Material identification control method for washing equipment

Abstract

The present invention relates to the field of smart home appliance technology, and specifically provides a material identification control method for a washing machine, aiming to solve the problem of inaccurate clothing material identification in existing washing machines. To this end, the identification control method of the present invention includes the following steps: after the clothing is placed in the inner drum of the washing machine, obtaining a static image of the clothing in a stationary state; determining the material type of the clothing based on the static image; judging whether the material type can be determined based on the static image; if it is determined that the material type cannot be determined based on the static image, controlling the rotation of the inner drum; during the rotation of the inner drum, obtaining a first dynamic video of the clothing in a rotating state; determining the material type based on the first dynamic video, and sequentially using different methods for identification of different types of materials until the material type of the clothing is determined, thereby improving the accuracy and precision of material type identification and thereby improving user experience.

Description

Technical Field

[0001] The present invention relates to the technical field of smart home appliances, and specifically provides a material recognition and control method for a washing device. Background Art

[0002] As people's living standards improve, their demand for smarter washing machines is growing. When it comes to clothing, the most obvious characteristic is its material. People expect washing machines to be even smarter, automatically acquiring information about clothing attributes such as material, and recommending appropriate washing programs and parameters based on these attributes. This ensures not only cleans clothing but also prevents damage. For example, silk clothing is not suitable for washing in water, which can often damage it.

[0003] Existing methods for identifying clothing materials typically rely on image recognition of static images of clothing to determine the clothing's material. However, during the image capture process, factors such as camera pixel density and ambient light intensity significantly impact image quality. Furthermore, the same method is used for identification of all clothing materials, yet some materials appear very similar. This results in low accuracy in static image material recognition, making it impossible to accurately identify clothing materials. This, in turn, leads to inappropriate wash program recommendations, negatively impacting the user experience.

[0004] Therefore, the art needs a new material identification control method for washing equipment to solve the above problems. Summary of the Invention

[0005] The present invention aims to solve the above technical problem, that is, to solve the problem of inaccurate clothing material identification in existing washing equipment.

[0006] The present invention provides a material identification control method for a washing device, the identification control method comprising the following steps: after the clothes are placed in the inner drum of the washing device, obtaining a static image of the clothes in a stationary state; determining the material type of the clothes based on the static image; judging whether the material type can be determined based on the static image; if it is determined that the material type cannot be determined based on the static image, controlling the rotation of the inner drum; during the rotation of the inner drum, obtaining a first dynamic video of the clothes in a rotating state; and determining the material type based on the first dynamic video.

[0007] In the preferred technical solution of the above-mentioned identification control method, the step of "determining the material type of the clothing based on the static image" specifically includes: calling the classification model to identify the static image; the step of "determining whether the material type can be determined based on the static image" specifically includes: judging whether the material type belongs to the first type based on the identification result; if the material type belongs to the first type, judging that the material type can be determined; and / or if the material type does not belong to the first type, judging that the material type cannot be determined; wherein, the material type that can be determined by the classification model based on the static image is the first type, and the first type includes wool, wool, cashmere and down.

[0008] In the preferred technical solution of the above-mentioned identification control method, the step of "determining the material type based on the first dynamic video" specifically includes: calling a preset model to analyze the first dynamic video; judging whether the material type can be determined based on the analysis result; if it is determined that the material type cannot be determined, controlling the washing device to perform a water injection operation; after the washing device performs the water injection operation, controlling the inner barrel to rotate; during the rotation of the inner barrel, obtaining a second dynamic video of the clothes in a rotating state; and determining the material type based on the second dynamic video.

[0009] In the preferred technical solution of the above-mentioned identification control method, the step of "calling a preset model to analyze the first dynamic video" specifically includes: converting the first dynamic video into multiple first images according to a first preset sampling period; preprocessing the multiple first images respectively to obtain target information of each first image; inputting all the target information into the preset model; the preset model analyzes all the target information; the step of "determining whether the material type can be determined based on the analysis results" specifically includes: determining whether the material type belongs to the second type; if the material type belongs to the second type, determining that the material type can be determined; and / or if the material type does not belong to the second type, determining that the material type cannot be determined; wherein, the material type that can be determined by the preset model according to the target information is the second type, and the second type includes silk.

[0010] In the preferred technical solution of the above-mentioned identification and control method, the step of "pre-processing the multiple first images separately to obtain target information of each of the first images" specifically includes: identifying the clothing in each of the first images and the outline of the clothing in the image; determining the target information based on the outline; wherein, the target information includes at least one of the outline pixel number curve, the instantaneous change curve and the average distance from the center of the circle of the clothing.

[0011] In the preferred technical solution of the above-mentioned identification and control method, the step of "determining the material type based on the second dynamic video" specifically includes: converting the second dynamic video into multiple second images according to a second preset sampling period; determining the motion information of the clothing corresponding to each second image based on the multiple second images; determining the material type based on all the motion information; wherein, the motion information includes the average velocity of the center of mass and / or the proportion of the total area of ​​the motion area; wherein, the material type that can be determined based on the motion information is the third type, and the third type includes cotton, polyester and cotton-polyester blends.

[0012] In the preferred technical solution of the above-mentioned identification and control method, the step of "converting the second dynamic video into multiple second images according to the second preset sampling period" specifically includes: segmenting the second dynamic video according to the second preset sampling period so that each of the obtained second images includes a complete image of the clothing in the second dynamic video.

[0013] In the preferred technical solution of the above-mentioned identification and control method, the step of "controlling the washing device to perform a water filling operation" specifically includes: obtaining the volume of the clothes; determining a preset water level based on the volume; and filling the washing device with water to the preset water level.

[0014] In the preferred technical solution of the above-mentioned identification and control method, the identification and control method also includes: when it is determined that the material type can be determined, judging whether the clothes are suitable for washing according to the material type; based on the judgment result, selectively recommending a washing program or sending a prompt message according to the material type.

[0015] In the preferred technical solution of the above identification control method, the step of "selectively recommending a washing and care program or sending a prompt message according to the material type based on the judgment result" specifically includes:

[0016] If the clothes are suitable for washing, a washing program is recommended according to the material type; and / or if the clothes are not suitable for washing, a prompt message is sent to remind the user that the clothes are not suitable for washing.

[0017] In the preferred technical solution of the identification control method of the present invention, after the clothes are placed in the inner drum of the washing machine, a static image of the clothes in a stationary state is obtained; the material type of the clothes is determined based on the static image; it is judged whether the material type can be determined based on the static image; if it is determined that the material type cannot be determined based on the static image, the inner drum is controlled to rotate; during the rotation of the inner drum, a first dynamic video of the clothes in a rotating state is obtained; and the material type is determined based on the first dynamic video.

[0018] Compared with the technical solution in the prior art that only performs material identification once based on the image of the clothes regardless of the type of material of the clothes, the present invention first determines the material type of the clothes based on the static image of the clothes in a stationary state, and determines the material type of the clothes for the first time; when it is determined that the material type of the clothes cannot be determined based on the static image, the material type is determined based on the clothes in a waterless rotation state, and the material type of the clothes is determined for the second time. In this process, different methods are used in turn to identify different types of materials until the material type of the clothes is determined. Different types of materials can be identified in a targeted manner, and the material type of the clothes can be accurately identified, thereby improving the accuracy and precision of material type identification, and thus improving user experience.

[0019] Furthermore, the classification model is called to recognize static images. Based on the recognition results, it can accurately identify whether the material type of the clothing is wool, wool, cashmere or down. It can specifically identify the materials of wool, wool, cashmere and down clothing, thereby improving the accuracy and precision of identifying wool, wool, cashmere and down materials.

[0020] Furthermore, a preset model is called to analyze the first dynamic video, and based on the analysis results, it is judged whether the material type can be determined, and whether the material type of the clothes is the material type corresponding to the preset model can be accurately identified, thereby accurately determining the material type of the identified clothes; when it is determined that the material type cannot be determined, it means that the material type of the clothes is a type that cannot be accurately determined based on both the static image and the first dynamic video. At this time, the washing machine is controlled to perform a water filling operation. After the washing machine is filled with water to a preset water level, the inner drum is controlled to rotate. During the rotation of the inner drum, a second dynamic video of the clothes in a rotating state is obtained, and the material type is determined based on the second dynamic video. The material type of the clothes is determined for the third time. In this process, different methods are used in turn for identification of specific material types until the material type of the clothes is determined. The material type of the clothes can be accurately identified, thereby improving the accuracy and precision of material type identification, thereby improving user experience.

[0021] Furthermore, the first dynamic video is converted into multiple first images according to a first preset sampling period; the multiple first images are preprocessed respectively to obtain target information of each first image; all target information is input into a preset model; the preset model analyzes all target information. Since the preset model can determine that the material type is silk based on the target information, according to the analysis results, it can accurately identify whether the material type of the clothing is silk, and can specifically identify the material of silk clothing, thereby improving the accuracy and precision of silk material identification.

[0022] Furthermore, the second dynamic video is converted into multiple second images according to a second preset sampling period; the motion information of the clothing corresponding to each second image is determined respectively according to the multiple second images; the material type is determined according to all the motion information. Since the material types that can be determined according to the motion information include cotton, polyester and cotton-polyester blends, it is possible to accurately identify whether the material type of the clothing is cotton, polyester or cotton-polyester blends according to the motion information, and it is possible to specifically identify the materials of cotton, polyester and cotton-polyester blends of clothing, thereby improving the accuracy and precision of identifying cotton, polyester and cotton-polyester blends.

[0023] Furthermore, when it is determined that the material type can be determined, it means that the material type of the clothes has been accurately identified. At this time, it is possible to accurately judge whether the clothes are suitable for washing based on the material type, and based on the judgment result, selectively recommend a washing program or send a prompt message based on the material type, avoiding washing the clothes when they are not suitable for washing, thereby avoiding damage to the clothes and further improving the user experience. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] The identification control method of the present invention will be described below with reference to the accompanying drawings and in combination with a washing machine, wherein:

[0025] Figure 1 This is the process of the identification control method of the present invention Figure 1 ;

[0026] Figure 2 is a flow chart of a method for determining a material type of clothing based on a static image according to the present invention;

[0027] Figure 3 The process of the method for determining the material type according to the first dynamic video of the present invention is Figure 1 ;

[0028] Figure 4 The process of the method for determining the material type according to the first dynamic video of the present invention is Figure 2 ;

[0029] Figure 5 is a flow chart of a method for controlling a washing machine to perform a water filling operation according to the present invention;

[0030] Figure 6 is a flow chart of a method for determining a material type based on a second dynamic video according to the present invention;

[0031] Figure 7 This is the process of the identification control method of the present invention Figure 2 ;

[0032] Figure 8 This is a flow chart of the present invention for selectively recommending a washing and care program or sending a reminder message;

[0033] Figure 9 It is a logic diagram of the identification control method of the present invention. DETAILED DESCRIPTION

[0034] The preferred embodiments of the present invention are described below with reference to the accompanying drawings. Those skilled in the art should understand that these embodiments are merely intended to explain the technical principles of the present invention and are not intended to limit the scope of protection of the present invention. For example, although the present application is described in conjunction with a washing machine, the technical solution of the present invention is not limited thereto. The identification and control method can obviously also be applied to other washing appliances such as a washer-dryer, and such a change does not deviate from the principles and scope of the present invention.

[0035] It should be noted that, in the description of the present invention, the terms "first", "second" and "third" are used for descriptive purposes only and should not be understood as indicating or implying relative importance.

[0036] Furthermore, it should be noted that, in the description of the present invention, unless otherwise expressly specified or limited, the term "disposed" should be understood in a broad sense. For example, it can mean a fixed connection, a detachable connection, or an integral connection; it can mean a mechanical connection or an electrical connection; it can mean a direct connection or an indirect connection through an intermediate medium; it can mean internal communication between two components. Those skilled in the art will understand the specific meanings of the above terms in the present invention based on the specific circumstances.

[0037] Based on the technical problems raised in the background technology, the present invention provides a material identification control method for a washing machine, which aims to first determine the material type of the clothes based on a static image of the clothes in a stationary state, and perform a first determination of the material type of the clothes; when it is determined that the material type of the clothes cannot be determined based on the static image, the material type is determined based on the clothes in a waterless rotation state, and the material type of the clothes is determined a second time. In this process, different methods are used in turn to identify different types of materials until the material type of the clothes is determined. Different types of materials can be identified in a targeted manner, and the material type of the clothes can be accurately identified, thereby improving the accuracy and precision of material type identification, and thereby improving user experience.

[0038] See first Figure 1 , the material identification control method for a washing machine of the present invention is described. Figure 1 This is the process of the identification control method of the present invention Figure 1 .

[0039] like Figure 1 As shown, the material identification control method for a washing machine of the present invention comprises the following steps:

[0040] S100, after the clothes are placed in the inner drum of the washing machine, a static image of the clothes in a stationary state is obtained;

[0041] S200, determining the material type of the clothing according to the static image;

[0042] S300, determining whether the material type can be determined based on the static image;

[0043] S400: If it is determined that the material type cannot be determined based on the static image, control the inner barrel to rotate;

[0044] S500: Acquire a first dynamic video of the clothes in a rotating state during the rotation of the inner tub;

[0045] S600: Determine a material type according to the first dynamic video.

[0046] In step S100, a static image of the clothes in the inner tub is captured by a camera or other photographing device installed on the washing machine. It should be noted that although the term "tub" is used, it is not limited to a pulsator washing machine and can also be a drum washing machine.

[0047] In step S400, when it is determined that the material type cannot be determined based on the static image, in order to further determine the material type of the clothes, the inner drum is controlled to rotate so that the camera device captures a first dynamic video of the clothes in a rotating state, so as to determine the material type of the clothes again based on the first dynamic video in a waterless state.

[0048] In step S500, a first dynamic video of the clothes in the inner tub is captured by a camera or other shooting device, for example, a first dynamic video of a first preset duration such as 30 seconds, 45 seconds or 60 seconds is captured.

[0049] Preferably, the material types of clothing include three types. The first type includes materials with obvious appearance characteristics such as wool, cashmere and down; the second type includes materials with a relatively smooth surface such as silk; and the third type includes cotton, polyester and cotton-polyester blends, and the materials of the second and third types are both fiber materials. When the inner barrel rotates, especially at the moment of changing the rotation direction, the instantaneous motion state of the two types of materials will be significantly different due to inertia. Of course, the specific material types included in each type are not limited to the material types listed above. For example, the first type can also include material types such as rabbit hair and mink fur, the second type can also include material types such as silk, and the third type also includes material types such as cotton, linen, and polyester fibers. The specific material types included in each type can be flexibly adjusted and set according to the actual material types and their performance.

[0050] Based on the above classification results, different methods are used in turn to identify the characteristics of different materials until the material type of the clothing is determined, ensuring the accuracy and precision of material type identification.

[0051] Refer to the following Figure 2 , the method of determining the material type of clothing based on a static image of the present invention is described. Figure 2 is a flow chart of a method for determining the material type of clothing based on a static image according to the present invention.

[0052] like Figure 2 As shown, in step S200, the step of "determining the material type of clothing according to the static image" specifically includes:

[0053] S211: Call the classification model to identify the static image.

[0054] The classification model is pre-installed on the washing machine and trained based on image samples of clothing materials such as wool, wool, cashmere, down, and fiber. The image samples are labeled, for example, with labels such as wool, wool, cashmere, down, and fiber. In other words, the classification model can identify material types such as wool, wool, cashmere, and down based on static images, specifically identifying clothing materials such as wool, wool, cashmere, and down, improving the accuracy and precision of material identification. However, if the clothing material type is a fiber material such as silk, cotton, polyester, or a cotton-polyester blend, the model can only identify the clothing material as fiber, but cannot determine the specific material type. The classification model can be a SENet model, a Keras model, a VGG model, an AtoC model, or other classification models. Regardless of the classification model used, the specific material identification method corresponding to any model should not constitute any limitation on the present invention.

[0055] Continue reading Figure 2 In step S300, the step of "determining whether the material type can be determined based on the static image" specifically includes:

[0056] S311, judging whether the material type belongs to the first type according to the recognition result; if so, executing step S312; if not, executing step S313;

[0057] S312, determining whether the material type can be determined;

[0058] S313: Determine that the material type cannot be determined.

[0059] In step S312, if the material type belongs to the first type, it means that the material type of the clothing corresponding to the static image is one of wool, wool, cashmere or down. Since the classification model can determine the first type of material such as wool, wool, cashmere and down based on the static image, it can determine the material type of the clothing, that is, wool, wool, cashmere or down.

[0060] In step S313, if the material type does not belong to the first type, it means that the material type of the clothing corresponding to the static image is not wool, wool, cashmere or down, but may be other fiber materials such as silk, real silk, cotton, polyester and cotton-polyester blends. The classification model can only determine that the material of this type of clothing is fiber, but cannot specify the specific material type, so it cannot identify the specific material type of the clothing.

[0061] For example, in the above process, the similarity between a static image and an image sample of a first type of clothing material, such as wool, cashmere, or down, can be calculated. When the similarity is greater than a preset similarity (e.g., 95%), the label corresponding to the image sample (e.g., wool label) is determined as the material type of the clothing. When the similarity is less than or equal to the preset similarity, the label corresponding to the image sample (e.g., wool label) is not determined as the material type of the clothing. The similarity can be represented by Euler distance or cosine distance.

[0062] It should be noted that, in the above process, step S312 and step S313 have no order but are parallel and are only related to the judgment result of whether the material type belongs to the first type. The corresponding steps can be executed according to different judgment results.

[0063] Refer to the following Figures 3 to 6 , the method of determining the material type based on the first dynamic video and the second dynamic video of the present invention is described. Figure 3 The process of the method for determining the material type according to the first dynamic video of the present invention is Figure 1 ; Figure 4 The process of the method for determining the material type according to the first dynamic video of the present invention is Figure 2 ; Figure 5 is a flow chart of a method for controlling a washing machine to perform a water filling operation according to the present invention; Figure 6 It is a flow chart of the method for determining the material type according to the second dynamic video of the present invention.

[0064] like Figure 3 As shown, in step S600, the step of "determining the material type according to the first dynamic video" specifically includes:

[0065] S611, calling a preset model to analyze the first dynamic video;

[0066] S612: Determine whether the material type can be determined based on the analysis results.

[0067] S613: If it is determined that the material type cannot be determined, controlling the washing machine to perform a water filling operation;

[0068] S614, after the washing machine performs the water filling operation, controlling the inner tub to rotate;

[0069] S615: Acquire a second dynamic video of the clothes in a rotating state during the rotation of the inner tub;

[0070] S616: Determine the material type according to the second dynamic video.

[0071] The preset model is pre-installed on the washing machine and trained based on dynamic video samples of second and third types of clothing rotating in the inner drum without water, as well as target information samples of the clothing corresponding to the dynamic video samples. During model training, using a pulsator washing machine as an example, dynamic video samples of second and third types of clothing rotating in the inner drum without water are captured. The videos are then converted into images according to a set sampling period, such as 0.03s, 0.05s, or 0.08s. Edge detection is used to identify the outline of the clothing. The number of pixels within the outline is related to the area of ​​clothing spread out in the drum. The difference in pixels within the outline between two adjacent frames reflects the instantaneous deformation of the clothing. Since the camera is fixed above the washing machine's inner drum, the center of the pulsator can be located in the image. Points within the clothing outline are randomly selected and the average distance from each point to the pulsator center is calculated, measuring distances related to clothing size, weight, and smoothness. The model records the number of outline pixels, instantaneous change curves, and average distance from the center of the circle for the second and third types of clothing, respectively, when placed in a washing machine and spun dry. Using a large amount of sample data, a pre-set model is trained to distinguish between the second and third types of materials. Since the second type only includes silk and the third type includes cotton, polyester, and cotton-polyester blends, and the instantaneous motion states of cotton, polyester, and cotton-polyester blends differ very little, it can accurately determine whether the material type of the clothing is silk. However, if the material type of the clothing is the third type, it can only be determined that the clothing is made of cotton and polyester (i.e., the third type), but the specific material type cannot be determined.

[0072] That is to say, the preset model can determine that the material type only includes silk based on the first dynamic video of the clothing and the corresponding target information of the clothing, so that it can accurately identify whether the material type of the clothing is silk, and can specifically identify the material of silk clothing, thereby improving the accuracy and precision of silk material identification.

[0073] Among them, the preset model can be other deep learning models or linear regression models such as CNN model, ResNet18 model, ResNet101 model, DeeplabV3+ model, ResNeXt model and HRNet model. Regardless of the model adopted, the specific method of identifying materials corresponding to any model should not constitute any limitation to the present invention.

[0074] The target information includes the number of pixels of the outline of the clothing, the instantaneous change curve and the average distance from the center of the circle. Of course, the target information can also include any one or two of the number of pixels of the outline of the clothing, the instantaneous change curve and the average distance from the center of the circle.

[0075] like Figure 4 As shown, in step S611, the step of "calling a preset model to analyze the first dynamic video" specifically includes:

[0076] S621: Convert a first dynamic video into a plurality of first images according to a first preset sampling period;

[0077] S622, pre-processing the multiple first images respectively to obtain target information of each first image;

[0078] S623, inputting all target information into a preset model;

[0079] S624: Analyze all target information using a preset model;

[0080] In step S621, the first preset sampling period corresponds to the set sampling period corresponding to the training of the preset model. For example, an image is captured every 0.03s, 0.05s, or 0.08s, and the first dynamic video is converted into a plurality of first images, such as 10, 20, or 30. Of course, the first preset sampling period can also be different from the set sampling period corresponding to the training of the preset model, which is not listed here.

[0081] In step S622, the preset model is used to preprocess the multiple first images to obtain target information for each first image. Specifically, the clothing item in each first image and its outline within the image are identified, and target information is determined based on the outline. The target information includes a pixel count curve, an instantaneous change curve, and the average distance from the center of the circle of the clothing item's outline. It should be noted that the specific content of the target information is consistent with the specific content of the target information used during training of the preset model and changes based on changes in the specific content of the target information used during training of the preset model.

[0082] Continue reading Figure 4 In step S612, the step of "determining whether the material type can be determined based on the analysis results" specifically includes:

[0083] S631, determine whether the material type belongs to the second type; if so, execute step S632; if not, execute step S633;

[0084] S632: Determine whether the material type can be determined;

[0085] S633: Determine that the material type cannot be determined.

[0086] In step S631, the target information, such as the pixel count curve, instantaneous change curve, and average distance from the center of the circle of the clothing, is compared one by one with the target information samples stored in the preset model to determine whether there is a target information sample in the preset model that is completely or substantially consistent with the target information. If so, the material type corresponding to the target information sample that is completely or substantially consistent with the target information is determined as the material type of the clothing, such as the second type. If not, the material type corresponding to the target information sample that is completely or substantially consistent with the target information is not determined as the material type of the clothing. Substantially consistent can be a degree of consistency such as 90%, 95%, etc.

[0087] In step S632, if the material type belongs to the second type, it means that the material type of the clothing corresponding to the first dynamic video is silk. Since the preset model can only determine the material type including silk based on the first dynamic video of the clothing and the corresponding target information of the clothing, the material type of the clothing can be determined, that is, silk.

[0088] In step S633, if the material type does not belong to the second type, it means that the material type of the clothing corresponding to the first dynamic video is not silk, but may be a third type such as cotton, polyester, and cotton-polyester blended. The preset model can only determine that the material of this type of clothing is cotton and polyester (that is, the third type), but cannot determine the specific material type, and therefore cannot identify the specific material type of the clothing.

[0089] It should be noted that, in the above process, step S632 and step S633 have no order but are parallel and are only related to the judgment result of whether the material type belongs to the second type. The corresponding steps can be executed according to different judgment results.

[0090] like Figure 5 As shown, in step S613, the step of "controlling the washing machine to perform the water injection operation" specifically includes:

[0091] S641, obtaining the volume of the clothes;

[0092] S642, determining a preset water level according to the volume;

[0093] S643: Fill the washing machine with water to a preset water level.

[0094] Correspondingly, in step S614, after the washing machine is filled with water to a preset water level, the inner tub is controlled to rotate.

[0095] Alternatively, in an alternative embodiment, the weight of the clothes can be obtained and the preset water level can be determined based on the weight; or, the preset water level can be determined based on the set washing program. Regardless of the method used to determine the preset water level, as long as the preset water level can be determined, the washing machine can be filled with water to the preset water level.

[0096] In steps S613 to S616, when it is determined that the material type cannot be determined, in order to further determine the material type of the clothes, the washing machine is controlled to perform the water injection operation and the inner drum is controlled to rotate so that the camera device can capture a second dynamic video of the clothes in a rotating state, so as to determine the material type again based on the second dynamic video in the water state.

[0097] Furthermore, a second dynamic video of the clothes in the inner tub is captured by a camera or other shooting device, for example, a second dynamic video of a second preset duration such as 20s, 40s or 60s is captured.

[0098] like Figure 6 As shown, in step S616, the step of "determining the material type according to the second dynamic video" specifically includes:

[0099] S651, converting the second dynamic video into a plurality of second images according to a second preset sampling period;

[0100] S652: Determine motion information of clothing corresponding to each second image according to the plurality of second images;

[0101] S653. Determine the material type based on all motion information.

[0102] The motion information includes the average speed of the center of mass of the clothing and the total area of ​​the motion region. Of course, the motion information may also only include the average speed of the center of mass of the clothing or the total area of ​​the motion region.

[0103] Among them, the material types that can be determined based on the second dynamic video and motion information include cotton, polyester and cotton-polyester blends.

[0104] In step S651, the second dynamic video is segmented according to a second preset sampling period, so that each second image includes a complete image of the clothing in the second dynamic video. The second preset sampling period may be to capture an image every 0.02 seconds, 0.04 seconds, or 0.05 seconds, etc., and the second dynamic video is converted into a plurality of second images, such as 5, 15, 25, or 35, each of which includes a complete image of the clothing.

[0105] In step S652, the average velocity of the center of mass and the total area ratio of the motion region are calculated according to the following method.

[0106] Method for calculating the average center of mass velocity: Estimate the center of mass position of the clothing through the segmented area of ​​the same clothing in each second image, calculate the coordinate difference between the corresponding center of mass positions in two adjacent frames of images, and then calculate the center of mass movement distance. Determine the interval time between the two frames of images according to the second preset sampling period, and calculate the average center of mass velocity v based on the center of mass movement distance d and the interval time t, that is, v = d / t.

[0107] Calculation method for the total motion area ratio: Define an empty image and assume that the grayscale value of each pixel in the empty image is 0. Starting from the first frame of the second image obtained by sampling, the segmented second image is sequentially superimposed on the empty image according to its original position. After the superposition is completed, the number of pixels with non-zero grayscale values ​​is calculated and divided by the total number of pixels to obtain the motion area ratio.

[0108] According to the above method, the average speed of the center of mass of the clothing and the proportion of the total area of ​​the movement area are calculated respectively. For example, the calculated average speed of the center of mass of the clothing is 0.409 m / s, and the proportion of the total area of ​​the movement area of ​​the clothing is 0.468%; for another example, the calculated average speed of the center of mass of the clothing is 0.441 m / s, and the proportion of the total area of ​​the movement area of ​​the clothing is 0.491%; for another example, the calculated average speed of the center of mass of the clothing is 0.408 m / s, and the proportion of the total area of ​​the movement area of ​​the clothing is 0.591%.

[0109] In step S653, the material type is determined based on all the motion information, and the determined material type is cotton, polyester, or a cotton-polyester blend. For example, the average speed of the center of mass of the clothes calculated in step S652 is 0.409 m / s, and the proportion of the cotton motion area is 0.468%, which respectively match the first preset average speed of the center of mass of cotton of 0.41 m / s and the first preset area of ​​motion of cotton of 0.47%, and the material type of the clothes is determined to be cotton; for another example, the average speed of the center of mass of the clothes calculated in step S652 is 0.441 m / s, and the proportion of the total area of ​​the motion area of ​​the clothes is 0.491%, which respectively match the second preset average speed of the center of mass of polyester of 0.44 m / s and the second preset area of ​​motion of polyester of 0.49%, and the material type of the clothes is determined to be polyester; for another example, the average speed of the center of mass of the clothes calculated in step S652 is 0.411 m / s, and the proportion of the total area of ​​the motion area of ​​the clothes is 0.591%, which respectively match the third preset average speed of the center of mass of cotton and polyester blended of 0.41 m / s and the third preset area of ​​motion of cotton and polyester of 0.59%, and the material type of the clothes is determined to be cotton and polyester.

[0110] That is, the material type of the clothing is determined in turn through static images, first dynamic videos, and second dynamic videos using different methods, and finally the material type of the clothing is determined to be cotton, polyester, or cotton-polyester blend, thereby improving the accuracy and precision of material type identification, and thus improving the user experience.

[0111] Furthermore, if the material type of the clothing belongs to the first type, the material type of the clothing, such as wool, can be determined through the static image of the clothing, and the clothing material type is only identified once.

[0112] If the material type of the clothing belongs to the second type, first, the static image of the clothing is used to determine that the material type of the clothing does not belong to the first type. Secondly, the first dynamic video of the clothing in a waterless state is used to determine the material type of the clothing, such as silk. The clothing material type is identified twice.

[0113] If the material type of the clothing belongs to the third type, first, through the static image of the clothing, it is determined that the material type of the clothing does not belong to the first type. Secondly, through the first dynamic video of the clothing in the water-free state, it is determined that the material type of the clothing does not belong to the second type. Finally, through the second dynamic video of the clothing in the water state, the material type of the clothing is determined, such as cotton. The clothing material type is identified three times.

[0114] Refer to the following Figure 7 and Figure 8 , further describing the identification control method of the present invention. Figure 7 This is the process of the identification control method of the present invention Figure 2 ; Figure 8 This is a flowchart of the present invention for selectively recommending a washing and care program or sending a reminder message.

[0115] like Figure 7 As shown, in the case where it is determined that the material type can be determined, the identification control method further includes:

[0116] S700, determining whether the clothing is suitable for washing based on the material type;

[0117] S800: Based on the judgment result, selectively recommend a washing and care program according to the material type or send a prompt message.

[0118] When it is determined that the material type can be determined, it means that the material type of the clothes has been accurately identified. At this time, it is possible to accurately judge whether the clothes are suitable for washing based on the material type, and based on the judgment result, selectively recommend a washing program or send a prompt message based on the material type, avoiding washing clothes when the clothes are not suitable for washing, thereby avoiding damage to the clothes and further improving the user experience.

[0119] like Figure 8 As shown, in step S800, the step of "selectively recommending a washing and care program or sending a prompt message based on the material type based on the judgment result" specifically includes:

[0120] S811. If the clothing is suitable for washing, recommend a washing program based on the material type;

[0121] S812. If the clothes are not suitable for washing, a prompt message is sent to remind the user that the clothes are not suitable for washing.

[0122] In step S811, if the clothes are suitable for washing, it means that the material of the clothes is cotton, polyester, cotton-polyester blended, etc., which will not cause damage to the clothes when washed with water, and a washing and care program is recommended according to the material type.

[0123] In step S812, if the clothes are not suitable for washing, it means that the material of the clothes is silk, wool, cashmere, down, etc., which will cause damage to the clothes if washed with water. In order to avoid damage to the clothes, a prompt message is sent to remind the user that the clothes are not suitable for washing.

[0124] Furthermore, prompt information can be sent to smart terminals such as mobile phones, tablets, smart bracelets and smart watches in the form of text, pictures, animations, etc.; the prompt information can also be sent directly by the prompt module in the form of voice, text, pictures, animations, etc.

[0125] It should be noted that in the above process, step S700, step S400, and step S613 are not sequentially executed, but are parallel to each other. They are solely related to whether the material type of the clothing has been determined, and the corresponding step is executed based on the different determination results. Steps S811 and S812 are not sequentially executed, but are parallel to each other. They are solely related to whether the clothing is suitable for washing, and the corresponding step is executed based on the different determination results.

[0126] Refer to the following Figure 9 , a possible control flow of the present invention is introduced. Among them, FIG9 is a logic diagram of the identification control method of the present invention.

[0127] like Figure 9 As shown, a possible complete process of the identification control method of the present invention is:

[0128] S901, after the clothes are placed in the inner drum of the washing machine, a static image of the clothes in a stationary state is obtained;

[0129] S902, calling a classification model to recognize a static image;

[0130] S903, judging whether the material type belongs to the first type according to the recognition result; if not, executing step S904; if so, executing step S916;

[0131] S904: Determine that the material type cannot be determined;

[0132] After step S904, execute step S905;

[0133] S905: Control the inner tub to rotate, and obtain a first dynamic video of the clothes in the rotating state during the inner tub rotation;

[0134] S906: Convert the first dynamic video into a plurality of first images according to a first preset sampling period;

[0135] S907: Preprocess the plurality of first images to obtain target information of each first image; wherein the target information includes a pixel count curve, an instantaneous change curve, and an average distance from the center of the clothing contour;

[0136] S908, inputting all target information into a preset model;

[0137] S909: Analyze all target information using a preset model;

[0138] S910, judging whether the material type belongs to the second type according to the analysis result; if not, executing step S911; if so, executing step S916;

[0139] S911, determine that the material type cannot be determined;

[0140] After step S911, execute step S912;

[0141] S912: Fill the washing machine with water to a preset water level. After the washing machine completes the water filling operation, control the inner tub to rotate, and obtain a second dynamic video of the clothes in the rotating state during the inner tub rotation;

[0142] S913. Convert the second dynamic video into a plurality of second images according to a second preset sampling period;

[0143] S914: Determine motion information of the clothing corresponding to each second image based on the plurality of second images; wherein the motion information includes an average velocity of the center of mass of the clothing and a total area of ​​a motion region;

[0144] S915. Determine the material type, such as cotton, based on all the motion information.

[0145] S916: Determine whether the material type can be determined; for example, in step S903, it is determined that the material type can be determined, and the determined material type is wool; in step S910, it is determined that the material type can be determined, and the determined material type is silk;

[0146] After step S915 or S916, execute step S917;

[0147] S917: Determine whether the clothing is suitable for washing based on the material type; if so, proceed to step S918; if not, proceed to step S919;

[0148] S918, recommend cleaning procedures based on material type;

[0149] S919: Send a prompt message to remind the user that the clothes are not suitable for washing.

[0150] It should be pointed out that the above embodiment is only a preferred embodiment of the present invention and is only used to illustrate the principle of the method of the present invention. It is not intended to limit the scope of protection of the present invention. In actual applications, those skilled in the art can allocate the above functions to different steps as needed, that is, decompose or combine the steps in the embodiment of the present invention. For example, the steps of the above embodiment can be combined into one step, or further divided into multiple sub-steps to complete all or part of the functions described above. The names of the steps involved in the embodiments of the present invention are only for distinguishing the steps and are not considered to be limitations of the present invention.

[0151] Thus far, the technical solutions of the present invention have been described in conjunction with the preferred embodiments shown in the accompanying drawings. However, it will be readily understood by those skilled in the art that the scope of protection of the present invention is obviously not limited to these specific embodiments. Without departing from the principles of the present invention, those skilled in the art may make equivalent changes or substitutions to the relevant technical features, and the technical solutions after such changes or substitutions will fall within the scope of protection of the present invention.

Claims

1. A material identification control method for washing equipment, characterized in that: The identification control method comprises the following steps: After the clothes are placed in the inner tub of the washing device, a static image of the clothes in a stationary state is acquired; determining a material type of clothing according to the static image; determining whether the material type can be determined based on the static image; If it is determined that the material type cannot be determined based on the static image, controlling the inner barrel to rotate; During the rotation of the inner tub, obtaining a first dynamic video of the clothes in a rotating state; determining the material type according to the first dynamic video; The step of "determining the material type according to the first dynamic video" specifically includes: Calling a preset model to analyze the first dynamic video; According to the analysis results, determining whether the material type can be determined; If it is determined that the material type cannot be determined, controlling the washing device to perform a water injection operation; After the washing device performs a water filling operation, controlling the inner tub to rotate; During the rotation of the inner tub, obtaining a second dynamic video of the clothes in a rotating state; The material type is determined according to the second dynamic video.

2. The identification control method according to claim 1, characterized in that: The step of “determining the material type of clothing according to the static image” specifically includes: Calling a classification model to identify the static image; The step of "determining whether the material type can be determined based on the static image" specifically includes: According to the recognition result, determining whether the material type belongs to the first type; If the material type belongs to the first type, determining that the material type can be determined; and / or If the material type does not belong to the first type, determining that the material type cannot be determined; The material type that can be determined by the classification model based on the static image is the first type, and the first type includes wool, wool, cashmere and down.

3. The identification control method according to claim 1, characterized in that: The step of "calling a preset model to analyze the first dynamic video" specifically includes: Converting the first dynamic video into a plurality of first images according to a first preset sampling period; Preprocessing the plurality of first images respectively to obtain target information of each of the first images; Inputting all the target information into the preset model; The preset model analyzes all the target information; The step of "determining whether the material type can be determined based on the analysis results" specifically includes: Determine whether the material type belongs to the second type; If the material type belongs to the second type, determining that the material type can be determined; and / or If the material type does not belong to the second type, determining that the material type cannot be determined; Among them, the material type that can be determined by the preset model according to the target information is the second type, and the second type includes silk.

4. The identification control method according to claim 3, characterized in that: The step of “respectively preprocessing the plurality of first images to obtain target information of each of the first images” specifically includes: identifying clothing in each of the first images and an outline of the clothing in the image; determining the target information according to the outline; The target information includes at least one of a pixel quantity curve of the clothing contour, an instantaneous change curve, and an average distance from the center of a circle.

5. The identification control method according to claim 1, characterized in that: The step of “determining the material type according to the second dynamic video” specifically includes: converting the second dynamic video into a plurality of second images according to a second preset sampling period; determining motion information of clothing corresponding to each of the second images according to the plurality of second images; Determining the material type based on all the motion information; The motion information includes the average velocity of the center of mass and / or the proportion of the total area of ​​the motion region; Among them, the material type that can be determined according to the motion information is the third type, and the third type includes cotton, polyester and cotton-polyester blended.

6. The identification control method according to claim 5, characterized in that: The step of “converting the second dynamic video into a plurality of second images according to a second preset sampling period” specifically includes: The second dynamic video is segmented according to the second preset sampling period so that each of the obtained second images includes a complete image of the clothing in the second dynamic video.

7. The identification control method according to claim 1, characterized in that: The step of "controlling the washing equipment to perform water injection operation" specifically includes: Get the volume of the clothes; determining a preset water level based on the volume; Fill the washing equipment with water to the preset water level.

8. The identification control method according to any one of claims 1 to 7, characterized in that: The identification control method further includes: If it is determined that the material type can be determined, determining whether the clothing is suitable for washing according to the material type; Based on the judgment result, a washing and care program is selectively recommended or a prompt message is sent according to the material type.

9. The identification control method according to claim 8, characterized in that: The step of "selectively recommending a washing and care program or sending a prompt message based on the material type based on the judgment result" specifically includes: If the garment is suitable for washing, recommend a wash cycle based on the material type; and / or If the clothes are not suitable for washing, a prompt message is sent to remind the user that the clothes are not suitable for washing.

Images