore, in order to further improve the accuracy of the obtained perceived information, this application focuses on introducing step S33, specifically the combination of the first attribute information and the second attribute information, as well as the information correction.

[0048] Step S33 includes, but is not limited to:

[0049] Method 1: Based on the first attribute information, obtain the area of ​​interest, use the second attribute information to measure the size within the area of ​​interest, correct the first attribute information, and obtain the perception data within the nursing space.

[0050] Option 1 can be considered as image-guided physical detection. The first attribute information can be regarded as guiding (prior) information. The range of interest includes the actual detection range of the sensor. The first attribute information is corrected by using the second attribute information within the actual detection range to improve the accuracy of the perceived data.

[0051] For example, by utilizing the image information of clothing, information such as the visual size and some detailed features of the clothing (i.e., the first attribute information) can be obtained. Then, based on the visual size and detailed features of the clothing, a detection area (i.e., the region of interest) is defined. The sensor is then controlled to measure the size of the clothing within the detection area (i.e., the second attribute information). This measurement method is more precise, and the measurement range can be adjusted in a timely manner according to the region of interest during the measurement process, thereby obtaining more accurate size information and reducing the problem of inaccurate clothing measurement results due to improper selection of sensor test contact points.

[0052] Method 2: Using the second attribute information as the basic information, use the first attribute information to determine whether the data in the second attribute information is within a reasonable range; if so, do not correct the data in the second attribute information and use the data in the second attribute information as the corresponding perception data; if not, use the corresponding data in the first attribute information to correct it and use the corrected data as the corresponding perception data.

[0053] Scheme 2 can be regarded as verifying the physical detection results based on image recognition, with sensor measurement results (second attribute information) as the main factor and image information (first attribute information) as an auxiliary factor or used to correct the sensor measurement results.

[0054] For example, since clothing types and sizes have certain size specifications, image recognition can be used to obtain the relevant clothing type and size. Combining the size specifications of the clothing type with the image recognition results (or historical measurement results), it is determined whether the sensor measurement results (i.e., the second attribute information) are within a reasonable size range. If they are within a reasonable range, the sensor measurement results are kept as the corresponding sensing data. If they are not, the size specifications and / or image recognition results are used to correct the sensor measurement results to obtain the final clothing measurement results, which are then used as the corresponding sensing data.

[0055] Method 3: Statistically analyze the first attribute information and the corresponding second attribute information to obtain the average error statistics of the attribute information. Fit the correction algorithm based on the average error statistics. Substitute the first attribute information and the second attribute information obtained in the actual clothing care process into the correction algorithm to obtain the perception data in the care space.

[0056] Option 3 can be viewed as a fusion and correction of image recognition and physical detection results. The two are fused according to certain rules, including but not limited to weighting. It should be noted that the process of obtaining statistical information, average error statistics, and correction algorithms can be conducted by the garment care system manufacturer through experimentation and by providing relevant correction algorithms. Alternatively, large amounts of data can be obtained from a cloud server for statistical calculation and the provision of relevant correction algorithms.

[0057] For example, since image recognition and physical detection results inevitably contain deviations, especially for certain types of clothing, the errors can be significant. During the manufacturer's experimental testing, based on a large amount of data, the average measurement errors of sensor measurement results and image recognition results for different types of clothing can be statistically analyzed. This allows for the creation of a statistical table of the average errors between sensor measurement results and image recognition results for different clothing types. Then, based on the table, the weight ratio of the two results can be calculated or a fusion function can be fitted (i.e., a correction algorithm can be obtained). It should be noted that the average error statistical table is prior information and does not need to be recalculated for each actual clothing care.

[0058] In actual garment care, the final garment size measurement result can be obtained by substituting the actual sensor measurement results and image recognition results into the weight ratio or into the fusion function.

[0059] Option 4: Obtain multiple sets of attribute information to train and obtain an error prediction model; import the first attribute information and / or the second attribute information obtained during the actual clothing care process into the error prediction model to obtain the perception data within the care space;

[0060] Each set of attribute information includes the first attribute information, the second attribute information corresponding to the first attribute information, the error data between the first attribute information and the actual information, and the error data between the second attribute information and the actual information.

[0061] Option 4 can be viewed as modifying and / or fusing the image recognition results and physical detection results, and then, based on a measurement error prediction (correction) model using a large amount of data, importing the actual detected first attribute information and / or second attribute information into the error prediction model for correction, thereby reducing measurement errors and obtaining more accurate data. The error prediction model can be established using existing AI large-scale model technology, or it can be achieved through manual calculation fitting or computer data fitting techniques.

[0062] The error prediction model is trained using a large amount of data (including attribute information and error data), enabling it to automatically predict errors. After importing actual attribute information, the model can automatically predict the error in that attribute information and correct it. For example, image recognition results (first attribute information) obtained during actual garment care are imported into the error prediction model to correct the image recognition results and improve their accuracy; or, physical detection results (second attribute information) obtained during actual garment care are imported into the error prediction model to correct the physical detection results and improve their accuracy; or, both image recognition results and physical detection results obtained during actual garment care are imported into the error prediction model to correct both and improve the overall accuracy of the result.

[0063] It should be noted that in step S33, any one of the methods from method one to method four can be used in combination with at least one of the other methods.

[0064] In this invention, for any garment care system, the multi-dimensional sensing method for garment care provided in this application can be used.

[0065] Based on the relevant technical solutions provided in this application, the garment care system also includes conventional components such as a care compartment, a care arm, sensors, and cameras. Existing technologies can be directly adopted for these conventional components, and they will not be described in detail here.

[0066] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A multi-dimensional sensing method for garment care, characterized in that, The method includes: S1. Obtain the first dimension information; S2, Obtain the second dimension information; S3. Generate perception data within the nursing space based on the first dimension information and the second dimension information; S4. Based on the perceived data, generate at least one of the following strategies: environmental control strategy, motion control strategy, nursing media control strategy, and prompting strategy.

2. The multi-dimensional sensing method for clothing care according to claim 1, characterized in that, The first dimension information is image information within the nursing space, and the second dimension information is physical measurement information within the nursing space. The physical measurement information includes at least one of temperature information, humidity information, size information, motion information, location information, and object recognition information.

3. The multi-dimensional sensing method for clothing care according to claim 1, characterized in that, Step S3 includes: S31. Based on the first dimension information, identify the first attribute information of at least one of the environmental objects, nursing objects, and execution objects within the nursing space; S32. Based on the second dimension information, identify the second attribute information of at least one of the environmental objects, nursing objects, and execution objects within the nursing space; S33. Combining the first attribute information and the second attribute information, information correction is performed to obtain the perception data within the nursing space.

4. The multi-dimensional sensing method for clothing care according to claim 3, characterized in that, The environmental object is the care space of the clothing care system; the care object is the item to be cared for; and the execution object is the care arm.

5. The multi-dimensional sensing method for clothing care according to claim 3, characterized in that, Step S33 includes any one of methods one to four, or step S33 includes any one of methods one to four combined with at least one of the others; One method is to use the first attribute information as the basic information, obtain the attention range, use the second attribute information to measure the size within the attention range, correct the first attribute information, and obtain the perception data within the nursing space. Method 2 is as follows: using the second attribute information as the basic information, the first attribute information is used to determine whether the data in the second attribute information is within a reasonable range; if so, the data in the second attribute information is not corrected and is used as the corresponding perception data; if not, the corresponding data in the first attribute information is used to correct it and the corrected data is used as the corresponding perception data. Method 3 is as follows: statistically analyze the first attribute information and the corresponding second attribute information to obtain the average error statistics of the attribute information, and fit a correction algorithm based on the average error statistics; substitute the first attribute information and the second attribute information obtained in the actual clothing care process into the correction algorithm to obtain the perception data in the care space; Option 4 involves acquiring multiple sets of attribute information to train and obtain an error prediction model; then importing the first attribute information and / or the second attribute information obtained during the actual clothing care process into the error prediction model to obtain the perception data within the care space.

6. The multi-dimensional sensing method for clothing care according to claim 1, characterized in that, The perceived data includes at least information on clothing material, clothing size, and foreign objects in the environment.

7. A multi-dimensional sensing method for clothing care according to claim 6, characterized in that, Step S4 includes: At least based on clothing material information, an environmental control strategy is generated to control environmental parameters within the care space. Based at least on clothing size information, a motion control strategy for controlling the movement of the nursing arm is generated; Based at least on clothing size information and / or material information, generate a care media control strategy to control care media parameters; A prompting strategy is generated based at least on foreign object information in the environment.

8. The multi-dimensional sensing method for clothing care according to claim 7, characterized in that, The environmental parameters include humidity and / or temperature; the nursing arm stroke includes the clamping stroke and ironing stroke of the nursing arm; the nursing medium parameters include the supply volume of the nursing medium per unit time, the temperature of the nursing medium, and the supply pressure of the nursing medium, wherein the nursing medium includes steam and hot air.

9. A garment care system, characterized in that, The garment care system employs the multi-dimensional sensing method for garment care as described in any one of claims 1-8.