Method and device for identifying moisture absorption and quick drying property of fabric based on dynamic image

Abstract

Embodiments of the present disclosure disclose a dynamic image-based fabric moisture absorption and quick-drying identification method and device. A specific embodiment of the method comprises: in response to a fabric sample being fixed on a sample holder and a pump liquid device starting to pump test liquid, performing parallel image acquisition by a first camera, a second camera and a third camera; for each first image in a first image sequence, performing image stitching on the first image and a corresponding second image of the first image in a second image sequence; respectively performing wet area positioning on a fourth image sequence and a third image sequence obtained; performing water absorption identification according to a first wet area information sequence to obtain water absorption performance description information; and performing quick-drying identification according to the first wet area information sequence and a second wet area information sequence to obtain quick-drying performance description information. The embodiment realizes accurate identification of the dynamic moisture absorption and quick-drying of the fabric sample.

Description

Technical Field

[0001] The embodiments disclosed herein relate to the fields of textile fabrics, performance measurement, image processing, and deep learning, and specifically to a method and apparatus for identifying the moisture absorption and quick-drying properties of fabrics based on dynamic images. Background Technology

[0002] The moisture-wicking and quick-drying properties of fabrics affect the conduction and evaporation of sweat. When fabrics cannot quickly absorb, conduct, and evaporate sweat, sweat accumulates on the skin's surface, causing a stuffy and sticky feeling. Especially in low-temperature environments, fabrics with poor quick-drying properties can continue to draw body heat, resulting in a "wind chill effect." Therefore, developing fabrics with excellent moisture-wicking and quick-drying properties has become one of the main research directions in the textile industry. Currently, GB / T 21655.1 and GB / T 21655.2 are mainly used to evaluate the moisture-wicking and quick-drying properties of fabrics.

[0003] However, the two assessment methods mentioned above often have the following problems: (1) GB / T 21655.1 uses a fixed experimental procedure to evaluate static single indicators, which has low evaluation efficiency and cannot reflect the dynamic moisture absorption and quick-drying properties of the fabric sample. (2) GB / T 21655.2 requires the use of a liquid water dynamic transfer performance tester, which evaluates the moisture absorption and quick-drying properties by combining the resistance changes of the contact surface of discrete point probes. This tester is expensive and cannot reflect the dynamic moisture absorption and quick-drying properties of the fabric sample. Summary of the Invention

[0004] The summary portion of this disclosure is intended to provide a brief overview of the concepts, which will be described in detail in the detailed description portion. This summary portion is not intended to identify key or essential features of the claimed technical solutions, nor is it intended to limit the scope of the claimed technical solutions.

[0005] Some embodiments of this disclosure propose a method and apparatus for identifying the moisture absorption and quick-drying properties of fabrics based on dynamic images, in order to solve the technical problems mentioned in the background section above.

[0006] In a first aspect, some embodiments of this disclosure provide a method for identifying the moisture absorption and quick-drying properties of fabrics based on dynamic images. The method includes: in response to a fabric sample being fixed on a sample stage and a pumping device starting to pump test liquid, parallel image acquisition is performed using a first camera, a second camera, and a third camera to obtain a first image sequence, a second image sequence, and a third image sequence; for each first image in the first image sequence, the first image and its corresponding second image in the second image sequence are stitched together to obtain a fourth image; wet area localization is performed on the obtained fourth image sequence and the third image sequence to obtain a first wet area information sequence and a second wet area information sequence; water absorption is identified based on the first wet area information sequence to obtain water absorption performance description information, wherein the water absorption performance description information includes: dynamic water absorption trend information and dynamic diffusion trend information; quick-drying properties are identified based on the first wet area information sequence and the second wet area information sequence to obtain quick-drying performance description information, wherein the quick-drying performance description information includes: drying half-life.

[0007] Secondly, some embodiments of this disclosure provide a fabric moisture absorption and quick-drying property identification device based on dynamic images. The device includes: an image acquisition unit configured to, in response to a fabric sample being fixed on a sample stage and a pumping device starting to pump test liquid, perform parallel image acquisition through a first camera, a second camera, and a third camera to obtain a first image sequence, a second image sequence, and a third image sequence; and an image stitching unit configured to, for each first image in the first image sequence, stitch together the first image and the corresponding second image in the second image sequence to obtain a fourth image. The wet area localization unit is configured to perform wet area localization on the obtained fourth image sequence and the aforementioned third image sequence respectively, to obtain a first wet area information sequence and a second wet area information sequence; the water absorption identification unit is configured to perform water absorption identification based on the aforementioned first wet area information sequence to obtain water absorption performance description information, wherein the aforementioned water absorption performance description information includes: dynamic water absorption trend information and dynamic diffusion trend information; the quick-drying identification unit is configured to perform quick-drying identification based on the aforementioned first wet area information sequence and the aforementioned second wet area information sequence to obtain quick-drying performance description information, wherein the aforementioned quick-drying performance description information includes: drying half-life.

[0008] Thirdly, some embodiments of this disclosure provide a fabric moisture-wicking and quick-drying property identification system. This system is applied to the fabric moisture-wicking and quick-drying property identification method based on dynamic images described in the first aspect. The system includes: an upper fixing component, a lower placement component, a test liquid storage tank, a pumping device, a first camera, a second camera, a third camera, and a control component. The upper fixing component comprises: an upper fixing component housing, a first fixing pin assembly, a test liquid conduit, a first reflector, and a second reflector. The first fixing pin is provided with... A cylindrical pin with a circular pin hole; the first fixing pin in the first fixing pin set is disposed at the four corners of the first surface of the upper fixing component housing; the test liquid conduit is disposed through the upper fixing component housing; the first end of the test liquid conduit is connected to a drop needle; the included angle between the mirror surfaces of the first reflector and the second reflector is 90 degrees; the plane formed by the first mirror edge of the first reflector and the first mirror edge of the second reflector is coplanar with the first surface of the upper fixing component housing; the drop needle passes through the center of the intersection line of the mirror surfaces of the first reflector and the second reflector. The lower mounting assembly includes a sample stage and a second set of fixing pins. The sample stage is made of transparent material, and the second fixing pins are cylindrical pins. The second fixing pins in the second set are located at the four corners of the second surface of the sample stage. The first fixing pins in the first set and the second fixing pins in the second set correspond one-to-one. The pump device is located between the test solution storage tank and the upper mounting assembly. The pump device and the test solution storage tank are connected by a delivery pipeline. The test solution conduit included in the pump device and the upper mounting assembly are connected... The test solution reservoir contains simulated sweat containing a fluorescent indicator, connected via an infusion line. The first and second cameras are positioned opposite each other, with the first camera facing the center of the first reflector and the second camera facing the center of the second reflector. The third camera faces the center of the first surface of the sample stage. All three cameras are high-speed cameras. The control components are communicatively connected to the pump, the first camera, the second camera, and the third camera.

[0009] Fourthly, some embodiments of this disclosure provide an electronic device, including: one or more processors; and a storage device having one or more programs stored thereon, wherein when the one or more programs are executed by the one or more processors, the one or more processors implement the method described in any implementation of the first aspect and the system described in any implementation of the third aspect.

[0010] Fifthly, some embodiments of this disclosure provide a computer-readable medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the method described in any implementation of the first aspect and the system described in any implementation of the third aspect.

[0011] The above embodiments of this disclosure have the following beneficial effects: The dynamic image-based fabric moisture absorption and quick-drying property identification method of some embodiments of this disclosure achieves accurate identification of the dynamic moisture absorption and quick-drying property of the fabric sample. Specifically, firstly, in response to the fabric sample being fixed on the sample stage and the pumping device starting to pump the test liquid, parallel image acquisition is performed using a first camera, a second camera, and a third camera to obtain a first image sequence, a second image sequence, and a third image sequence. This, combined with a high-speed camera, quickly captures the diffusion, penetration, and drying of the test liquid in the fabric sample. Secondly, for each first image in the first image sequence, the first image and its corresponding second image in the second image sequence are stitched together to obtain a fourth image. In practice, considering the need to avoid positional interference between the camera and the dispensing needle when identifying the fabric's moisture absorption and quick-drying property, the corresponding images are acquired using a first camera, a second camera, and a V-shaped reflector. Therefore, before identification, the first image and the corresponding second image need to be stitched together to form a complete image of the fabric sample. Next, wet zone localization is performed on the obtained fourth image sequence and the aforementioned third image sequence to obtain a first wet zone information sequence and a second wet zone information sequence. This allows for wet zone localization from both the skin-friendly and non-skin-friendly sides of the fabric sample, thereby determining the diffusion changes after the test liquid comes into contact with the textile sample. Further, water absorption is identified based on the aforementioned first wet zone information sequence to obtain water absorption performance description information, which includes dynamic water absorption trend information and dynamic diffusion trend information. By combining continuous first wet zone information, the dynamic moisture absorption performance of the fabric sample is quantified. Finally, quick-drying performance is identified based on the aforementioned first and second wet zone information sequences to obtain quick-drying performance description information, which includes the drying half-life. By combining continuous first and second wet zone information, the dynamic quick-drying performance of the fabric sample is quantified. This method achieves accurate identification of the dynamic moisture absorption and quick-drying performance of the fabric sample. Attached Figure Description

[0012] The above and other features, advantages, and aspects of the embodiments of this disclosure will become more apparent from the accompanying drawings and the following detailed description. Throughout the drawings, the same or similar reference numerals denote the same or similar elements. It should be understood that the drawings are schematic, and elements are not necessarily drawn to scale.

[0013] Figure 1 This is a schematic diagram of the relationship between some components of the fabric moisture absorption and quick-drying property identification system according to some embodiments of this disclosure, viewed from the front. Figure 2 This is a schematic diagram of the relationship between some components of the fabric moisture absorption and quick-drying property identification system according to some embodiments of this disclosure, viewed from a top view. Figure 3 This is a schematic diagram of the relationship between some components of the fabric moisture absorption and quick-drying identification system according to some embodiments of this disclosure, viewed from a side view. Figure 4 This is a schematic diagram showing the connection relationship between the test liquid storage tank and the pump device in a fabric moisture absorption and quick-drying identification system according to some embodiments of this disclosure; Figure 5 This is a flowchart of some embodiments of the fabric moisture absorption and quick-drying property identification method based on dynamic images according to the present disclosure; Figure 6 This is a schematic diagram of the structure of some embodiments of the fabric moisture absorption and quick-drying identification device based on dynamic images according to the present disclosure. Figure 7 This is a schematic diagram of the structure of an electronic device suitable for implementing some embodiments of the present disclosure. Detailed Implementation

[0014] Embodiments of this disclosure will now be described in more detail with reference to the accompanying drawings. While some embodiments of this disclosure are shown in the drawings, it should be understood that this disclosure can be implemented in various forms and should not be construed as limited to the embodiments set forth herein. Rather, these embodiments are provided to provide a more thorough and complete understanding of this disclosure. It should be understood that the accompanying drawings and embodiments of this disclosure are for illustrative purposes only and are not intended to limit the scope of protection of this disclosure.

[0015] It should also be noted that, for ease of description, only the parts relevant to the invention are shown in the accompanying drawings. Unless otherwise specified, the embodiments and features described in this disclosure can be combined with each other.

[0016] It should be noted that the concepts of "first" and "second" mentioned in this disclosure are used only to distinguish different devices, modules or units, and are not used to limit the order of functions performed by these devices, modules or units or their interdependencies.

[0017] It should be noted that the terms "a" and "a plurality of" used in this disclosure are illustrative rather than restrictive, and those skilled in the art should understand that, unless otherwise expressly indicated in the context, they should be understood as "one or more".

[0018] The names of messages or information exchanged between multiple devices in the embodiments of this disclosure are for illustrative purposes only and are not intended to limit the scope of such messages or information.

[0019] This disclosure will now be described in detail with reference to the accompanying drawings and embodiments.

[0020] First, see Figure 1 The diagram shown illustrates the relationship between some components of the fabric moisture absorption and quick-drying recognition system from a frontal view. Figure 2 The diagram shown illustrates the relationships between some components of the fabric moisture-wicking and quick-drying identification system from a top-down view. Figure 3 The diagram shown illustrates the relationship between some components of the fabric moisture-wicking and quick-drying recognition system from a side view perspective. Figure 4 The diagram shown illustrates the connection relationship between the test solution storage tank and the pump device in the fabric moisture absorption and quick-drying property identification system. Figures 1 to 4 This is only used to reveal the components and connections between them in the fabric moisture-wicking and quick-drying identification system, and does not impose any constraints on their dimensions. Specifically: The aforementioned fabric moisture absorption and quick-drying identification system includes: an upper fixing component, a lower placement component, a test liquid storage tank 10, a pumping device 9, a first camera 11, a second camera 12, a third camera 13, and a control component (not shown in the figure).

[0021] The upper fixing component includes: an upper fixing component housing 1, a first fixing pin assembly, a test liquid conduit 3, a first reflector 4, and a second reflector 5.

[0022] The first fixing pin 2 is a cylindrical pin with a circular pin hole. The first fixing pins 2 in the aforementioned set of first fixing pins are located at the four corners of the first surface of the upper fixing assembly housing 1. Specifically, with... Figure 1 For example, the first surface of the upper fixing component housing 1 is the lower surface of the upper fixing component housing 1. In particular, the first surface of the upper fixing component housing 1 is the contact surface between the upper end of the first fixing pin 2 and the upper fixing component housing 1. The circular pin hole of the first fixing pin 2 faces the corresponding second fixing pin 8.

[0023] The test fluid conduit 3 is disposed within the upper fixing component housing 1. Specifically, the test fluid conduit 3 extends from the center of the second surface of the upper fixing component housing 1, through the upper fixing component housing 1, to the center of the first surface of the upper fixing component housing 1. Figure 1 For example, the second surface of the upper fixing component housing 1 is the upper surface of the upper fixing component housing 1. The test liquid conduit 3 is a flexible conduit. The first end of the test liquid conduit 3 is connected to a dripping needle 6. Among them, taking... Figure 1For example, the end of the first conduit is the lower end of the test liquid conduit 3. The dispensing needle 6 is a miniature stainless steel needle. Optionally, a flow rate sensor can also be installed between the test liquid conduit 3 and the dispensing needle 6 to monitor the test liquid flow rate in real time.

[0024] The included angle between the mirror surfaces of the first reflector 4 and the second reflector 5 is 90 degrees. Both the first reflector 4 and the second reflector 5 are planar reflectors. The plane formed by the first mirror edge of the first reflector 4 and the first mirror edge of the second reflector 5 is coplanar with the first surface of the upper fixing assembly housing 1. Figure 1 For example, the first mirror edge is the upper edge of the first reflecting mirror 4 and the second reflecting mirror 5. The aforementioned dispensing needle 6 passes through the center of the intersection line of the mirror surfaces of the first reflecting mirror 4 and the second reflecting mirror 5. Since the included angle between the mirror surfaces of the first reflecting mirror 4 and the second reflecting mirror 5 is 90 degrees, the first reflecting mirror 4 and the second reflecting mirror 5 are connected at the same edge. Figure 1 For example, the first reflector 4 and the second reflector 5 are connected to the lower edge of the reflector, so the dripping needle 6 passes through the center of the lower edge where the first reflector 4 and the second reflector 5 are connected, so as to ensure that the test liquid is dripped onto the fabric sample 14 through the dripping needle 6.

[0025] The aforementioned lower placement assembly includes a sample stage 7 and a second set of fixing pins. The sample stage 7 is made of a transparent material, such as high-transparency acrylic or glass. The second fixing pins 8 are cylindrical pins, and are located at the four corners of the second surface of the sample stage 7. Figure 1 For example, the second surface of the sample stage 7 is the upper surface of the sample stage 7. The first fixing pin 2 in the first fixing pin set and the second fixing pin 7 in the second fixing pin set correspond one-to-one. Specifically, the fixing pin diameter of the second fixing pin 7 is the same as the hole diameter of the first fixing pin 2. When the upper fixing component is connected to the lower placement component, the second fixing pin 7 and the corresponding first fixing pin 2 slide in contact.

[0026] The aforementioned pumping device 9 is disposed between the test solution storage tank 10 and the upper fixing assembly. The pumping device 9 and the test solution storage tank 10 are connected via a delivery pipeline. The pumping device 9 and the test solution conduit 3 included in the upper fixing assembly are also connected via a delivery pipeline. Specifically, the delivery pipeline is a flexible pipeline. The pumping device 9 is a miniature liquid pump. The test solution storage tank contains simulated sweat containing a fluorescent indicator. The simulated sweat can be prepared according to standards such as AATCC TM15 and GB / T 3922. The mass fraction of the fluorescent indicator in the test solution is 0.05%.

[0027] The first camera 11 and the second camera 12 are arranged facing each other, with the first camera 11 facing the center of the first reflector 4 and the second camera 12 facing the center of the second reflector 5. Specifically, the optical axis of the first camera 11 passes through the center of the first reflector 4 and forms a 45-degree angle with it. The optical axis of the second camera 12 passes through the center of the second reflector 5 and forms a 45-degree angle with it. In practice, when the fabric sample 14 is placed horizontally on the center of the second surface of the sample stage 7, light is directed vertically from the fabric sample 14 to the first reflector 4 and the second reflector 5, and after being bent at a 45-degree angle, it is directed horizontally to the first camera 11 and the second camera 12. This ensures that the first image (acquired by the first camera 11) and the second image (acquired by the second camera 12) can represent the upper surface state of the fabric sample 14 with almost no distortion.

[0028] The third camera 13 faces the center of the first surface of the sample stage 7. All three cameras—the first camera 11, the second camera 12, and the third camera 13—are high-speed cameras. Specifically, the optical axis of the third camera 13 passes through the center of the first surface of the sample stage 7 and is perpendicular to it. In practice, when the fabric sample 14 is placed horizontally at the center of the second surface of the sample stage 7, this ensures that the third camera 13 can acquire a third image characterizing the lower surface state of the fabric sample 13.

[0029] The control component is communicatively connected to the pumping device 9, the first camera 11, the second camera 12, and the third camera 13. The control component controls whether the pumping device 9 is turned on and the pumping speed when it is on. The control component also controls the first camera 11, the second camera 12, and the third camera 13 to simultaneously acquire images, performs image processing on the acquired first, second, and third images, and performs corresponding water absorption and quick-drying property recognition. Specifically, the control component can be composed of a control chip and corresponding chip circuitry.

[0030] The aforementioned fabric moisture-wicking and quick-drying property identification system, as an inventive aspect of this disclosure, introduces a high-speed camera and redesigns the system architecture and corresponding hardware from the perspectives of image acquisition completeness and test liquid delivery accuracy. Specifically, firstly, compared to the conventional method of evaluating static single indicators using fixed experimental procedures, by leveraging the rapid image capture capabilities of a high-speed camera, it achieves continuous and rapid capture of diffusion and penetration changes when the test liquid comes into contact with the fabric sample, thereby significantly improving the identification and evaluation efficiency. Furthermore, compared to the method of evaluating moisture-wicking and quick-drying property based on expensive liquid water dynamic transfer performance testers, it significantly reduces instrument costs. Simultaneously, compared to the method of measuring resistance changes at discrete points, the image-based identification method can accurately identify the state changes at various points on the fabric sample. Comparing identification and evaluation costs, efficiency, and accuracy across multiple dimensions, the aforementioned fabric moisture-wicking and quick-drying property identification system is significantly superior to existing identification and evaluation methods.

[0031] Continue to refer to Figure 5 The diagram illustrates a flowchart 500 of some embodiments of a method for identifying the moisture-wicking and quick-drying properties of fabrics based on dynamic images according to the present disclosure. This method includes the following steps: Step 501: In response to the fabric sample being fixed on the sample stage and the pumping device starting to pump the test liquid, parallel image acquisition is performed through the first camera, the second camera and the third camera to obtain the first image sequence, the second image sequence and the third image sequence.

[0032] In some embodiments, the execution entity (e.g., a computing device) of the fabric moisture absorption and quick-drying property identification method based on dynamic imaging can, in response to the fabric sample being fixed on the sample stage and the pumping device starting to pump the test liquid, perform parallel image acquisition using a first camera, a second camera, and a third camera to obtain a first image sequence, a second image sequence, and a third image sequence. Specifically, the fabric sample is illuminated with a fluorescent lamp during image acquisition for the first, second, and third images.

[0033] The first image in the first image sequence is a continuous image captured by the first camera. The second image in the second image sequence is a continuous image captured by the second camera. The third image in the third image sequence is a continuous image captured by the third camera. The image size of the first image is N×M. The image size of the second image is N×M. The image size of the third image is M×M. M = 2N. The fabric sample is a fabric specimen to be tested for moisture absorption and quick-drying properties. The sample size of the fabric sample is K×K. In practice, the skin-friendly side of the fabric sample faces upwards. The non-skin-friendly side of the fabric sample faces downwards. Specifically, the non-skin-friendly side of the fabric sample is coplanar with the second surface of the sample stage. The skin-friendly side of the fabric sample faces the first surface of the upper fixing component housing. The fabric sample should be placed horizontally and flat at the center of the second surface of the sample stage. The test solution is simulated sweat containing a fluorescent indicator with a mass fraction of 0.05%.

[0034] In practice, firstly, after the fabric sample is placed on the sample stage, the computing device sends a pumping command to the pumping device, causing the test solution in the test solution reservoir to be delivered through the infusion line to the test solution conduit included in the upper fixing component, and then dripped onto the fabric sample below the dripping needle via the dripping needle connected to the test solution conduit. Next, when the pumping device starts pumping, it controls the first, second, and third cameras to simultaneously acquire images, obtaining a first image sequence, a second image sequence, and a third image sequence. Because of the synchronous image acquisition method, the number of first images in the first image sequence is the same as the number of second images in the second image sequence, and the number of second images in the second image sequence is the same as the number of third images in the third image sequence. Specifically, the first, second, and third cameras are all set to external trigger mode by default. Therefore, by sending a unified image acquisition signal (e.g., PPS pulse per second) to the first, second, and third cameras, it is ensured that the three cameras start exposure synchronously. This method achieves extremely high synchronization accuracy, thus ensuring the clock synchronization of image acquisition.

[0035] It should be noted that the aforementioned computing device can be either hardware or software. When the computing device is hardware, it can be implemented as a single server or a single terminal device. When the computing device is software, it can be installed within the hardware devices listed above. It can be implemented as a single software program or a software module. No specific limitations are made here. Specifically, the aforementioned computing device can be a control component included in a fabric moisture-wicking and quick-drying recognition system.

[0036] Step 502: For each first image in the first image sequence, the first image and the corresponding second image in the second image sequence are stitched together to obtain the fourth image.

[0037] In some embodiments, the execution entity may, for each first image in the first image sequence, stitch together the first image and the corresponding second image in the second image sequence to obtain a fourth image.

[0038] The fourth image is obtained by stitching the first image and the corresponding second image together. The size of the fourth image is M×M.

[0039] In practice, since light is reflected through the first and second reflectors when acquiring images of the fabric sample, the first and second images acquired by the first and second cameras respectively correspond to half of the fabric sample. Therefore, it is necessary to stitch the first image and the corresponding second image together. Specifically, based on the positional relationship between the fabric sample, the first reflector, the second reflector, the first camera, and the second camera, the stitching boundary is determined. The first image and the corresponding second image are then stitched together at the stitching boundary to obtain the fourth image.

[0040] Step 503: Perform wet area localization on the obtained fourth image sequence and third image sequence respectively to obtain the first wet area information sequence and the second wet area information sequence.

[0041] In some embodiments, the aforementioned execution entity may perform wet area localization on the obtained fourth image sequence and third image sequence respectively to obtain a first wet area information sequence and a second wet area information sequence.

[0042] The first wet area information represents the location of the wet area corresponding to the wet area in the fourth image. This information may include: wet area location coordinates, a wet area identifier, and a wet area confidence score. The wet area location coordinates represent the coordinates of the identified wet area within the fourth image. Specifically, the wet area location coordinates can be represented by a sequence of edge coordinates corresponding to the wet area. The wet area identifier indicates whether a wet area exists in the fourth image; if no wet area exists, the wet area identifier can be represented by "0". If a wet area exists in the fourth image, the wet area identifier can be represented by "1". The wet area confidence score represents the confidence level when a wet area exists in the fourth image. The second wet area information represents the location of the wet area corresponding to the wet area in the third image. This information may include: wet area location coordinates, a wet area identifier, and a wet area confidence score. The wet area location coordinates represent the coordinates of the identified wet area within the third image. Specifically, the wet area location coordinates can be represented by a sequence of edge coordinates corresponding to the wet area. The wet area identifier indicates whether a wet area exists in the third image; if no wet area exists in the third image, the wet area identifier can be represented by "0". When a wet area exists in the third image, the wet area identifier can be represented by "1". The wet area confidence score represents the confidence level when a wet area exists in the third image. The wet area confidence score ranges from [0,1]. The first wet area information in the first wet area information sequence corresponds one-to-one with the fourth image in the fourth image sequence. Specifically, the number of first wet area information items in the first wet area information sequence is the same as the number of fourth images in the fourth image sequence. The second wet area information in the second wet area information sequence corresponds one-to-one with the third image in the third image sequence. Specifically, the number of second wet area information items in the second wet area information sequence is the same as the number of third images in the third image sequence.

[0043] In practice, wet area localization can be performed on each fourth image in the fourth image sequence and each third image in the third image sequence using object detection methods, such as the YOLO (You Only Look Once) model, to obtain the first wet area information sequence and the second wet area information sequence.

[0044] In some optional implementations of some embodiments, the execution entity performs wet area localization on the obtained fourth image sequence and the obtained third image sequence respectively to obtain a first wet area information sequence and a second wet area information sequence, including: Step S1: For each fourth image in the above fourth image sequence, perform the following first wet area localization step: Step S11: Extract color features from the fourth image to obtain the color features of the first fabric.

[0045] The first fabric color feature representation is based on the color feature expression of the fabric sample obtained from the fourth image.

[0046] In practice, the aforementioned execution entity extracts color features from the fourth image using a color feature encoder to obtain the color features of the first fabric. The color feature encoder consists of a backbone color feature encoding network and a color feature shaping network. These two networks are connected sequentially. The backbone color feature encoding network uses a ResNet-50 network. The color feature shaping network consists of three fully connected layers.

[0047] Step S12: Extract texture features from the fourth image to obtain the first fabric texture features.

[0048] The first fabric texture feature representation is based on the texture feature expression of the fabric sample obtained in the fourth image.

[0049] In practice, the aforementioned execution entity extracts texture features from the fourth image using a texture feature encoder to obtain the first fabric texture features. The texture feature encoder consists of a backbone texture feature encoding network and a texture feature shaping network. The backbone texture feature encoding network and the texture feature shaping network are connected sequentially. The backbone texture feature encoding network uses the MobileNet V3 network. The texture feature shaping network consists of three fully connected layers.

[0050] Step S13: Extract fluorescence features from the fourth image to obtain the fluorescence features of the first fabric.

[0051] The first fabric fluorescence feature characterization is based on the fluorescence feature expression of the fabric sample obtained from the fourth image. Since the test solution contains a fluorescent indicator, it emits specific fluorescence after being illuminated by a fluorescent lamp. Therefore, capturing the first fabric fluorescence feature can be used to assist in subsequent wet area localization. The feature dimensions of the first fabric color feature, the first fabric texture feature, and the first fabric fluorescence feature are the same.

[0052] In practice, the aforementioned execution entity extracts fluorescence features from the fourth image using a fluorescence feature encoder to obtain the first fabric fluorescence features. The fluorescence feature encoder consists of a backbone fluorescence feature encoding network and a fluorescence feature shaping network. The backbone fluorescence feature encoding network and the fluorescence feature shaping network are connected serially. The backbone fluorescence feature encoding network uses a ResNet-50 network. The fluorescence feature shaping network consists of three fully connected layers. The color feature encoder, texture feature encoder, and fluorescence feature encoder are configured in parallel.

[0053] Specifically, firstly, considering the different textile specifications and colors of the fabric samples, to ensure the accuracy of subsequent wet area localization, corresponding features are extracted from three specific dimensions: fabric color, fabric texture, and fabric fluorescence. Compared with conventional feature extraction methods, this significantly reduces the number of features and lowers the probability of overfitting. Furthermore, the encoders involved in feature extraction all employ lightweight network models to ensure a balance between extraction accuracy and speed. Next, by setting up color feature shaping networks, texture feature shaping networks, and fluorescence feature shaping networks, the feature dimensions of the first fabric color feature, the first fabric texture feature, and the first fabric fluorescence feature are ensured to be identical, facilitating subsequent feature fusion.

[0054] Step S14: Perform feature fusion on the color features, texture features, and fluorescence features of the first fabric to obtain the fused features of the first fabric.

[0055] The first fabric fusion feature is the fusion feature corresponding to the first fabric color feature, the aforementioned first fabric texture feature, and the aforementioned first fabric fluorescence feature.

[0056] In practice, the aforementioned execution entity can determine the channel weights corresponding to the first fabric color feature, the first fabric texture feature, and the first fabric fluorescence feature through a channel attention mechanism. Since the first fabric color feature, the first fabric texture feature, and the first fabric fluorescence feature have the same feature dimensions, a weighted feature superposition method can be used to obtain the first fabric fusion feature. Specifically, the channel weights are learned through two fully connected layers.

[0057] Step S15: In response to the image position corresponding to the fourth image being the first position of the sequence, multi-scale feature extraction is performed on the first fabric fusion feature corresponding to the fourth image to obtain the first multi-scale fabric fusion feature.

[0058] Among them, the first multi-scale fabric fusion feature is a feature expression under multiple receptive fields.

[0059] In practice, the first multi-scale fabric fusion feature can be obtained by using the FPN (Feature Pyramid Networks) model to extract multi-scale features from the first fabric fusion feature corresponding to the fourth image.

[0060] Step S16: In response to the non-sequence first position of the image position corresponding to the fourth image, perform multi-scale feature extraction on the target first fabric fusion feature set and the first fabric fusion feature corresponding to the fourth image to obtain the first multi-scale fabric fusion feature.

[0061] Wherein, assuming that the image position of the above fourth image in the fourth image sequence is k (k≥2), then the target first fabric fusion feature in the target first fabric fusion feature set is the first fabric fusion feature corresponding to all the fourth images from image position k-1 to the previous one.

[0062] In practice, when the image position corresponding to the fourth image is not the first position of the sequence, the ConvLSTM (Convolutional Long Short Term Memory) model is first used to extract the temporal features of the target first fabric fusion feature set and the first fabric fusion feature corresponding to the fourth image. The result of the temporal feature extraction is used as the input of the FPN model to obtain the first multi-scale fabric fusion feature, thereby capturing the temporal feature expression in the diffusion process of the test liquid.

[0063] Step S17: Decode the features based on the first multi-scale fabric fusion features to obtain the first wet area mask corresponding to the fourth image above.

[0064] The first wet area mask represents the image mask corresponding to the wet area in the fourth image.

[0065] In practice, the first multi-scale fabric fusion features can be used as input to the decoding network to obtain the first wet area mask corresponding to the fourth image. The decoding network uses the U-Net model as its backbone and also includes a mask localization head and a confidence regression head. The mask localization head uses the Softmax function to regress the edge coordinates. The confidence regression head uses the Sigmoid function to perform confidence regression for each edge coordinate.

[0066] Step S18: Based on the first wet area mask, generate the first wet area information corresponding to the fourth image in the first wet area information sequence.

[0067] In practice, when the first wet region mask is empty, the wet region information corresponding to the fourth image can include a wet region identifier of "0", and the wet region location coordinates and wet region confidence scores can both be empty (NULL). When the first wet region mask is not empty, the wet region information corresponding to the fourth image can include a wet region identifier of "1", the wet region location coordinates can be the mask boundary coordinates corresponding to the first wet region mask, and the wet region confidence score can be the confidence score result output by the confidence score regression head.

[0068] In some optional implementations of some embodiments, the execution entity performs wet area localization on the obtained fourth image sequence and the obtained third image sequence to obtain a first wet area information sequence and a second wet area information sequence, and further includes: Step S2: For each third image in the above third image sequence, perform the following second wet area localization step: Step S21: Extract features from the third image to obtain the second fabric color features, the second fabric texture features, and the second fabric fluorescence features.

[0069] The second fabric color feature is a color feature representation of the fabric sample obtained based on the third image. The second fabric texture feature is a texture feature representation of the fabric sample obtained based on the third image. The second fabric fluorescence feature is a fluorescence feature representation of the fabric sample obtained based on the third image. The feature dimensions of the second fabric color feature, the second fabric texture feature, and the second fabric fluorescence feature are the same.

[0070] In practice, feature extraction can be performed in parallel using a third image as input through a color feature encoder, a texture feature encoder, and a fluorescence feature encoder to obtain the second fabric color features, the second fabric texture features, and the second fabric fluorescence features.

[0071] Step S22: Perform feature fusion on the above-mentioned color features, texture features and fluorescence features of the second fabric to obtain the fused features of the second fabric.

[0072] The second fabric fusion feature is the fusion feature corresponding to the second fabric color feature, the aforementioned second fabric texture feature, and the aforementioned second fabric fluorescence feature.

[0073] In practice, the aforementioned execution entity can determine the channel weights corresponding to the second fabric color features, the second fabric texture features, and the second fabric fluorescence features through a channel attention mechanism. Since the feature dimensions of the second fabric color features, the second fabric texture features, and the second fabric fluorescence features are the same, a weighted feature superposition method can be used to obtain the second fabric fusion features. Specifically, the channel weights are learned through two fully connected layers.

[0074] Step S23: In response to the image position corresponding to the third image being the first position of the sequence, multi-scale feature extraction is performed on the second fabric fusion feature corresponding to the third image to obtain the second multi-scale fabric fusion feature.

[0075] Among them, the second multi-scale fabric fusion feature is a feature expression under multiple receptive fields.

[0076] In practice, the second multi-scale fabric fusion feature can be obtained by using the FPN (Feature Pyramid Networks) model to extract multi-scale features from the second fabric fusion feature corresponding to the third image.

[0077] Step S24: In response to the non-sequence first position of the image position corresponding to the third image, perform multi-scale feature extraction on the target second fabric fusion feature set and the second fabric fusion feature corresponding to the third image to obtain the second multi-scale fabric fusion feature.

[0078] Wherein, assuming that the image position of the above third image in the third image sequence is k (k≥2), then the target second fabric fusion feature in the target second fabric fusion feature set is the second fabric fusion feature corresponding to all the third images from image position k-1 to the previous one.

[0079] In practice, when the image position corresponding to the third image is not the first position of the sequence, the ConvLSTM (Convolutional Long Short Term Memory) model is first used to extract the temporal features of the target second fabric fusion feature set and the second fabric fusion feature corresponding to the third image. The result of the temporal feature extraction is used as the input of the FPN model to obtain the second multi-scale fabric fusion feature, thereby capturing the temporal feature expression in the diffusion process of the test liquid.

[0080] Step S25: Perform mask flipping on the first wet area mask corresponding to the third image above to obtain the flipped wet area mask.

[0081] Since there is a flip-mirror relationship between the third image and the corresponding fourth image, and the test liquid diffuses from the center of the fabric sample to the edge, the first wet area mask can be used for auxiliary positioning of the wet area corresponding to the third image.

[0082] In practice, the first wet area mask can be flipped and mirrored to obtain the flipped wet area mask.

[0083] Step S26: Apply Gaussian blur to the above-mentioned flipped wet area mask to obtain the blurred wet area mask.

[0084] In practice, due to the penetration of the test liquid into the fabric sample, the wet area identified in the fourth image is not exactly the same size as the wet area identified in the corresponding third image. Therefore, Gaussian blur is used to expand the mask boundary based on flipping the wet area mask, thereby enhancing the accuracy of subsequent wet area localization. At the same time, the feature representation of non-interest areas is weakened, reducing the amount of subsequent unnecessary feature calculations and improving the speed of wet area localization.

[0085] Step S27: Based on the above blurred wet area mask, perform local feature enhancement on the above second multi-scale fabric fusion feature to obtain the third multi-scale fabric fusion feature.

[0086] In practice, a weight matrix can be constructed based on the blurred wet area mask. In this matrix, the value corresponding to the blurred wet area mask is 1, and the value decreases with distance from the mask. Therefore, the weight matrix can be updated using distance decay in conjunction with the blurred wet area mask. The matrix dimension of the weight matrix is ​​the same as the feature dimension of the second multi-scale fabric fusion feature. Therefore, the second multi-scale fabric fusion feature can be updated with weights using the weight matrix to achieve local feature enhancement, thereby obtaining the third multi-scale fabric fusion feature.

[0087] Step S28: Perform feature decoding based on the third multi-scale fabric fusion features to obtain the second wet area mask corresponding to the third image.

[0088] The second wet area mask represents the image mask corresponding to the wet area in the third image.

[0089] In practice, the third multi-scale fabric fusion features can be used as input to the decoding network to obtain the second wet area mask corresponding to the third image. The decoding network uses the U-Net model as its backbone and also includes a mask localization head and a confidence regression head. The mask localization head uses the Softmax function to regress the edge coordinates. The confidence regression head uses the Sigmoid function to perform confidence regression for each edge coordinate.

[0090] Step S29: Based on the second wet area mask, generate the second wet area information corresponding to the third image in the second wet area information sequence.

[0091] In practice, when the second wet region mask is empty, the wet region information corresponding to the third image can include a wet region identifier of "0", and the wet region location coordinates and wet region confidence scores can both be empty (NULL). When the second wet region mask is not empty, the wet region information corresponding to the third image can include a wet region identifier of "1", the wet region location coordinates can be the mask boundary coordinates corresponding to the second wet region mask, and the wet region confidence score can be the confidence score result output by the confidence score regression head.

[0092] Step 504: Identify water absorption based on the first wet zone information sequence to obtain water absorption performance description information.

[0093] In some embodiments, the aforementioned execution entity can identify water absorption based on the first wet zone information sequence to obtain water absorption performance description information.

[0094] The water absorption performance description information includes: dynamic water absorption trend information and dynamic diffusion trend information. Dynamic water absorption trend information characterizes the dynamic water absorption rate of the fabric. Dynamic diffusion trend information characterizes the diffusion state of the test liquid on the surface of the fabric sample.

[0095] In practice, the area change of the wetland region can be determined by combining the location coordinates of the wetland region included in the first wetland region information sequence, thereby calculating the dynamic water absorption rate and the corresponding diffusion trend rate, which can be used as dynamic water absorption trend information and dynamic diffusion trend information.

[0096] In some optional implementations of certain embodiments, the execution entity performs water absorption identification based on the first wet zone information sequence to obtain water absorption performance description information, including: Step S1: For each piece of information about the first wetland in the above first wetland information sequence, generate the area of ​​the first wetland region based on the above first wetland information to obtain the first wetland area area sequence.

[0097] Among them, the area of ​​the first wetland is the area of ​​the wetland corresponding to the information of the first wetland.

[0098] In practice, the mask area of ​​the first wet area mask corresponding to the first wet area information can be used as the area of ​​the first wet area region.

[0099] Step S2: Based on the area of ​​each two adjacent first wet areas in the above first wet area area sequence, generate discrete moisture absorption rate values ​​to obtain a discrete moisture absorption rate value sequence.

[0100] The discrete moisture absorption rate value represents the rate of area change between two adjacent first wet zone regions. Discrete moisture absorption rate value = area difference between two adjacent first wet zone regions / image acquisition time difference. The image acquisition time difference represents the time difference between acquiring two adjacent first images. Image acquisition time difference = unit time / frame rate. For example, the unit time can be 1 second.

[0101] Step S3: Fit the rate curve based on the discrete moisture absorption rate value sequence above to obtain the moisture absorption rate curve.

[0102] The moisture absorption rate curve represents the change in moisture absorption rate after the test droplet is applied to the fabric sample.

[0103] In practice, the least squares method can be used to fit the rate curve based on the discrete moisture absorption rate value sequence mentioned above, thus obtaining the moisture absorption rate curve. Since the cameras are all high-speed cameras, multiple nearly continuous discrete moisture absorption rate values ​​will be obtained. Compared with the method of evaluating moisture absorption and drying speed based on the resistance change corresponding to the contact surface of discrete probes, this method can express the dynamic diffusion changes of the test liquid more cost-effectively and accurately.

[0104] Step S4: Based on the above moisture absorption rate curve, generate dynamic moisture absorption trend information, including the moisture absorption performance description information.

[0105] In practice, the average and maximum moisture absorption rates within each unit time window can be determined based on the moisture absorption rate curve, serving as dynamic moisture absorption trend information.

[0106] Step S5: For each piece of information about the first wetland in the above sequence of information about the first wetland, determine the diffusion trend characteristics based on the above information about the first wetland.

[0107] The diffusion trend characteristics include: diffusion uniformity characteristics and diffusion directionality characteristics. Diffusion uniformity characteristics characterize the diffusion uniformity of the test liquid. Diffusion directionality characteristics characterize the diffusion direction of the test liquid.

[0108] In practice, when the test solution comes into contact with the fabric sample, it diffuses from the center of the fabric sample outwards. Therefore, firstly, the distance from the edge coordinates of the first wet zone to the center of the fabric sample can be calculated. Specifically, 360 diffusion distance values ​​from the center of the fabric sample to the boundary of the first wet zone can be determined at 1-degree angular intervals, serving as the diffusion directionality feature. That is, the feature dimension of the diffusion directionality feature is 1×360. The angular interval can be flexibly adjusted according to the actual calculation accuracy and efficiency requirements. Next, based on the diffusion directionality feature, the mean diffusion distance and the standard deviation of the diffusion distance are determined. Diffusion uniformity feature = 1 - Standard deviation of diffusion distance / Mean diffusion distance.

[0109] Step S6: Based on the obtained diffusion trend feature sequence, generate dynamic diffusion trend information including water absorption performance description information.

[0110] In practice, the obtained diffusion trend feature sequence can be used as dynamic diffusion trend information included in the description information of water absorption performance.

[0111] Step 505: Based on the first wet area information sequence and the second wet area information sequence, quick-drying performance is identified to obtain quick-drying performance description information.

[0112] In some embodiments, the aforementioned execution entity can perform quick-drying identification based on the first wet area information sequence and the second wet area information sequence to obtain quick-drying performance description information.

[0113] The quick-drying performance description information includes: drying half-life. The drying half-life characterizes the length of time from the start of the maximum wet area on the skin-friendly side and the non-skin-friendly side of the fabric sample until the wet area is ≤ 0.5 × maximum wet area.

[0114] In practice, firstly, for the skin-friendly side of the fabric sample, by combining the first wet area information sequence, the first wet area information corresponding to the maximum wet area area can be determined, as well as the first corresponding wet area area ≤ 0.5 × maximum wet area starting from the first wet area information corresponding to the maximum wet area area. Therefore, the image acquisition interval of the two first images or the two corresponding second images corresponding to these two first wet area information is taken as the drying half-life corresponding to the skin-friendly side of the fabric sample. Next, for the non-skin-friendly side of the fabric sample, by combining the second wet area information sequence, the second wet area information corresponding to the maximum wet area can be determined, as well as the first corresponding second wet area area ≤ 0.5 × maximum wet area starting from the second wet area information corresponding to the maximum wet area area. Therefore, the image acquisition interval of the two third images corresponding to these two second wet area information is taken as the drying half-life corresponding to the non-skin-friendly side of the fabric sample.

[0115] In some optional implementations of certain embodiments, the execution entity performs quick-drying performance identification based on the first wet area information sequence and the second wet area information sequence to obtain quick-drying performance description information, including: Step S1: For each piece of information about the second wetland in the above second wetland information sequence, generate the area of ​​the second wetland region based on the above second wetland information to obtain the area sequence of the second wetland region.

[0116] The area of ​​the second wetland is the area of ​​the wetland corresponding to the information of the second wetland.

[0117] In practice, the mask area of ​​the second wet area mask corresponding to the second wet area information can be used as the area of ​​the second wet area region.

[0118] Step S2: Determine the peak area of ​​the first wetland and the peak area of ​​the second wetland based on the above-mentioned area sequence of the first wetland and area sequence of the second wetland.

[0119] The peak area of ​​the first wetland is the area of ​​the first wetland region with the largest area value in the area sequence of the first wetland regions. The peak area of ​​the second wetland region is the area of ​​the second wetland region with the largest area value in the area sequence of the second wetland regions.

[0120] In practice, the peak areas of the first and second wetlands can be determined by iterating through the area values.

[0121] Step S3: Determine the first candidate drying half-life based on the above-mentioned first wet region area sequence and the above-mentioned first wet region peak area.

[0122] In practice, for the first wet region area that meets the filtering condition and where the peak area of ​​the first wet region begins to appear, the image acquisition time difference between the first or second image corresponding to the peak area of ​​the first wet region and the first or second image corresponding to the wet region area that meets the filtering condition is used as the first candidate drying half-life. The filtering condition is: wet region area ≤ 0.5 × peak area of ​​the first wet region.

[0123] Step S4: Determine the second candidate drying half-life based on the above-mentioned second wet region area sequence and the above-mentioned second wet region peak area.

[0124] In practice, for the first second wet region area that meets the filtering condition where the peak area of ​​the second wet region begins to appear, the image acquisition time difference between the third image corresponding to the peak area of ​​the second wet region and the third image corresponding to the second wet region area that meets the filtering condition is used as the second candidate drying half-life. The filtering condition is: wet region area ≤ 0.5 × peak area of ​​the second wet region.

[0125] Step S5: Based on the first candidate drying half-life and the second candidate drying half-life, generate the drying half-life included in the quick-drying performance description information.

[0126] In practice, the aforementioned implementing entity may use the first candidate drying half-life and the second candidate drying half-life as the drying half-life included in the quick-drying performance description information.

[0127] In some optional implementations of some embodiments, the above method further includes: Step S1: Extract the second wetland information that meets the filtering conditions from the above second wetland information sequence and use it as the target wetland information.

[0128] Among them, the target wetland information is the first wetland information in the second wetland information sequence that is identified as "1".

[0129] Step S2: Generate the liquid penetration time based on the image acquisition time of the third image corresponding to the above target wet area information.

[0130] Wherein, liquid penetration time = image acquisition time of the third image corresponding to the target wet area information - time of first contact between the test liquid and the fabric sample.

[0131] Step S3: Perform pseudo-humidity value mapping based on the first wet area information sequence and the second wet area information sequence to obtain the first dynamic humidity cloud map and the second dynamic humidity cloud map.

[0132] The first dynamic humidity cloud map represents the change in the wet area on the skin-friendly side of the fabric sample. The second dynamic humidity cloud map represents the change in the wet area on the non-skin-friendly side of the fabric sample.

[0133] In practice, since the information of the first wetland area includes the location coordinates of the wetland area and the information of the second wetland area also includes the location coordinates of the wetland area, dynamic cloud maps that represent the changes in the wetland area can be constructed by combining the location coordinates of the wetland area in a time sequence. These dynamic cloud maps serve as the first dynamic humidity cloud map and the second dynamic humidity cloud map.

[0134] The above embodiments of this disclosure have the following beneficial effects: The dynamic image-based fabric moisture absorption and quick-drying property identification method of some embodiments of this disclosure achieves accurate identification of the dynamic moisture absorption and quick-drying property of the fabric sample. Specifically, firstly, in response to the fabric sample being fixed on the sample stage and the pumping device starting to pump the test liquid, parallel image acquisition is performed using a first camera, a second camera, and a third camera to obtain a first image sequence, a second image sequence, and a third image sequence. This, combined with a high-speed camera, quickly captures the diffusion, penetration, and drying of the test liquid in the fabric sample. Secondly, for each first image in the first image sequence, the first image and its corresponding second image in the second image sequence are stitched together to obtain a fourth image. In practice, considering the need to avoid positional interference between the camera and the dispensing needle when identifying the fabric's moisture absorption and quick-drying property, the corresponding images are acquired using a first camera, a second camera, and a V-shaped reflector. Therefore, before identification, the first image and the corresponding second image need to be stitched together to form a complete image of the fabric sample. Next, wet zone localization is performed on the obtained fourth image sequence and the aforementioned third image sequence to obtain a first wet zone information sequence and a second wet zone information sequence. This allows for wet zone localization from both the skin-friendly and non-skin-friendly sides of the fabric sample, thereby determining the diffusion changes after the test liquid comes into contact with the textile sample. Further, water absorption is identified based on the aforementioned first wet zone information sequence to obtain water absorption performance description information, which includes dynamic water absorption trend information and dynamic diffusion trend information. By combining continuous first wet zone information, the dynamic moisture absorption performance of the fabric sample is quantified. Finally, quick-drying performance is identified based on the aforementioned first and second wet zone information sequences to obtain quick-drying performance description information, which includes the drying half-life. By combining continuous first and second wet zone information, the dynamic quick-drying performance of the fabric sample is quantified. This method achieves accurate identification of the dynamic moisture absorption and quick-drying performance of the fabric sample.

[0135] Further reference Figure 6As an implementation of the methods shown in the above figures, this disclosure provides some embodiments of a fabric moisture-wicking and quick-drying property identification device based on dynamic images. These device embodiments are similar to... Figure 5 Corresponding to the method embodiments shown, this dynamic image-based fabric moisture absorption and quick-drying identification device can be specifically applied to various electronic devices.

[0136] like Figure 6 As shown, a fabric moisture-wicking and quick-drying identification device 600 based on dynamic images in some embodiments includes: an image acquisition unit 601, an image stitching unit 602, a wet area positioning unit 603, a water-wicking identification unit 604, and a quick-drying identification unit 605. The image acquisition unit 601 is configured to, in response to a fabric sample being fixed on a sample stage and a pumping device starting to pump test liquid, acquire images in parallel using a first camera, a second camera, and a third camera to obtain a first image sequence, a second image sequence, and a third image sequence. The image stitching unit 602 is configured to, for each first image in the first image sequence, stitch the first image and the first image into the second image sequence. The corresponding second image is stitched together to obtain a fourth image; the wet area positioning unit 603 is configured to perform wet area positioning on the obtained fourth image sequence and the aforementioned third image sequence respectively to obtain a first wet area information sequence and a second wet area information sequence; the water absorption recognition unit 604 is configured to perform water absorption recognition based on the aforementioned first wet area information sequence to obtain water absorption performance description information, wherein the aforementioned water absorption performance description information includes: dynamic water absorption trend information and dynamic diffusion trend information; the quick-drying recognition unit 605 is configured to perform quick-drying recognition based on the aforementioned first wet area information sequence and the aforementioned second wet area information sequence to obtain quick-drying performance description information, wherein the aforementioned quick-drying performance description information includes: drying half-life.

[0137] It is understandable that the units described in the dynamic image-based fabric moisture-wicking and quick-drying recognition device 600 are similar to the reference units. Figure 5 The steps in the described method correspond accordingly. Therefore, the operations, features, and beneficial effects described above for the method also apply to the fabric moisture-wicking and quick-drying property recognition device 600 based on dynamic images and the units contained therein, and will not be repeated here.

[0138] The following is for reference. Figure 7 It shows a schematic diagram of the structure of an electronic device (e.g., a computing device) 700 suitable for implementing some embodiments of the present disclosure. Figure 7 The electronic device shown is merely an example and should not be construed as limiting the functionality and scope of the embodiments of this disclosure.

[0139] like Figure 7As shown, the electronic device 700 may include a processing unit (e.g., a central processing unit, a graphics processing unit, etc.) 701, which can perform various appropriate actions and processes according to a program stored in a read-only memory 702 or a program loaded from a storage device 708 into a random access memory 703. The random access memory 703 also stores various programs and data required for the operation of the electronic device 700. The processing unit 701, the read-only memory 702, and the random access memory 703 are interconnected via a bus 704. An input / output interface 705 is also connected to the bus 704.

[0140] Typically, the following devices can be connected to the input / output interface 705: input devices 706 including, for example, a touchscreen, touchpad, keyboard, mouse, camera, microphone, accelerometer, gyroscope, etc.; output devices 707 including, for example, a liquid crystal display (LCD), speaker, vibrator, etc.; storage devices 708 including, for example, magnetic tape, hard disk, etc.; and communication devices 709. Communication device 709 allows electronic device 700 to communicate wirelessly or wiredly with other devices to exchange data. Although Figure 7 An electronic device 700 with various devices is shown; however, it should be understood that it is not required to implement or possess all of the devices shown. More or fewer devices may be implemented or possessed alternatively. Figure 7 Each box shown can represent a device or multiple devices as needed.

[0141] In particular, according to some embodiments of this disclosure, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, some embodiments of this disclosure include a computer program product comprising a computer program carried on a computer-readable medium, the computer program containing program code for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via a communication device 709, or installed from a storage device 708, or installed from a read-only memory 702. When the computer program is executed by the processing device 701, it performs the functions defined in the methods of some embodiments of this disclosure.

[0142] It should be noted that, in some embodiments of this disclosure, the computer-readable medium may be a computer-readable signal medium or a computer-readable storage medium, or any combination thereof. A computer-readable storage medium may be, for example,—but not limited to—an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples of a computer-readable storage medium may include, but are not limited to: an electrical connection having one or more wires, a portable computer disk, a hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage device, magnetic storage device, or any suitable combination thereof. In some embodiments of this disclosure, a computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, apparatus, or device. In some embodiments of this disclosure, a computer-readable signal medium may include a data signal propagated in baseband or as part of a carrier wave, carrying computer-readable program code. Such propagated data signals may take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof. A computer-readable signal medium can be any computer-readable medium other than a computer-readable storage medium, which can send, propagate, or transmit a program for use by or in connection with an instruction execution system, apparatus, or device. The program code contained on the computer-readable medium can be transmitted using any suitable medium, including but not limited to: wires, optical fibers, RF (radio frequency), etc., or any suitable combination thereof.

[0143] In some implementations, clients and servers can communicate using any currently known or future-developed network protocol such as HTTP (Hypertext Transfer Protocol) and can interconnect with digital data communication (e.g., communication networks) of any form or medium. Examples of communication networks include local area networks (“LANs”), wide area networks (“WANs”), the Internet (e.g., the Internet of Things), and peer-to-peer networks (e.g., ad hoc peer-to-peer networks), as well as any currently known or future-developed networks.

[0144] The aforementioned computer-readable medium may be included in the aforementioned electronic device; or it may exist independently and not assembled into the electronic device. The aforementioned computer-readable medium carries one or more programs. When the electronic device executes the aforementioned one or more programs, the electronic device causes the following actions: in response to the fabric sample being fixed on the sample stage and the pumping device starting to pump the test liquid, the electronic device performs parallel image acquisition via a first camera, a second camera, and a third camera to obtain a first image sequence, a second image sequence, and a third image sequence; for each first image in the aforementioned first image sequence, the first image and the corresponding second image in the aforementioned second image sequence are stitched together to obtain a fourth image; the fourth image sequence and the third image sequence are respectively subjected to wet area localization to obtain a first wet area information sequence and a second wet area information sequence; water absorption is identified based on the aforementioned first wet area information sequence to obtain water absorption performance description information, wherein the water absorption performance description information includes: dynamic water absorption trend information and dynamic diffusion trend information; quick-drying performance is identified based on the aforementioned first wet area information sequence and the aforementioned second wet area information sequence to obtain quick-drying performance description information, wherein the quick-drying performance description information includes: drying half-life.

[0145] Computer program code for performing operations of some embodiments of this disclosure can be written in one or more programming languages ​​or a combination thereof, including object-oriented programming languages ​​such as Java, Smalltalk, C++, and Python, and conventional procedural programming languages ​​such as the "C" language or similar programming languages. The program code can be executed entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving remote computers, the remote computer can be connected to the user's computer via any type of network—including a local area network (LAN) or a wide area network (WAN)—or can be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0146] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of this disclosure. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two consecutively indicated blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, can be implemented using a dedicated hardware-based system that performs the specified function or operation, or using a combination of dedicated hardware and computer instructions.

[0147] The functions described above in this document can be performed at least in part by one or more hardware logic components. For example, exemplary types of hardware logic components that can be used, without limitation, include: field-programmable gate arrays (FPGAs), application-specific integrated circuits (ASICs), application-specific standard products (ASSPs), system-on-a-chip (SoCs), complex programmable logic devices (CPLDs), and so on.

[0148] The above description is merely a selection of preferred embodiments of this disclosure and an explanation of the technical principles employed. Those skilled in the art should understand that the scope of the invention involved in the embodiments of this disclosure is not limited to technical solutions formed by specific combinations of the above-described technical features, but should also cover other technical solutions formed by arbitrary combinations of the above-described technical features or their equivalents without departing from the above-described inventive concept. For example, technical solutions formed by substituting the above-described features with (but not limited to) technical features with similar functions disclosed in the embodiments of this disclosure.

Claims

1. A method for identifying the moisture absorption and quick-drying properties of fabrics based on dynamic images, characterized in that, include: In response to the fabric sample being fixed on the sample stage and the pumping device starting to pump the test liquid, parallel image acquisition is performed through the first camera, the second camera and the third camera to obtain the first image sequence, the second image sequence and the third image sequence. For each first image in the first image sequence, the first image and the corresponding second image in the second image sequence are stitched together to obtain the fourth image; Wet zone localization is performed on the obtained fourth image sequence and the third image sequence respectively to obtain the first wet zone information sequence and the second wet zone information sequence; Based on the first wet area information sequence, water absorption is identified to obtain water absorption performance description information, wherein the water absorption performance description information includes: dynamic water absorption trend information and dynamic diffusion trend information; Based on the first wet area information sequence and the second wet area information sequence, quick-drying performance is identified to obtain quick-drying performance description information, wherein the quick-drying performance description information includes: drying half-life.

2. The method for identifying the moisture absorption and quick-drying properties of fabrics based on dynamic images according to claim 1, characterized in that, The step of identifying water absorption capacity based on the first wet area information sequence to obtain water absorption performance description information includes: For each piece of information about a first wetland in the first wetland information sequence, the area of ​​the first wetland region is generated based on the first wetland information, resulting in a sequence of areas of the first wetland region. Based on the area of ​​each two adjacent first wetlands in the first wetland area sequence, a discrete moisture absorption rate value is generated to obtain a discrete moisture absorption rate value sequence. Based on the discrete moisture absorption rate value sequence, a rate curve is fitted to obtain the moisture absorption rate curve; Based on the moisture absorption rate curve, dynamic moisture absorption trend information, including moisture absorption performance description information, is generated. For each piece of information about a first wetland in the first wetland information sequence, diffusion trend characteristics are determined based on the first wetland information, wherein the diffusion trend characteristics include: diffusion uniformity characteristics and diffusion direction characteristics; Based on the obtained diffusion trend feature sequence, dynamic diffusion trend information is generated to describe the water absorption performance.

3. The method for identifying the moisture absorption and quick-drying properties of fabrics based on dynamic images according to claim 2, characterized in that, The step of identifying quick-drying properties based on the first wetland information sequence and the second wetland information sequence to obtain quick-drying performance description information includes: For each piece of information about the second wetland in the second wetland information sequence, the area of ​​the second wetland region is generated based on the second wetland information, resulting in a sequence of areas of the second wetland region. Based on the area sequence of the first wetland region and the area sequence of the second wetland region, determine the peak area of ​​the first wetland region and the peak area of ​​the second wetland region; The first candidate drying half-life is determined based on the area sequence of the first wetland region and the peak area of ​​the first wetland region. The second candidate drying half-life is determined based on the area sequence of the second wet region and the peak area of ​​the second wet region; Based on the first candidate drying half-life and the second candidate drying half-life, the drying half-life is used to generate quick-drying performance description information including the drying half-life.

4. The method for identifying the moisture absorption and quick-drying properties of fabrics based on dynamic images according to claim 3, characterized in that, The method further includes: Extract the second wetland information that meets the filtering conditions from the second wetland information sequence, and use it as the target wetland information; Based on the image acquisition time of the third image corresponding to the target wet area information, the liquid penetration time is generated; Pseudo-humidity values ​​are mapped based on the first and second wet area information sequences, respectively, to obtain a first dynamic humidity cloud map and a second dynamic humidity cloud map.

5. The method for identifying the moisture absorption and quick-drying properties of fabrics based on dynamic images according to claim 1, characterized in that, The step of performing wet area localization on the obtained fourth image sequence and the third image sequence respectively to obtain a first wet area information sequence and a second wet area information sequence includes: For each fourth image in the fourth image sequence, perform the following first wet area localization step: Color features are extracted from the fourth image to obtain the color features of the first fabric; Texture features are extracted from the fourth image to obtain the first fabric texture features; Fluorescence features were extracted from the fourth image to obtain the fluorescence features of the first fabric; The color features, texture features, and fluorescence features of the first fabric are fused to obtain the first fabric fused features. In response to the image position corresponding to the fourth image being the first position of the sequence, multi-scale feature extraction is performed on the first fabric fusion feature corresponding to the fourth image to obtain the first multi-scale fabric fusion feature. In response to the non-sequence first position of the image position corresponding to the fourth image, multi-scale feature extraction is performed on the target first fabric fusion feature set and the first fabric fusion feature corresponding to the fourth image to obtain the first multi-scale fabric fusion feature. Feature decoding is performed based on the first multi-scale fabric fusion features to obtain the first wet area mask corresponding to the fourth image; Based on the first wet area mask, the first wet area information corresponding to the fourth image is generated in the first wet area information sequence.

6. The method for identifying the moisture absorption and quick-drying properties of fabrics based on dynamic images according to claim 2, characterized in that, The step of performing wet area localization on the obtained fourth image sequence and the third image sequence respectively to obtain a first wet area information sequence and a second wet area information sequence further includes: For each third image in the third image sequence, perform the following second wet area localization step: Feature extraction is performed on the third image to obtain the color features, texture features, and fluorescence features of the second fabric. The color features, texture features, and fluorescence features of the second fabric are fused to obtain the fused features of the second fabric. In response to the image position corresponding to the third image being the first position of the sequence, multi-scale feature extraction is performed on the second fabric fusion feature corresponding to the third image to obtain the second multi-scale fabric fusion feature. In response to the non-sequence first position of the image position corresponding to the third image, multi-scale feature extraction is performed on the target second fabric fusion feature set and the second fabric fusion feature corresponding to the third image to obtain the second multi-scale fabric fusion feature. The first wet area mask corresponding to the third image is flipped to obtain the flipped wet area mask; Gaussian blurring is applied to the inverted wet area mask to obtain the blurred wet area mask; Based on the blurred wet area mask, the second multi-scale fabric fusion feature is locally enhanced to obtain the third multi-scale fabric fusion feature. Based on the third multi-scale fabric fusion features, feature decoding is performed to obtain the second wet area mask corresponding to the third image; Based on the second wet area mask, the second wet area information corresponding to the third image is generated in the second wet area information sequence.

7. A fabric moisture-wicking and quick-drying property identification system, applied to the fabric moisture-wicking and quick-drying property identification method based on dynamic images as described in any one of claims 1 to 6, wherein, The fabric moisture absorption and quick-drying property identification system includes: an upper fixing component, a lower placement component, a test liquid storage tank, a pumping device, a first camera, a second camera, a third camera, and a control component, characterized in that it includes: The upper fixing component includes: an upper fixing component housing, a first fixing pin set, a test liquid conduit, a first reflector, and a second reflector. The first fixing pin is a cylindrical pin with a circular pin hole. The first fixing pins in the first fixing pin set are located at the four corners of the first surface of the upper fixing component housing. The test liquid conduit is disposed inside the upper fixing component housing. The first end of the test liquid conduit is connected to a drop needle. The mirror surface angle between the first reflector and the second reflector is 90 degrees. The plane formed by the first mirror surface edge of the first reflector and the first mirror surface edge of the second reflector is coplanar with the first surface of the upper fixing component housing. The drop needle passes through the center of the intersection line of the mirror surfaces of the first reflector and the second reflector. The lower placement component includes: a sample placement stage and a second set of fixing pins, wherein the sample placement stage is made of transparent material, the second fixing pin is a cylindrical pin, the second fixing pin in the second set of fixing pins is disposed at the four corners of the second surface of the sample placement stage, and the first fixing pin in the first set of fixing pins and the second fixing pin in the second set of fixing pins correspond one-to-one. The pumping device is disposed between the test liquid storage tank and the upper fixing assembly. The pumping device and the test liquid storage tank are connected by a liquid infusion line. The test liquid conduit included in the pumping device and the upper fixing assembly is connected by a liquid infusion line. The test liquid storage tank stores simulated sweat containing a fluorescent indicator. The first camera and the second camera are arranged facing each other, wherein the first camera faces the center of the first reflector and the second camera faces the center of the second reflector. The third camera faces the center of the first surface of the sample stage, wherein the first camera, the second camera, and the third camera are all high-speed cameras; The control component is communicatively connected to the pumping device, the first camera, the second camera, and the third camera.

8. A fabric moisture absorption and quick-drying property identification device based on dynamic imaging, characterized in that, include: The image acquisition unit is configured to perform parallel image acquisition through the first camera, the second camera and the third camera in response to the fabric sample being fixed on the sample stage and the pumping device starting to pump the test liquid, to obtain the first image sequence, the second image sequence and the third image sequence. The image stitching unit is configured to stitch together the first image and the corresponding second image in the second image sequence for each first image in the first image sequence to obtain a fourth image; The wet area localization unit is configured to perform wet area localization on the obtained fourth image sequence and the third image sequence respectively, to obtain a first wet area information sequence and a second wet area information sequence; The water absorption identification unit is configured to identify water absorption based on the first wet area information sequence to obtain water absorption performance description information, wherein the water absorption performance description information includes: dynamic water absorption trend information and dynamic diffusion trend information; The quick-drying performance identification unit is configured to perform quick-drying performance identification based on the first wet area information sequence and the second wet area information sequence to obtain quick-drying performance description information, wherein the quick-drying performance description information includes: drying half-life.

9. An electronic device, characterized in that, include: One or more processors; A storage device on which one or more programs are stored; When the one or more programs are executed by the one or more processors, the one or more processors implement the method as described in any one of claims 1 to 6 and the system as described in claim 7.

10. A computer-readable medium, characterized in that, It stores a computer program thereon, wherein the computer program, when executed by a processor, implements the method as described in any one of claims 1 to 6 and the system as described in claim 7.

Images