Ironing equipment control method and ironing equipment

Abstract

The invention discloses a control method of ironing equipment and the ironing equipment, and the control method comprises the steps: responding to a starting instruction of the ironing equipment, carrying out the image collection of a fabric through a camera of the ironing equipment, obtaining at least two frames of fabric images, and enabling each fabric image to belong to the same region of the fabric; dividing each fabric image into a plurality of blocks according to a target division strategy; according to each block, the material category of the fabric is determined by using a target detection model, and the target detection model comprises an inconvolution operator, a convolution attention module and a fusion layer; and determining target ironing parameters of the ironing equipment according to the material category. According to the technical scheme, misjudgment caused by slight shaking, instantaneous light reflection and the like of the fabric is effectively reduced, the material type of the fabric is rapidly and accurately determined, the target ironing parameters matched with the material type can be determined, the better ironing effect is obtained, and the user experience is improved.

Description

Technical Field

[0001] This application relates to the field of ironing technology, and in particular to a control method for ironing equipment and an ironing equipment. Background Technology

[0002] In order for ironing equipment to iron fabrics with appropriate ironing parameters, some ironing equipment has a fabric recognition function.

[0003] For example, patent CN106436246A discloses an intelligent garment steamer system, including a garment identification tag for recording garment characteristic information; a garment information acquisition device for collecting garment characteristic information by reading the garment identification tag; a control and storage module, including a control unit, a storage unit, and a transmission unit. The storage unit stores ironing plans, the control unit matches and retrieves the ironing plans from the storage unit based on the garment characteristic information read by the garment information acquisition device, and the transmission unit transmits the ironing plans; and an ironing mechanism controlled by the control unit to perform ironing operations on the garment. It is evident that this intelligent garment steamer system uses garment identification tags to determine garment characteristic information, but the garment composition identified by the identification tag sometimes does not match the actual fabric type, causing fabric identification errors. Furthermore, the intelligent garment steamer system cannot identify fabrics without identification tags. Summary of the Invention

[0004] This application provides a control method and ironing device for an ironing equipment. Based on a fabric image, a target detection model including an inward roll operator, a convolutional attention module, and a fusion layer is used to perform material recognition, thereby improving the accuracy of fabric material recognition.

[0005] In a first aspect, a control method for an ironing device is provided, comprising: responding to a start command for the ironing device, acquiring images of a fabric using a camera of the ironing device to obtain at least two fabric images, each fabric image belonging to the same region on the fabric; dividing each fabric image into multiple blocks according to a target segmentation strategy; determining the material category of the fabric using a target detection model based on each block, the target detection model including an involute operator, a convolutional attention module, and a fusion layer, wherein the involute operator is used to extract target spatial features from each block, the convolutional attention module is used to determine key regions in each block, and the fusion layer is used to fuse the target spatial features of blocks belonging to the same location in different fabric images; and determining target ironing parameters of the ironing device based on the material category.

[0006] In some embodiments, determining the material category of the fabric using an object detection model based on each of the blocks includes: extracting target spatial features of each block using the involute operator and the convolutional attention module; fusing the target spatial features of blocks belonging to the same location in different fabric images using the fusion layer to determine fusion features; and determining the material category and confidence level of the fabric using the classifier of the object detection model based on the fusion features.

[0007] In this way, the target spatial features of each block are determined by the involution operator and the convolutional attention module, and the target spatial features are fused by the fusion layer, thereby improving the accuracy and robustness of material recognition.

[0008] In some embodiments, before dividing each of the fabric images into multiple blocks according to the target partitioning strategy, the method further includes: preprocessing each of the fabric images, the preprocessing including at least one of the following: illumination compensation, contrast stretching.

[0009] By preprocessing the fabric images, the image quality can be further improved, thereby increasing the accuracy of material identification.

[0010] In some embodiments, each fabric image employs a 2-megapixel resolution. Dividing each fabric image into multiple blocks according to a target segmentation strategy includes uniformly dividing each frame of the fabric image into 9 blocks of 3×3 pixels. This achieves efficient and accurate determination of the fabric's material type while improving the clarity of the fabric image.

[0011] In some embodiments, acquiring at least two frames of fabric images using the camera of the ironing device includes: continuously acquiring three frames of fabric images using the camera at a target distance from the fabric surface. This improves both the accuracy and efficiency of material recognition.

[0012] In some embodiments, the target distance is 50±0.5mm. This can improve the clarity of the captured fabric image while avoiding the high temperature of the heating plate of the ironing equipment from affecting the camera.

[0013] In some embodiments, the camera is used to acquire images of the fabric at a target magnification, and each fabric image is a microscopic image conforming to the target magnification. This allows the fabric images to more accurately represent the texture of the fabric, thereby improving the accuracy of fabric material identification.

[0014] In some embodiments, the target detection model employs the YOLO-CLS model. This makes the target detection model more lightweight while improving the real-time performance of material recognition.

[0015] In some embodiments, the target spatial features include at least one of the following: fiber orientation, interlacing point.

[0016] By using at least one of the fiber orientation and interlacing points as target spatial features, the target spatial features can be made more consistent with the actual material of the fabric, which can further improve the accuracy of fabric material identification.

[0017] Secondly, an ironing device is provided, including a camera, an image processing chip, and a controller. The image processing chip is equipped with a target detection model, and the controller is configured to: in response to a start command of the ironing device, use the camera of the ironing device to acquire images of a fabric, obtaining at least two frames of fabric images, each of the fabric images belonging to the same region on the fabric; divide each of the fabric images into multiple blocks according to a target segmentation strategy; determine the material category of the fabric based on each block using the target detection model, wherein the target detection model includes an involute operator, a convolutional attention module, and a fusion layer, the involute operator is used to extract target spatial features from each block, the convolutional attention module is used to determine key regions in each block, and the fusion layer is used to fuse the target spatial features of blocks belonging to the same location in different fabric images; and determine the target ironing parameters of the ironing device based on the material category.

[0018] By applying the above technical solutions, since fabric images can accurately represent the actual texture features of the fabric, and the involute operator can extract the target spatial features of each block, the convolutional attention module can determine the key areas in each block, and the fusion layer can fuse the target spatial features of blocks belonging to the same position in different fabric images, the misjudgment caused by slight fabric shaking, instantaneous reflection, etc. can be effectively reduced, and the material category of the fabric can be quickly and accurately determined. Thus, the target ironing parameters that match the material category can be determined, resulting in better ironing effect and improved user experience. Attached Figure Description

[0019] To more clearly illustrate the technical solutions of this application, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments recorded in this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0020] Figure 1 This is a flowchart of a control method for an ironing device according to an embodiment of this application; Figure 2 This is a flowchart illustrating the determination of the material category of a fabric according to an embodiment of this application. Figure 3 This is a schematic diagram illustrating the principle of fusing target spatial features of blocks belonging to the same location in different fabric images using a fusion layer, according to an embodiment of this application. Figure 4 This is a structural block diagram of an ironing device according to an embodiment of this application. Detailed Implementation

[0021] Various embodiments and features of this application are described herein with reference to the accompanying drawings.

[0022] It should be understood that various modifications can be made to the embodiments described herein. Therefore, the above description should not be considered as limiting, but merely as an example of embodiments. Other modifications within the scope and spirit of this application will be apparent to those skilled in the art.

[0023] The accompanying drawings, which are included in and form part of this specification, illustrate embodiments of the present application and, together with the general description of the present application given above and the detailed description of the embodiments given below, serve to explain the principles of the present application.

[0024] These and other features of this application will become apparent from the following description of preferred forms of embodiments given as non-limiting examples, with reference to the accompanying drawings.

[0025] It should also be understood that although this application has been described with reference to some specific examples, those skilled in the art can certainly implement many other equivalent forms of this application.

[0026] The above and other aspects, features and advantages of this application will become more apparent when taken in conjunction with the accompanying drawings and in view of the following detailed description.

[0027] Specific embodiments of this application are described thereafter with reference to the accompanying drawings; however, it should be understood that the claimed embodiments are merely examples of this application, which can be implemented in various ways. Well-known and / or repeated functions and structures are not described in detail to avoid unnecessary or redundant details that could obscure the application. Therefore, the specific structural and functional details claimed herein are not intended to be limiting, but merely serve as the basis and representative basis for the claims to teach those skilled in the art to use this application in a variety of substantially any suitable detailed structures.

[0028] This specification may use the phrases “in one embodiment,” “in another embodiment,” “in yet another embodiment,” or “in other embodiments,” all of which may refer to one or more of the same or different embodiments according to this application.

[0029] This application discloses a control method for an ironing device. The method utilizes the ironing device's camera to acquire at least two frames of fabric images. Each fabric image is divided into multiple blocks according to a target segmentation strategy. Based on each block, a target detection model is used to determine the fabric's material category. Then, the target ironing parameters of the ironing device are determined based on the material category. Since fabric images can accurately represent the actual texture features of the fabric, and the involute operator can extract target spatial features from each block, a convolutional attention module identifies key regions within each block, and a fusion layer fuses the target spatial features of blocks belonging to the same location in different fabric images, this method effectively reduces misjudgments caused by slight fabric shaking or momentary reflections. This allows for rapid and accurate determination of the fabric's material category, thereby enabling the determination of target ironing parameters adapted to the material category, resulting in better ironing effects and improved user experience.

[0030] like Figure 1 As shown, the control method includes the following steps: Step S101: In response to the start command of the ironing device, the camera of the ironing device is used to capture images of the fabric to obtain at least two fabric images, each of which belongs to the same area on the fabric.

[0031] In this embodiment, the ironing device includes, for example, any one of an electric iron, a garment steamer, or a steam iron. The start command for the ironing device can be triggered by the user pressing a start button on the device, by the user powering on the device, by a voice command, or by the user starting the device on a user terminal (such as a mobile phone). The ironing device is equipped with a camera. When the ironing device is placed on fabric (such as clothing or cloth) or when the user holds the ironing device, in response to the start command, the camera captures images of the same area on the fabric, obtaining at least two frames of fabric images. The camera can be a CMOS (Complementary Metal-Oxide Semiconductor) camera or a CCD (Charge Coupled Device) camera. There can be one or more cameras.

[0032] Step S102: Divide each fabric image into multiple blocks according to the target segmentation strategy.

[0033] In this embodiment, since it is necessary to fuse the fabric images in the future, each fabric image is divided into multiple blocks according to a predetermined target segmentation strategy. Each fabric image is segmented using the target segmentation strategy, and different blocks in the same fabric image correspond to different positions of the fabric image.

[0034] Step S103: Based on each block, the material category of the fabric is determined using a target detection model. The target detection model includes an involute operator, a convolutional attention module, and a fusion layer. The involute operator is used to extract target spatial features from each block. The convolutional attention module is used to determine key regions in each block. The fusion layer is used to fuse target spatial features of blocks belonging to the same location in different fabric images.

[0035] In this embodiment, the trained target detection model is used to detect targets in each area and determine the material category of the fabric. The target detection model includes an involution operator, a convolutional attention module (CBAM module), and a fusion layer. The involution operator is used to extract target spatial features from each block, the convolutional attention module is used to determine key regions in each block, and the fusion layer is used to fuse the target spatial features of blocks belonging to the same location in different fabric images.

[0036] The material category of the fabric may include any one of the following: pure cotton, silk, polyester blend, etc. The material category of the fabric may also include unidentified materials. In some embodiments of this application, the target detection model may also output the confidence level corresponding to the material category.

[0037] Step S104: Determine the target ironing parameters of the ironing equipment according to the material category.

[0038] In this embodiment, different material categories can correspond to different target ironing parameters. A pre-established correspondence between different material categories and target ironing parameters can be established, and the target ironing parameters of the ironing equipment can be determined based on the material category and the correspondence. Target ironing parameters may include one or more parameters such as target temperature, target steam volume, target water volume, and steam jet mode. For example, when the target ironing parameters include target temperature, if the material category is cotton / linen, the target temperature is 160℃; if the material category is wool, the target temperature is 150℃; if the material category is silk, the target temperature is 120℃; if the material category is synthetic fiber, the target temperature is 100℃; and if the material category is unidentified, the target temperature is 100℃, to minimize the risk of burns. Furthermore, since the temperature threshold for burns to wool is 170℃, by setting the target temperature for cotton / linen to 160℃, even if wool is mistakenly identified as cotton / linen, it will not cause burns to the fabric, thereby improving safety during the ironing process.

[0039] After determining the target ironing parameters, the ironing equipment can then be used to iron the fabric according to these parameters. Since the target ironing parameters are compatible with the fabric material type, a good ironing effect can be obtained.

[0040] In some embodiments of this application, the ironing device may include a display screen. After determining the target ironing parameters, the target ironing parameters and material category are displayed on the display screen, allowing the user to intuitively understand the current working status of the ironing device. If the material category does not match the actual material category, the user can also make timely adjustments, thereby further improving the user experience.

[0041] An embodiment of this application discloses a control method for an ironing device. In response to a start command for the ironing device, the method uses the device's camera to acquire images of the fabric, obtaining at least two fabric images, each belonging to the same region on the fabric. Each fabric image is divided into multiple blocks according to a target segmentation strategy. Based on each block, a target detection model is used to determine the material category of the fabric. The target detection model includes an involute operator, a convolutional attention module, and a fusion layer. The involute operator is used to extract target spatial features from each block, the convolutional attention module is used to determine key regions within each block, and the fusion layer is used to fuse the target spatial features of blocks belonging to the same location in different fabric images. The target ironing parameters of the ironing device are then determined based on the material category. Since fabric images can accurately represent the actual texture features of the fabric, and the involute operator can extract the target spatial features of each block, the convolutional attention module can determine the key areas in each block, and the fusion layer can fuse the target spatial features of blocks belonging to the same position in different fabric images, it can effectively reduce misjudgments caused by slight fabric shaking, instantaneous reflection, etc., and achieve fast and accurate determination of the fabric material category. In this way, the target ironing parameters that match the material category can be determined, resulting in better ironing effect and improved user experience.

[0042] In some embodiments of this application, the step of determining the material category of the fabric using a target detection model based on each of the said blocks is as follows: Figure 2 As shown, it includes the following steps: Step S1031: Extract the target spatial features of each block using the involution operator and the convolutional attention module.

[0043] In this embodiment, the target spatial features can be fabric structural features related to fabric material recognition, and the key region can be the area in the fabric image whose discriminative power for fabric material recognition is higher than a threshold. After inputting each block into the target detection model, the involute operator and convolutional attention module are used to extract the target spatial features of each block.

[0044] Step S1032: The target spatial features of blocks belonging to the same location in different fabric images are fused using the fusion layer to determine the fusion features.

[0045] In this embodiment, the target spatial features of each block are input into the fusion layer. The fusion layer fuses the target spatial features of blocks belonging to the same location in different fabric images to obtain multiple block fusion features. The fusion features are determined by the fusion features of each block.

[0046] Step S1033: Based on the fusion features, the material category and confidence level of the fabric are determined using the classifier of the target detection model.

[0047] In this embodiment, the fused features are input into the classifier of the target detection model, and the classifier performs classification processing to determine the material category and confidence level of the fabric.

[0048] In this way, the target spatial features of each block are determined by the involution operator and the convolutional attention module, and the target spatial features are fused by the fusion layer, thereby improving the accuracy and robustness of material recognition.

[0049] In some embodiments of this application, before dividing each of the fabric images into multiple blocks according to the target segmentation strategy, the method further includes: The fabric images are preprocessed, and the preprocessing includes at least one of the following: illumination compensation and contrast stretching.

[0050] In this embodiment, by performing illumination compensation on the fabric image, the color imbalance problem caused by uneven illumination or light source deviation in the fabric image can be improved. Illumination compensation algorithms can be used to compensate for the illumination of the fabric image, including any one of the following algorithms: GrayWorld color equalization algorithm, histogram equalization, nonlinear transformation, local region processing, etc. Contrast stretching specifically enhances the image contrast by adjusting the grayscale value distribution or brightness range of the image.

[0051] By preprocessing the fabric images, the image quality can be further improved, thereby increasing the accuracy of material identification.

[0052] In some embodiments of this application, each of the fabric images adopts a 2-megapixel resolution, and the step of dividing each of the fabric images into multiple blocks according to a target segmentation strategy includes: Each frame of the fabric image is evenly divided into 9 blocks of 3×3.

[0053] In this embodiment, each fabric image can be at a resolution of 2 megapixels, which can improve the clarity of the fabric images. By evenly dividing each frame of fabric image into 9 blocks of 3×3, a reasonable division of the 2-megapixel resolution fabric image is achieved, thereby efficiently and accurately determining the material category of the fabric.

[0054] It should be noted that the resolution of the above fabric image and the way it is divided into blocks are one specific implementation of the embodiments of this application. Those skilled in the art can flexibly adopt other resolutions and block division methods as needed, and different resolutions and block division methods are all within the protection scope of this application.

[0055] In some embodiments of this application, the step of using the camera of the ironing device to capture images of the fabric and obtain at least two frames of fabric images includes: At a target distance from the fabric surface, the camera continuously captures three frames of images of the fabric.

[0056] In this embodiment, by using a camera to continuously capture three frames of fabric images, the accuracy and efficiency of material identification can be improved.

[0057] It should be noted that the above method of obtaining at least two fabric images is a specific implementation of the embodiments of this application. Those skilled in the art can flexibly use fabric images with other frame numbers as needed, and fabric images with different frame numbers are all within the protection scope of this application.

[0058] For example, such as Figure 3 As shown, three consecutive frames of fabric images were acquired, designated as P1, P2, and P3. P1, P2, and P3 were evenly divided into 3×3 blocks, totaling 9 blocks. The target spatial features of each block in P1 are a1, b1, c1, d1, e1, f1, g1, h1, and i1, respectively. The target spatial features of each block in P2 are a2, b2, c2, d2, e2, f2, g2, h2, and i2, respectively. The target spatial features of each block in P3 are a3, b3, c3, d3, e3, f3, g3, h3, and i3, respectively.

[0059] a1, a2, and a3 belong to the same block at the same location on P1, P2, and P3. A fusion layer is used to fuse a1, a2, and a3 to obtain block fusion feature A; b1, b2, and b3 belong to the same block at the same location on P1, P2, and P3. A fusion layer is used to fuse b1, b2, and b3 to obtain block fusion feature B; c1, c2, and c3 belong to the same block at the same location on P1, P2, and P3. A fusion layer is used to fuse c1, c2, and c3 to obtain block fusion feature C, and so on, to obtain the fused image P4. The fusion features of each block in P4 are A, B, C, D, E, F, G, H, and I, respectively.

[0060] In some embodiments of this application, the target distance is 50±0.5mm.

[0061] In this embodiment, the camera is placed inside the ironing equipment. By setting the target distance to 50±0.5mm, the high temperature of the heating plate of the ironing equipment can be avoided from affecting the camera, while ensuring the clarity of the captured fabric image. In particular, for thick and soft fabrics (such as fluffy wool or multi-layered cotton and linen) that can provide a depth of field of about 1cm, a clear fabric image can still be obtained, avoiding blurring problems caused by changes in fabric thickness.

[0062] In some embodiments of this application, the camera is used to acquire images of the fabric at a target magnification, and each fabric image is a microscopic image that conforms to the target magnification.

[0063] In this embodiment, the camera can be a microscope camera, which can directly obtain a microscopic image of the target magnification after taking a picture of the fabric surface, thereby making the fabric image more accurately represent the texture of the fabric, and thus improving the accuracy of fabric material identification.

[0064] In some embodiments of this application, the target magnification is not less than 10x magnification.

[0065] By setting the target magnification to at least 10x, a clearer fabric texture can be obtained, thereby further improving the accuracy of fabric material identification.

[0066] In some embodiments of this application, the target detection model adopts the YOLO-cls model.

[0067] In this embodiment, YOLO-cls is a variant of convolutional neural networks (CNN) specifically designed for image classification tasks. By employing the YOLO-cls model, the object detection model is made more lightweight while improving the real-time performance of material recognition.

[0068] It should be noted that the above-mentioned type of object detection model is a specific implementation of the embodiments of this application. Those skilled in the art can flexibly adopt different types of object detection models (such as SSD (Single Shot MultiBox Detector), Faster R-CNN, etc.) as needed. Different types of object detection models are all within the protection scope of this application.

[0069] In some embodiments of this application, the target spatial features include at least one of the following: fiber orientation, interlacing point.

[0070] In this embodiment, fiber orientation refers to the direction in which fibers are arranged in the fabric, while interlacing points are structural nodes formed by the intersection of fibers. By using at least one of fiber orientation and interlacing points as target spatial features, the target spatial features are made more consistent with the actual material of the fabric, which can further improve the accuracy of fabric material identification.

[0071] This application also proposes an ironing device, such as... Figure 4 As shown, the device includes a camera, an image processing chip, and a controller. The image processing chip contains a target detection model, and the controller is configured to: In response to a start command for the ironing equipment, the camera of the ironing equipment is used to capture images of the fabric, and at least two fabric images are obtained, each of which belongs to the same area on the fabric. The image processing chip divides each fabric image into multiple blocks according to a target segmentation strategy. The image processing chip determines the material category of the fabric based on each of the blocks using the target detection model. The target detection model includes an involute operator, a convolutional attention module, and a fusion layer. The involute operator is used to extract target spatial features from each of the blocks. The convolutional attention module is used to determine key regions in each of the blocks. The fusion layer is used to fuse target spatial features from blocks belonging to the same location in different fabric images. The target ironing parameters of the ironing equipment are determined based on the material category.

[0072] In this embodiment, the image processing chip can be, for example, the Ingenic T32 chip with an NPU, and the controller can be, for example, an MCU (Microcontroller Unit). The image processing chip includes a target detection model.

[0073] After the ironing device is placed on the fabric (such as clothing or cloth) or when the user holds the ironing device, the controller responds to the start command of the ironing device by using the camera of the ironing device to acquire images of the fabric, obtaining at least two frames of fabric images, each belonging to the same area on the fabric. The controller sends at least two frames of fabric images to the image processing chip, which divides each fabric image into multiple blocks according to a target segmentation strategy, and determines the material category of the fabric based on each block using a target detection model. The image processing chip then sends the material category of the fabric to the controller, which determines the target ironing parameters of the ironing device based on the material category. Subsequently, the ironing device can be used to iron the fabric according to the target ironing parameters. Because the target ironing parameters are adapted to the material category of the fabric, a good ironing effect can be obtained.

[0074] Since fabric images can accurately represent the actual texture features of the fabric, and the involute operator can extract the target spatial features of each block, the convolutional attention module can determine the key areas in each block, and the fusion layer can fuse the target spatial features of blocks belonging to the same position in different fabric images, it can effectively reduce misjudgments caused by slight fabric shaking, instantaneous reflection, etc., and achieve fast and accurate determination of the fabric material category. In this way, the target ironing parameters that match the material category can be determined, resulting in better ironing effect and improved user experience.

[0075] Alternatively, as an alternative, the controller can also perform the step of dividing each fabric image into multiple blocks according to the target segmentation strategy.

[0076] Alternatively, as an alternative, the image processing chip may respond to the start command of the ironing device by using the camera of the ironing device to capture images of the fabric and obtain at least two frames of fabric images.

[0077] Optionally, the image processing chip can be set up independently of the controller or integrated into the controller.

[0078] Other embodiments of an ironing device described in this application can be found in the corresponding embodiments of a control method for an ironing device described in this application, which will not be repeated here.

[0079] In the above embodiments, implementation can be achieved entirely or partially through software, hardware, firmware, or any combination thereof. When implemented using software, it can be implemented entirely or partially in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, all or part of the processes or functions described in the embodiments of this application are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., coaxial cable, fiber optic, digital subscriber line) or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that integrates one or more available media. The available medium can be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium (e.g., solid-state drive), etc.

[0080] The above embodiments are merely exemplary embodiments of this application and are not intended to limit this application. The scope of protection of this application is defined by the claims. Those skilled in the art can make various modifications or equivalent substitutions to this application within its substance and scope of protection, and such modifications or equivalent substitutions should also be considered to fall within the scope of protection of this application.

Claims

1. A control method of an ironing apparatus, characterized by, The method comprises the following steps: in response to a start instruction of an ironing device, image acquisition of a fabric is performed by using a camera of the ironing device, at least two frames of fabric images are obtained, and each fabric image belongs to a same region on the fabric; each fabric image is divided into a plurality of blocks according to a target division strategy; a material category of the fabric is determined according to each block by using a target detection model, the target detection model comprises an inner convolution operator, a convolution attention module and a fusion layer, the inner convolution operator is used for extracting target spatial features of each block, the convolution attention module is used for determining a key region in each block, and the fusion layer is used for fusing target spatial features of blocks belonging to a same position in different fabric images; a target ironing parameter of the ironing device is determined according to the material category.

2. The control method of the ironing apparatus as claimed in claim 1, characterized in that, The method comprises the following steps: target spatial features of each block are extracted by using the inner convolution operator and the convolution attention module; fused features are determined by fusing target spatial features of blocks belonging to a same position in different fabric images by using the fusion layer; a material category and a confidence of the fabric are determined by using a classifier of the target detection model according to the fused features.

3. The control method of the ironing apparatus as claimed in claim 1, characterized in that, Before each fabric image is divided into a plurality of blocks according to a target division strategy, the method further comprises the following steps: each fabric image is preprocessed, and the preprocessing comprises at least one of the following: illumination compensation, contrast stretching.

4. The control method of the ironing apparatus as claimed in claim 1, characterized in that, Each fabric image adopts a 2 million pixel resolution, and each fabric image is divided into a plurality of blocks according to a target division strategy, which comprises the following steps: each fabric image is uniformly divided into 3*3=9 blocks.

5. The control method of the ironing apparatus as claimed in claim 4, characterized in that, The method comprises the following steps: three frames of fabric images are continuously acquired by using the camera at a target distance from the surface of the fabric.

6. The control method of the ironing apparatus as claimed in claim 5, characterized in that, The target distance is 50±0.5 mm.

7. The control method of the ironing apparatus as claimed in claim 1, characterized in that, The camera is used for image acquisition of the fabric at a target magnification, and each fabric image is a microscopic image conforming to the target magnification.

8. The control method of the ironing apparatus as claimed in claim 1, characterized in that, The target detection model adopts a YOLO-cls model.

9. The control method of the ironing apparatus as claimed in claim 1, characterized in that, The target spatial features comprise at least one of the following: fiber direction, interlaced point.

10. An ironing device, characterized by A camera, an image processing chip and a controller are comprised, the target detection model is arranged in the image processing chip, and the controller is configured to: in response to a start instruction of an ironing device, image acquisition of a fabric is performed by using a camera of the ironing device, at least two frames of fabric images are obtained, and each fabric image belongs to a same region on the fabric; each fabric image is divided into a plurality of blocks according to a target division strategy by using the image processing chip; The image processing chip determines the material category of the fabric according to each of the blocks by using the target detection model, the target detection model comprises an inner convolution operator, a convolution attention module and a fusion layer, the inner convolution operator is used for extracting target spatial features of each of the blocks, the convolution attention module is used for determining key regions in each of the blocks, and the fusion layer is used for fusing target spatial features of blocks belonging to the same position in different fabric images. According to the material category, a target ironing parameter of the ironing equipment is determined.

Images