WO2021098486A1 - 衣物颜色识别的处理方法、装置、设备及存储介质 - Google Patents

衣物颜色识别的处理方法、装置、设备及存储介质 Download PDF

Info

Publication number
WO2021098486A1
WO2021098486A1 PCT/CN2020/125427 CN2020125427W WO2021098486A1 WO 2021098486 A1 WO2021098486 A1 WO 2021098486A1 CN 2020125427 W CN2020125427 W CN 2020125427W WO 2021098486 A1 WO2021098486 A1 WO 2021098486A1
Authority
WO
WIPO (PCT)
Prior art keywords
image data
laundry
washing machine
color
recognition
Prior art date
Application number
PCT/CN2020/125427
Other languages
English (en)
French (fr)
Inventor
赵龙
许升
黄振兴
刘一丁
丁晓鹏
Original Assignee
重庆海尔洗衣机有限公司
海尔智家股份有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Priority claimed from CN201911152484.0A external-priority patent/CN112836706A/zh
Priority claimed from CN201911152480.2A external-priority patent/CN112831982A/zh
Application filed by 重庆海尔洗衣机有限公司, 海尔智家股份有限公司 filed Critical 重庆海尔洗衣机有限公司
Publication of WO2021098486A1 publication Critical patent/WO2021098486A1/zh

Links

Images

Classifications

    • D—TEXTILES; PAPER
    • D06—TREATMENT OF TEXTILES OR THE LIKE; LAUNDERING; FLEXIBLE MATERIALS NOT OTHERWISE PROVIDED FOR
    • D06F—LAUNDERING, DRYING, IRONING, PRESSING OR FOLDING TEXTILE ARTICLES
    • D06F34/00—Details of control systems for washing machines, washer-dryers or laundry dryers
    • D06F34/14—Arrangements for detecting or measuring specific parameters
    • D06F34/18—Condition of the laundry, e.g. nature or weight
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V10/00—Arrangements for image or video recognition or understanding
    • G06V10/40—Extraction of image or video features

Definitions

  • the colors of clothing vary greatly. In addition to the common red, orange, yellow, green, cyan, blue, and purple, there are also various combinations of colors. As to what colors of clothing can be mixed and washed, the existing technology usually depends on people. Judging by experience, the judgment result is not accurate enough, which may easily cause the clothes that cannot be mixed to be washed together, resulting in cross-color, resulting in a poor user experience.
  • the first aspect of this application provides a method for processing clothing color recognition, including:
  • a pre-trained classification network model is used to determine the segmented image data of the laundry
  • the identification result is sent to the washing machine, so that the washing machine performs corresponding processing according to the identification result.
  • the step of using a pre-trained classification network model to determine the segmented image data of the laundry according to the original image data includes:
  • a pre-trained two-class neural network model based on deep learning is used to determine the segmented image data of the laundry.
  • the step of using a clothing color recognition algorithm based on HSV space to perform color recognition on the segmented image data to obtain a recognition result includes:
  • the color composition of the laundry is determined as the recognition result.
  • the method further includes:
  • the color composition of the main hue of the laundry is used as the recognition result.
  • the second aspect of the present application provides a processing method for clothing color recognition, including:
  • the washing machine is controlled to perform corresponding processing according to the target control program.
  • the determining the corresponding target control program according to the recognition result includes:
  • the corresponding target control program is determined, and the preset control logic rule includes the corresponding relationship between the color and the control program.
  • the determining the corresponding target control program according to the recognition result and preset control logic rules includes:
  • the target control program is determined to be an alarm prompt program
  • the controlling the washing machine to perform corresponding processing according to the target control program includes:
  • a third aspect of the present application provides a processing device for clothing color recognition, including:
  • the first receiving module is used to receive the original image data in the drum sent by the washing machine
  • the first determining module is configured to determine the segmented image data of the laundry by using a pre-trained classification network model according to the original image data;
  • a processing module configured to use a clothing color recognition algorithm based on the HSV space to perform color recognition on the segmented image data to obtain a recognition result
  • the first sending module is configured to send the identification result to the washing machine, so that the washing machine performs corresponding processing according to the identification result.
  • the first determining module is specifically configured to:
  • a pre-trained two-class neural network model based on deep learning is used to determine the segmented image data of the laundry.
  • the processing module is specifically used for:
  • the color composition of the laundry is determined as the recognition result.
  • the processing module is further used for:
  • the color composition of the main hue of the laundry is used as the recognition result.
  • a fourth aspect of the present application provides a processing device for clothing color recognition, including:
  • the acquisition module is used to collect the original image data in the washing machine barrel
  • the second sending module is used to send the original image data to the server
  • the second receiving module is configured to receive the recognition result of the color of the laundry returned by the server;
  • the second determining module is used to determine the corresponding target control program according to the recognition result
  • the control module is used to control the washing machine to perform corresponding processing according to the target control program.
  • the corresponding target control program is determined, and the preset control logic rule includes the corresponding relationship between the color and the control program.
  • the second determining module is specifically configured to:
  • the target control program is determined to be an alarm prompt program
  • the control module is specifically used for:
  • a fifth aspect of the present application provides a server, including: at least one processor and a memory;
  • the memory is configured to store computer-executable instructions, so that the at least one processor executes the computer-executable instructions to implement the method provided in the first aspect.
  • a sixth aspect of the present application provides a washing machine, including: at least one processor and a memory;
  • the memory is used to store computer-executable instructions, so that the at least one processor executes the computer-executable instructions to implement the method provided in the second aspect.
  • a seventh aspect of the present application provides a computer-readable storage medium in which computer-executable instructions are stored, and when the computer-executable instructions are executed by a processor, they are used to implement the method provided in the first aspect.
  • An eighth aspect of the present application provides a computer-readable storage medium in which computer-executable instructions are stored, and the computer-executable instructions are used to implement the method provided in the second aspect when the computer-executable instructions are executed by a processor.
  • a ninth aspect of the present application provides a computer program, including program code.
  • the program code executes the method provided in the first aspect.
  • the tenth aspect of the present application provides a computer program, including program code.
  • the program code executes the method provided in the second aspect.
  • An eleventh aspect of the present application provides a method for processing clothing color recognition, including:
  • the identification result is sent to the washing machine, so that the washing machine performs corresponding processing according to the identification result.
  • the image contour information includes position information of the contour of the laundry on the original image
  • the determining the segmented image data of the laundry according to the original image data and the image contour information includes:
  • the segmented image data of the laundry is segmented from the original image data.
  • the step of using a clothing color recognition algorithm based on HSV space to perform color recognition on the segmented image data to obtain a recognition result includes:
  • the color composition of the main hue of the laundry is used as the recognition result.
  • a twelfth aspect of this application provides a processing method for clothing color recognition, including:
  • the method further includes:
  • the image contour path is a closed path, it is determined that the image contour is qualified
  • the image contour path is a non-closed path, it is determined that the image contour is unqualified
  • the determining image contour information according to the image contour path includes:
  • the image contour information is determined according to the image contour path.
  • the determining image contour information according to the image contour path includes:
  • a preset number of position information is acquired from the image contour path as the image contour information.
  • a thirteenth aspect of the present application provides a processing device for clothing color recognition, including:
  • the first receiving module is configured to receive the raw image data and the image outline information of the laundry sent by the terminal;
  • the first determining module is configured to determine the segmented image data of the laundry according to the original image data and the image contour information
  • a processing module configured to use a clothing color recognition algorithm based on the HSV space to perform color recognition on the segmented image data to obtain a recognition result
  • the first sending module is configured to send the identification result to the washing machine, so that the washing machine performs corresponding processing according to the identification result
  • the image contour information includes position information of the contour of the laundry on the original image; the first determining module is specifically configured to:
  • the segmented image data of the laundry is segmented from the original image data.
  • the processing module is specifically used for:
  • the color composition of the laundry is determined as the recognition result.
  • the processing module is further used for:
  • the color composition of the main hue of the laundry is used as the recognition result.
  • a fourteenth aspect of the present application provides a processing device for clothing color recognition, including:
  • the acquisition module is used to acquire the raw image data taken
  • the second sending module is used to send the original image data to the cloud server
  • the display module is used to display the corresponding original image on the contour input interface according to the original image data
  • the acquiring module is further configured to acquire the image contour path of the laundry input by the user on the contour input interface;
  • the second determining module is configured to determine image contour information according to the image contour path
  • the second sending module is further configured to send the image outline information to the cloud server;
  • the second receiving module is used for the recognition result of the color of the laundry returned by the cloud server;
  • the display module is also used to display the recognition result of the color of the laundry returned by the cloud server.
  • the second determining module is further used for:
  • the image contour path is a closed path, it is determined that the image contour is qualified
  • the image contour path is a non-closed path, it is determined that the image contour is unqualified
  • the second determining module is specifically used for:
  • the image contour information is determined according to the image contour path.
  • the second determining module is specifically configured to:
  • a preset number of position information is acquired from the image contour path as the image contour information.
  • the fifteenth aspect of the present application provides a server, including: at least one processor and a memory;
  • the memory is configured to store computer-executable instructions, so that the at least one processor executes the computer-executable instructions to implement the method provided in the eleventh aspect.
  • a sixteenth aspect of the present application provides a terminal, including: at least one processor and a memory;
  • the memory is used to store computer-executable instructions, so that the at least one processor executes the computer-executable instructions to implement the method provided in the twelfth aspect.
  • a seventeenth aspect of the present application provides a computer-readable storage medium having computer-executable instructions stored in the computer-readable storage medium, and when the computer-executable instructions are executed by a processor, the method.
  • An eighteenth aspect of the present application provides a computer-readable storage medium having computer-executable instructions stored in the computer-readable storage medium, and when the computer-executable instructions are executed by a processor, they are used to implement the method.
  • a nineteenth aspect of the present application provides a computer program, including program code.
  • the program code executes the method provided in the eleventh aspect.
  • the twentieth aspect of the present application provides a computer program, including program code.
  • the program code executes the method provided in the twelfth aspect.
  • the clothing color recognition processing method, device, equipment and storage medium of the present application uses raw image data of the clothing to determine the segmented image data of the laundry, and adopts the clothing color based on the HSV space.
  • the recognition algorithm performs color recognition on the segmented image data of the laundry to obtain the color recognition result, which effectively improves the accuracy of the color recognition of the laundry, and sends the recognition result to the washing machine, so that the washing machine can perform corresponding processing according to the recognition result, which can be based on the color Recognition provides users with accurate laundry services and improves user experience.
  • FIG. 1 is a schematic diagram of the architecture of a processing system provided by an embodiment of the application
  • FIG. 2 is a schematic flowchart of a method for processing clothing color recognition provided by an embodiment of the application
  • FIG. 3 is a schematic flowchart of a method for processing clothing color recognition according to another embodiment of the application.
  • FIG. 4 is a schematic flowchart of a processing method for clothing color recognition provided by still another embodiment of the application.
  • FIG. 5 is a schematic flowchart of a method for processing clothing color recognition according to another embodiment of the application.
  • FIG. 6 is a schematic structural diagram of a processing device for clothing color recognition provided by an embodiment of the application.
  • FIG. 7 is a schematic structural diagram of a processing device for clothing color recognition provided by another embodiment of the application.
  • FIG. 8 is a schematic structural diagram of a server provided by an embodiment of this application.
  • FIG. 9 is a schematic structural diagram of a washing machine provided by an embodiment of the application.
  • FIG. 10 is a schematic structural diagram of a processing system provided by another embodiment of this application.
  • FIG. 11 is a schematic flowchart of a method for processing clothing color recognition provided by an embodiment of the application.
  • FIG. 12 is a schematic flowchart of a method for processing clothing color recognition according to another embodiment of the application.
  • FIG. 13 is a schematic flowchart of a method for processing clothing color recognition according to still another embodiment of the application.
  • FIG. 15 is a schematic diagram of a work flow provided by an embodiment of the application.
  • 16 is a schematic structural diagram of a processing device for clothing color recognition provided by an embodiment of the application.
  • FIG. 17 is a schematic structural diagram of a processing device for clothing color recognition provided by another embodiment of the application.
  • FIG. 18 is a schematic structural diagram of a server provided by an embodiment of this application.
  • FIG. 19 is a schematic structural diagram of a terminal provided by an embodiment of this application.
  • RGB space based on the three basic colors of R (Red: Red), G (Green: Green), and B (Blue: Blue), different degrees of superposition are performed to produce a rich and wide range of colors, so it is commonly known as the three-primary color mode .
  • the color is represented by a cube of unit length.
  • the 8 common colors of black, blue, green, red, purple, yellow and white are located at the 8 vertices of the cube.
  • the black is placed at the origin of the three-dimensional rectangular coordinate system, and the red, green and blue are placed respectively. With 3 coordinate axes, the entire cube is placed within the limit of the first hexagram.
  • cyan and red, purple (or magenta) and green, yellow and blue are complementary colors.
  • each parameter is: R: 0-255; G: 0-255; B: 0-255.
  • HSV space refers to the HSV (Hue (hue), Saturation (saturation), Value (lightness)) color space, also known as the hexagonal pyramid model space.
  • FIG. 1 is a schematic diagram of the architecture of the processing system based on the following embodiments of the application related to the interaction between the washing machine and the server.
  • the processing system may include a washing machine and a server.
  • the server may be a cloud server or other servers.
  • the washing machine can be a drum washing machine or other types of washing machines. Taking the drum washing machine as an example, the server and the washing machine can be equipped with corresponding laundry color recognition processing devices.
  • the original image data can be sent to the server through network transmission.
  • the laundry color recognition processing device of the washing machine obtains the original image data taken by the camera and sends it to the server via network transmission, and the laundry color recognition processing device on the server can obtain the original image data.
  • the server uses the pre-trained classification network model to determine the segmented image data of the laundry according to the original image data.
  • the segmented image data refers to the segmented image data from the original image data and only includes the image of the laundry Part of the image data; use the clothing color recognition algorithm based on HSV space to perform color recognition on the segmented image data to obtain the recognition result; send the recognition result to the washing machine. After the washing machine receives the recognition result, it controls the washing machine according to the recognition result and preset control logic Proceed accordingly.
  • An embodiment of the present application provides a processing method for identifying the color of clothes, which is used to identify the color of clothes to be washed in a washing machine barrel.
  • the execution subject of this embodiment is a processing device for clothing color recognition, and the device may be installed in a server.
  • FIG. 2 is a schematic flowchart of a method for processing clothing color recognition provided by this embodiment. The method includes:
  • Step 101 Receive original image data in the drum sent by the washing machine.
  • a camera can be installed above the door and window of the washing machine, such as a mini camera, used to capture the scene image (or original image) of the underwear of the drum, obtain the original image data in the drum, and send it to the server through the washing machine.
  • a mini camera used to capture the scene image (or original image) of the underwear of the drum, obtain the original image data in the drum, and send it to the server through the washing machine.
  • the user can be triggered when the user opens the door and window of the washing machine, puts clothes in, and closes the door and window, and triggers the camera of the washing machine to capture the scene image of the tube underwear, or it can be triggered when the user closes the door and window and starts the washing machine, which can be set according to actual needs.
  • This embodiment is not limited.
  • Step 102 According to the original image data, a pre-trained classification network model is used to determine the segmented image data of the laundry.
  • the original training image data in the cylinder can be collected in advance and the original training image data can be labeled to obtain the label data corresponding to the original training image data; and the classification neural network can be established based on the original training image data and the corresponding label data training. Good classification neural network, get classification network model.
  • the pre-trained classification network model can be used to determine the segmented image data of the laundry based on the original image data.
  • Segmented image data refers to the image data segmented from the original image data that only includes the image part of the laundry, that is, through the classification network model, the clothing and background in the original image data are separated to obtain which of the original image data The parts belong to the background part, which parts belong to the clothes part, and the parts belong to the clothes part are the segmented image data of the laundry.
  • the classification network model is a two-class neural network model based on deep learning, which takes into account the processing rate and accuracy, and divides the original image into two categories: clothing and background.
  • the network architecture of the classification network model can include a cascade network.
  • the cascade network is divided into four layers. Each layer will perform convolutional down-sampling of the original object (or original image data) to different degrees to extract semantic features.
  • the output of each layer The feature maps will finally be fused into a feature map, and then through the up-sampling layer to extract detailed features to obtain the target feature map, and finally the obtained target feature map is subjected to conventional semantic segmentation to obtain the final segmented image.
  • Step 103 Perform color recognition on the segmented image data by using a clothing color recognition algorithm based on the HSV space to obtain a recognition result.
  • a clothing color recognition algorithm based on the HSV space can be used to perform color recognition on the segmented image data to obtain the recognition result.
  • the recognition result includes the color composition of the laundry, such as 80% red and 20% white.
  • the clothing color recognition algorithm based on the HSV space is an algorithm that converts the segmented image data from the RGB space to the HSV space for color recognition.
  • the segmented image data is converted from the RGB space to the HSV space to obtain the converted segmented image data; the clothing color representation range is acquired; based on the converted segmented image data and the clothing color representation range, the color composition of the laundry is determined ,
  • the clothing color characterization range includes different ranges representing different colors
  • the clothing color characterization range is determined by combining the HSV hexagonal pyramid and a large number of clothing color characterization realizations with different colors. For example, combining HSV hexagonal pyramid and 1000 pieces of clothes with different colors to realize the color characterization, determine the HSV characterization range of 10 common colors.
  • the representation range of black is:
  • h, s, and v respectively represent the values of the three channels.
  • Step 104 Send the recognition result to the washing machine, so that the washing machine performs corresponding processing according to the recognition result.
  • the recognition result is sent to the washing machine.
  • the washing machine receives the recognition result of the color of the laundry, it can determine the corresponding target control program according to the recognition result, and control the washing machine to perform corresponding processing according to the target control program.
  • the target control program is determined to be a program that does not start the washing machine and gives an alarm. Start the washing machine and control the washing machine to send out an alarm.
  • the specific alarm notification method can be set according to actual needs, for example, it can be an alarm notification through voice broadcast, or the corresponding alarm information can be displayed on the display interface of the washing machine and an alarm notification sound can be issued.
  • the recognition result is white and is greater than the second threshold, it is determined that the corresponding target control program is the color protection washing program, and the color protection washing program may be selected to control the washing machine to perform the color protection washing process.
  • the mixed washing program may be selected to control the washing machine to perform the mixed washing process.
  • preset control logic rules may be set in the washing machine in advance, which may include the corresponding relationship between colors and control programs. After the washing machine receives the recognition result, it determines the target control program according to the recognition result and the preset control logic rules.
  • the laundry color recognition processing method uses a camera on the washing machine to capture the original image in the drum, the washing machine sends the original image data to the server, and the server receives the original image data in the drum sent by the washing machine; according to the original image data,
  • the pre-trained classification network model determines the segmented image data of the laundry; uses the clothing color recognition algorithm based on the HSV space to perform color recognition on the segmented image data to obtain the recognition result; sends the recognition result to the washing machine so that the washing machine can recognize according to As a result, corresponding processing is performed, which effectively improves the accuracy of clothing color recognition, thereby being able to provide users with accurate laundry services and improve user experience.
  • Another embodiment of the present application further supplements the method provided in the foregoing embodiment.
  • FIG. 3 is a schematic flowchart of a processing method for clothing color recognition provided by this embodiment.
  • a pre-trained classification network model is used to determine the segmented image data of the laundry, including:
  • Step 1021 According to the original image data, a pre-trained deep learning-based two-class neural network model is used to determine the segmented image data of the laundry.
  • the classification network model is a two-class neural network model based on deep learning, which takes into account the processing speed and accuracy, and divides the original image into two categories: clothing and background.
  • the network architecture of the classification network model can include a cascade network.
  • the cascade network is divided into four layers. Each layer will perform convolutional down-sampling of the original object (or original image data) to different degrees to extract semantic features.
  • the output of each layer The feature maps will finally be fused into a feature map, and then through the up-sampling layer to extract detailed features to obtain the target feature map, and finally the obtained target feature map is subjected to conventional semantic segmentation to obtain the final segmented image.
  • the classification network model Before adopting the classification network model, the classification network model needs to be obtained through training data training. Specifically, the original training image data in the cylinder can be collected in advance and the original training image data can be annotated to obtain the label data corresponding to the original training image data; and a classification neural network can be established to train based on the original training image data and the corresponding label data Establish a good classification neural network and obtain a classification network model.
  • a large number of clothing images can be taken with the help of supplementary light, and after cleaning (cleaning out images with unclear color information and excessive environmental interference noise) and labeling, the original training image data and corresponding label data can be obtained, based on the original training image data Train the pre-established two-class neural network based on deep learning with the corresponding label data to obtain a two-class neural network model based on deep learning.
  • a clothing color recognition algorithm based on HSV space to perform color recognition on the segmented image data to obtain a recognition result, including:
  • Step 1031 Convert the segmented image data from the RGB space to the HSV space, and obtain the converted segmented image data.
  • Step 1032 Obtain the clothing color characterization range.
  • Step 1033 Determine the color composition of the laundry based on the converted segmented image data and the clothing color representation range.
  • Step 1034 Use the color composition of the laundry as a recognition result.
  • the clothing color representation range includes different ranges representing different colors.
  • the clothing color representation range can be determined by combining the HSV hexagonal pyramid and a large number of clothing color representations with different colors. For example, combining HSV hexagonal pyramid and 1000 pieces of clothes with different colors to realize the color characterization, determine the HSV characterization range of 10 common colors.
  • the representation range of black is:
  • h, s, and v respectively represent the values of the three channels.
  • the method further includes:
  • Step 1035 Filter the color composition of the laundry to obtain the main hue color composition of the laundry.
  • step 1036 the main color composition of the laundry is used as the recognition result.
  • the color composition that may be obtained includes a very small proportion of other colors, and the color composition of the laundry can be filtered.
  • the main hue color composition of the laundry is obtained, and the main hue color composition of the laundry is sent to the washing machine as a recognition result.
  • the clothing color recognition processing method provided by the embodiment of the application realizes the automatic recognition of clothing color, and combines the two-class neural network model of deep learning and the color recognition algorithm based on HSV space, which effectively improves the accuracy of color recognition and saves
  • the user judges the color of the clothes based on experience, manually selects the link of the washing mode, frees the user's hands, improves the convenience of the washing link, and solves the problem that the user judges the color of the clothes based on experience and does not select the washing mode accurately.
  • the laundry color recognition processing method uses a camera on the washing machine to capture the original image in the drum, the washing machine sends the original image data to the server, and the server receives the original image data in the drum sent by the washing machine; according to the original image data,
  • the pre-trained classification network model determines the segmented image data of the laundry; uses the clothing color recognition algorithm based on the HSV space to perform color recognition on the segmented image data to obtain the recognition result; sends the recognition result to the washing machine so that the washing machine can recognize according to As a result, corresponding processing is performed, which effectively improves the accuracy of clothing color recognition, thereby being able to provide users with accurate laundry services and improve user experience.
  • Yet another embodiment of the present application provides a processing method for identifying the color of clothes, which is used to identify the color of clothes to be washed in a washing machine barrel.
  • the execution subject of this embodiment is a processing device for clothing color recognition, and the device may be installed in a washing machine.
  • FIG. 4 is a schematic flowchart of a method for processing clothing color recognition provided by this embodiment. The method includes:
  • Step 301 Collect the original image data in the drum of the washing machine.
  • Step 302 Send the original image data to the server, so that the server obtains the color recognition result of the laundry according to the original image data.
  • Step 303 Receive the identification result of the color of the laundry returned by the server.
  • Step 304 Determine the corresponding target control program according to the recognition result.
  • Step 305 Control the washing machine to perform corresponding processing according to the target control program.
  • this embodiment is an embodiment of a method on the washing machine side that interacts with the above server-side embodiment.
  • the specific execution process has been described in detail in the interactive description of the above embodiment.
  • the above embodiment involves execution on the washing machine side.
  • the parts can be used to explain the above steps of this embodiment, and will not be repeated here.
  • the original image in the drum is captured by the camera on the washing machine, the washing machine sends the original image data to the server, and the server receives the original image data in the drum sent by the washing machine; according to the original image data,
  • the pre-trained classification network model determines the segmented image data of the laundry; uses the clothing color recognition algorithm based on HSV space to perform color recognition on the segmented image data to obtain the recognition result; sends the recognition result to the washing machine so that the washing machine can recognize according to As a result, corresponding processing is performed, which effectively improves the accuracy of clothing color recognition, thereby being able to provide users with accurate laundry services and improve user experience.
  • Another embodiment of the present application further supplements the method provided in the foregoing embodiment.
  • FIG. 5 is a schematic flowchart of the processing method for clothing color recognition provided by this embodiment.
  • determining the corresponding target control program according to the recognition result includes:
  • Step 3041 Determine the corresponding target control program according to the recognition result and preset control logic rules.
  • the preset control logic rule includes the corresponding relationship between the color and the control program.
  • determining the corresponding target control program according to the recognition result and preset control logic rules including:
  • step 30411 if it is determined that the laundry has a risk of cross-color according to the recognition result, the target control program is determined to be an alarm prompt program.
  • the washing machine is controlled to perform corresponding processing according to the target control program, including:
  • Step 3051 Control the washing machine to issue an alarm prompt according to the alarm prompt program.
  • the target control program is determined to be a program that does not start the washing machine and gives an alarm. Start the washing machine and control the washing machine to send out an alarm.
  • the specific alarm notification method can be set according to actual needs, for example, it can be an alarm notification through voice broadcast, or the corresponding alarm information can be displayed on the display interface of the washing machine and an alarm notification sound can be issued.
  • the recognition result is white and is greater than the second threshold, it is determined that the corresponding target control program is a color protection washing program, and the color protection washing program can be selected to control the washing machine to perform the color protection washing process.
  • the mixed washing program may be selected to control the washing machine to perform the mixed washing process.
  • this embodiment is an embodiment of a method on the washing machine side that interacts with the above server-side embodiment.
  • the specific execution process has been described in detail in the interactive description of the above embodiment.
  • the above embodiment involves execution on the washing machine side.
  • the parts can be used to explain the above steps of this embodiment, and will not be repeated here.
  • the laundry color recognition processing method uses a camera on the washing machine to capture the original image in the drum, the washing machine sends the original image data to the server, and the server receives the original image data in the drum sent by the washing machine; according to the original image data,
  • the pre-trained classification network model determines the segmented image data of the laundry; uses the clothing color recognition algorithm based on the HSV space to perform color recognition on the segmented image data to obtain the recognition result; sends the recognition result to the washing machine so that the washing machine can recognize according to As a result, corresponding processing is performed, which effectively improves the accuracy of clothing color recognition, thereby being able to provide users with accurate laundry services and improve user experience.
  • Yet another embodiment of the present application provides a processing device for clothing color recognition, which is used to execute the method provided by the foregoing server-side embodiment.
  • the processing device 50 for clothing color recognition includes a first receiving module 51, a first determining module 52, a processing module 53, and a first sending module 54.
  • the first receiving module is used to receive the original image data in the drum sent by the washing machine; the first determining module is used to determine the segmented image data of the laundry according to the original image data and using a pre-trained classification network model;
  • the processing module is used to recognize the segmented image data by using the clothing color recognition algorithm based on the HSV space to obtain the recognition result;
  • the first sending module is used to send the recognition result to the washing machine, so that the washing machine performs corresponding processing according to the recognition result .
  • Another embodiment of the present application further supplements the device provided in the foregoing embodiment.
  • the first determining module is specifically used for:
  • a pre-trained two-class neural network model based on deep learning is used to determine the segmented image data of the laundry.
  • the processing module is specifically used for:
  • processing module is also used to:
  • the original image in the drum is captured by the camera on the washing machine, the washing machine sends the original image data to the server, and the server receives the original image data in the drum sent by the washing machine; according to the original image data, Use the pre-trained classification network model to determine the segmented image data of the laundry; use the HSV space-based clothing color recognition algorithm to perform color recognition on the segmented image data to obtain the recognition result; send the recognition result to the washing machine so that the washing machine can be based on The recognition results are processed accordingly, which effectively improves the accuracy of clothing color recognition, so as to provide users with accurate laundry services and improve user experience.
  • Yet another embodiment of the present application provides a processing device for clothing color recognition, which is used to implement the method provided by the above-mentioned washing machine-side embodiment.
  • the processing device 70 for clothing color recognition includes an acquiring module 71, a second sending module 72, a second receiving module 73, a second determining module 74, and a control module 75.
  • the acquisition module is used to collect the original image data in the washing machine barrel; the second sending module is used to send the original image data to the server; the second receiving module is used to receive the color recognition result of the laundry returned by the server.
  • the second determination module is used to determine the corresponding target control program according to the recognition result; the control module is used to control the washing machine to perform corresponding processing according to the target control program.
  • the original image in the drum is captured by the camera on the washing machine, the washing machine sends the original image data to the server, and the server receives the original image data in the drum sent by the washing machine; according to the original image data, Use the pre-trained classification network model to determine the segmented image data of the laundry; use the HSV space-based clothing color recognition algorithm to perform color recognition on the segmented image data to obtain the recognition result; send the recognition result to the washing machine so that the washing machine can be based on The recognition results are processed accordingly, which effectively improves the accuracy of clothing color recognition, so as to provide users with accurate laundry services and improve user experience.
  • the second determining module is specifically used for:
  • the corresponding target control program is determined, and the preset control logic rule includes the corresponding relationship between the color and the control program.
  • the target control program is determined to be an alarm prompt program; the control module is specifically used for:
  • the original image in the drum is captured by the camera on the washing machine, the washing machine sends the original image data to the server, and the server receives the original image data in the drum sent by the washing machine; according to the original image data, Use the pre-trained classification network model to determine the segmented image data of the laundry; use the HSV space-based clothing color recognition algorithm to perform color recognition on the segmented image data to obtain the recognition result; send the recognition result to the washing machine so that the washing machine can be based on The recognition results are processed accordingly, which effectively improves the accuracy of clothing color recognition, so as to provide users with accurate laundry services and improve user experience.
  • Yet another embodiment of the present application provides a server for executing the method provided in the foregoing server-side embodiment.
  • the server 80 includes: at least one processor 81 and a memory 82;
  • the memory is used to store computer-executable instructions, so that at least one processor executes the computer-executable instructions to implement the methods provided in the foregoing embodiments.
  • the original image in the drum is captured by the camera on the washing machine, the washing machine sends the original image data to the server, and the server receives the original image data in the drum sent by the washing machine; according to the original image data, a pre-trained classification is used
  • the network model determines the segmented image data of the laundry; uses the clothing color recognition algorithm based on the HSV space to perform color recognition on the segmented image data to obtain the recognition result; sends the recognition result to the washing machine so that the washing machine performs corresponding processing according to the recognition result , Which effectively improves the accuracy of clothing color recognition, which can provide users with accurate laundry services and improve user experience.
  • Another embodiment of the present application provides a washing machine, which is used to implement the method provided in the above-mentioned washing machine-side embodiment.
  • FIG. 9 is a schematic structural diagram of the washing machine provided in this embodiment.
  • the washing machine 90 includes: at least one processor 91 and a memory 92;
  • the memory is used to store computer-executable instructions, so that at least one processor executes the computer-executable instructions to implement the method provided by the foregoing washing machine-side embodiment.
  • the washing machine may also include a sound broadcasting device.
  • the original image in the drum is captured by the camera on the washing machine, the washing machine sends the original image data to the server, and the server receives the original image data in the drum sent by the washing machine; according to the original image data, a pre-trained classification is used
  • the network model determines the segmented image data of the laundry; uses the clothing color recognition algorithm based on the HSV space to perform color recognition on the segmented image data to obtain the recognition result; sends the recognition result to the washing machine so that the washing machine performs corresponding processing according to the recognition result , Which effectively improves the accuracy of clothing color recognition, which can provide users with accurate laundry services and improve user experience.
  • the computer-readable storage medium stores computer-executable instructions.
  • the computer-executable instructions are executed by a processor, they are used to implement the server-side execution method provided by any of the foregoing embodiments. .
  • the original image in the drum is captured by the camera on the washing machine, the washing machine sends the original image data to the server, and the server receives the original image data in the drum sent by the washing machine;
  • the trained classification network model determines the segmented image data of the laundry; uses the clothing color recognition algorithm based on HSV space to perform color recognition on the segmented image data to obtain the recognition result; sends the recognition result to the washing machine so that the washing machine can be based on the recognition result
  • Corresponding processing can effectively improve the accuracy of clothing color recognition, so as to provide users with accurate laundry services and improve user experience.
  • Another embodiment of the present application provides a computer-readable storage medium in which computer-executable instructions are stored.
  • the computer-executable instructions are executed by a processor, the computer-executable instructions are used to implement any one of the above-mentioned embodiments to provide a washing machine-side execution method.
  • the original image in the drum is captured by the camera on the washing machine, the washing machine sends the original image data to the server, and the server receives the original image data in the drum sent by the washing machine;
  • the trained classification network model determines the segmented image data of the laundry; uses the clothing color recognition algorithm based on HSV space to perform color recognition on the segmented image data to obtain the recognition result; sends the recognition result to the washing machine so that the washing machine can be based on the recognition result
  • Corresponding processing can effectively improve the accuracy of clothing color recognition, so as to provide users with accurate laundry services and improve user experience.
  • the laundry color recognition processing method provided by another embodiment of the present application is applicable to an application scenario in which a user automatically recognizes the color of the laundry when the user washes the laundry, and selects a suitable laundry program for washing according to the color of the laundry.
  • FIG. 10 is a schematic diagram of the architecture of a processing system related to interaction between a terminal, a washing machine, and a server in the following embodiments of the application.
  • the processing system may include a terminal, a washing machine, and a server.
  • the server may be a cloud server or other servers.
  • the terminal can be installed with the corresponding client application APP of the washing machine or the corresponding processing device for clothing color recognition can be set on the terminal.
  • the washing machine can be a drum washing machine or other types of washing machines.
  • the server can set the corresponding clothing color recognition ⁇ Processing device.
  • the terminal can be a terminal with a camera such as a user's mobile phone or tablet computer.
  • the clothing scene image (or original image) can be captured through the terminal camera, and the flash can also be used to fill light to help capture clear and analyzable images.
  • the contour of the laundry the user can outline the contour of the laundry in the original image on the original image of the contour input interface, the terminal can obtain the image contour path outlined by the user, and determine the image contour information according to the image contour path.
  • the recognition result is displayed, and the washing machine receives After the recognition result, the washing machine is controlled to perform corresponding processing according to the recognition result and preset control logic.
  • the clothing color recognition algorithm based on HSV space is used to recognize the segmented image data, which effectively improves the accuracy of clothing color recognition, thereby providing users with accurate The laundry service improves the user experience.
  • This application does not need to add any additional sensors in the washing machine, and borrows the camera integrated in the user terminal to complete the image collection work. This does not increase the complexity of the washing machine's main structure, and at the same time does not increase the cost, and at the same time increase the level of intelligence and the washing machine. Convenience.
  • An embodiment of the present application provides a processing method for clothing color recognition, which is used to recognize the color of laundry.
  • the execution subject of this embodiment is a processing device for clothing color recognition, and the device may be installed in a server.
  • FIG. 11 is a schematic flowchart of a method for processing clothing color recognition provided by this embodiment. The method includes:
  • Step 1101 Receive the raw image data and the image outline information of the laundry sent by the terminal.
  • the terminal can be installed with the corresponding client application APP of the washing machine or the terminal can be equipped with a corresponding clothing color recognition processing device, and the clothing scene image (or original image) can be captured through the terminal camera to obtain the original image data, which can be obtained through the network
  • the transmission sends the original image data to the server, and displays the captured original image on the terminal.
  • it can be displayed on the contour input interface for the user to outline the contour of the laundry in the original image.
  • the user can display the original image on the contour input interface Outline the outline of the laundry in the original image.
  • the terminal can obtain the image outline path outlined by the user, and determine the image outline information according to the image outline path.
  • the image outline information includes the determined location information of the outline of the laundry in the original image
  • the terminal can send the image outline information to the server.
  • the terminal may send the original image data and image contour information to the server together after obtaining the original image data, or send it once after obtaining the original image data, and send it again after obtaining the image contour information.
  • the actual demand setting is not limited in this embodiment.
  • the clothing color recognition processing device on the server can obtain the original image data and the determined image contour information.
  • Step 1102 Determine the segmented image data of the laundry according to the original image data and the image outline information.
  • the server determines the segmented image data of the laundry according to the original image data and the image contour information.
  • the segmented image data refers to the segmented image data that only includes the laundry The image data of the object image part.
  • the image contour information includes position information of the contour of the laundry on the original image; the server may segment the segmented image data of the laundry from the original image data according to the image contour information.
  • the image contour information may include the position coordinates of the top, left, most square, and bottom four positions of the contour of the laundry in the original image, and the server may extract the four position coordinates from the original image data according to these four position coordinates.
  • the rectangular area including the laundry is segmented as the segmented image data of the laundry.
  • the image contour information may also include more position coordinates of the contour of the laundry in the original image, so that the segmented image data of the laundry is more accurate. Specifically, it can be set according to actual needs, which is not limited in this embodiment.
  • Step 1103 Use the clothing color recognition algorithm based on the HSV space to perform color recognition on the segmented image data to obtain a recognition result.
  • a clothing color recognition algorithm based on the HSV space can be used to perform color recognition on the segmented image data to obtain the recognition result.
  • the recognition result includes the color composition of the laundry, such as 80% red and 20% white.
  • the clothing color recognition algorithm based on the HSV space is an algorithm that converts the segmented image data from the RGB space to the HSV space for color recognition.
  • the segmented image data is converted from the RGB space to the HSV space to obtain the converted segmented image data; the clothing color representation range is acquired; based on the converted segmented image data and the clothing color representation range, the color composition of the laundry is determined ,
  • the clothing color characterization range includes different ranges representing different colors
  • the clothing color characterization range is determined by combining the HSV hexagonal pyramid and a large number of clothing color characterization realizations with different colors. For example, combining HSV hexagonal pyramid and 1000 pieces of clothes with different colors to realize the color characterization, determine the HSV characterization range of 10 common colors.
  • the representation range of black is:
  • h, s, and v respectively represent the values of the three channels.
  • the usually obtained segmented image data is three RGB channel values, each channel is represented by 0-255.
  • the determination of each color requires three values to change at the same time. It is impossible to accurately determine which range a color is in, and it is converted from RGB space to After the HSV space, H represents the hue of a color, and the general attributes of the color can be basically determined, and the depth of the color can be refined by subdividing S and V.
  • the color composition of the laundry can be determined by traversing each pixel in the segmented image data and counting the proportion of pixels attributed to each color characterization range.
  • Step 1104 Send the recognition result to the washing machine, so that the washing machine performs corresponding processing according to the recognition result.
  • the server After the server obtains the recognition result of the color of the laundry, it can send the recognition result to the washing machine.
  • the washing machine can select an appropriate target control program according to the recognition result.
  • the target control program controls the washing machine to perform the corresponding processing.
  • the server may also send the recognition result to the terminal, and the terminal may display the recognition result on the terminal for the user to view, so that the user can understand the color composition of the laundry to be washed.
  • Preset control logic rules may be set in the washing machine in advance, which may include the corresponding relationship between colors and control programs. After the washing machine receives the recognition result, the target control program is determined according to the recognition result and the preset control logic rules.
  • the target control program is determined to be a program that does not start the washing machine and gives an alarm. Start the washing machine and control the washing machine to send out an alarm.
  • the specific alarm notification method can be set according to actual needs, for example, it can be an alarm notification through voice broadcast, or the corresponding alarm information can be displayed on the display interface of the washing machine and an alarm notification sound can be issued.
  • the recognition result is white and is greater than the second threshold, it is determined that the corresponding target control program is the color protection washing program, and the color protection washing program may be selected to control the washing machine to perform the color protection washing process.
  • the mixed washing program may be selected to control the washing machine to perform the mixed washing process.
  • the original image data can be sent to the server through network transmission, and the captured original image can be displayed on the terminal for the user to outline the original image.
  • the terminal can obtain the image contour path outlined by the user, and determine the image contour information according to the image contour path, and the terminal can send the image contour information to the server.
  • the user terminal realizes image shooting without installing a camera on the washing machine, which reduces the cost of the washing machine.
  • Another embodiment of the present application further supplements the method provided in the foregoing embodiment.
  • FIG. 12 is a schematic flowchart of the processing method for clothing color recognition provided by this embodiment.
  • the image contour information includes the position information of the contour of the laundry on the original image; the laundry is determined according to the original image data and the image contour information
  • the segmented image data includes:
  • Step 11021 According to the image contour information, segment the segmented image data of the laundry from the original image data.
  • the image contour information includes the position information of the contour of the laundry on the original image, which may include a preset amount of position information, and the server can segment the divided image of the laundry from the original image data according to the image contour information. data.
  • the image contour information may include the position coordinates of the top, left, most square, and bottom four positions of the contour of the laundry in the original image, and the server may extract the four position coordinates from the original image data according to these four position coordinates.
  • the rectangular area including the laundry is segmented as the segmented image data of the laundry.
  • the image contour information may also include more position coordinates of the contour of the laundry in the original image, so that the segmented image data of the laundry is more accurate. Specifically, it can be set according to actual needs, which is not limited in this embodiment.
  • a clothing color recognition algorithm based on HSV space to perform color recognition on the segmented image data to obtain a recognition result, including:
  • Step 11031 Convert the segmented image data from the RGB space to the HSV space, and obtain the converted segmented image data.
  • Step 11032 Obtain the clothing color characterization range.
  • Step 11033 Determine the color composition of the laundry based on the converted segmented image data and the clothing color representation range.
  • Step 11034 Use the color composition of the laundry as a recognition result.
  • the clothing color representation range includes different ranges representing different colors.
  • the clothing color representation range can be determined by combining the HSV hexagonal pyramid and a large number of clothing color representations with different colors. For example, combining HSV hexagonal pyramid and 1000 pieces of clothes with different colors to realize the color characterization, determine the HSV characterization range of 10 common colors.
  • the representation range of black is:
  • h, s, and v respectively represent the values of the three channels.
  • the method further includes:
  • Step 11035 Filter the color composition of the laundry to obtain the main color composition of the laundry.
  • Step 11036 Use the main color composition of the laundry as the recognition result.
  • the color composition that may be obtained includes a very small proportion of other colors, and the color composition of the laundry can be filtered.
  • the original image data can be sent to the server through network transmission, and the captured original image can be displayed on the terminal for the user to outline the original image.
  • the terminal can obtain the image contour path outlined by the user, and determine the image contour information according to the image contour path, and the terminal can send the image contour information to the server.
  • the clothing color recognition processing device on the server can obtain the original image data and the determined image contour information, and determine the segmented image data of the laundry based on the original image data and the image contour information, and use the clothing color recognition algorithm based on the HSV space Perform color recognition on the segmented image data to obtain the recognition result; send the recognition result to the terminal and the washing machine. After the terminal receives the recognition result, it will display the recognition result. After the washing machine receives the recognition result, it will control according to the recognition result and preset control logic. The washing machine carries out the corresponding treatment.
  • the user terminal realizes image shooting without installing a camera on the washing machine, which reduces the cost of the washing machine.
  • Another embodiment of the present application provides a processing method for identifying the color of clothes, which is used to identify the color of clothes to be washed.
  • the execution subject of this embodiment is a processing device for clothing color recognition, and the device may be installed in a terminal.
  • FIG. 13 is a schematic flowchart of a method for processing clothing color recognition provided by this embodiment. The method includes:
  • Step 1301 Acquire raw image data taken.
  • Step 1303 Obtain the image contour path of the laundry input by the user on the contour input interface.
  • Step 1304 Determine image contour information according to the image contour path.
  • Step 1305 Send the image outline information to the cloud server.
  • Step 1306 Receive and display the color recognition result of the laundry returned by the cloud server.
  • this embodiment is an embodiment of the method on the terminal side that interacts with the above-mentioned server-side embodiment.
  • the specific execution process has been described in detail in the interactive description of the above-mentioned embodiment.
  • the above-mentioned embodiment relates to the terminal side.
  • the executed part can be used to explain the above steps of this embodiment, and will not be repeated here.
  • the original image data can be sent to the server through network transmission, and the captured original image can be displayed on the terminal for the user to outline the original image.
  • the terminal can obtain the image contour path outlined by the user, and determine the image contour information according to the image contour path, and the terminal can send the image contour information to the server.
  • the clothing color recognition processing device on the server can obtain the original image data and the determined image contour information, and determine the segmented image data of the laundry based on the original image data and the image contour information, and use the clothing color recognition algorithm based on the HSV space Perform color recognition on the segmented image data to obtain the recognition result; send the recognition result to the terminal and the washing machine. After the terminal receives the recognition result, it will display the recognition result. After the washing machine receives the recognition result, it will control according to the recognition result and preset control logic. The washing machine carries out the corresponding treatment.
  • the user terminal realizes image shooting without installing a camera on the washing machine, which reduces the cost of the washing machine.
  • Another embodiment of the present application further supplements the method provided in the foregoing embodiment.
  • FIG. 14 is a schematic flowchart of the processing method for clothing color recognition provided by this embodiment.
  • the method further includes:
  • Step 12011 Determine whether the image contour is qualified according to the image contour path.
  • the image contour path is a closed path, the image contour is determined to be qualified; if the image contour path is a non-closed path, the image contour is determined to be unqualified.
  • Determine the image contour information according to the image contour path including:
  • step 13041 if the image contour is qualified, the image contour information is determined according to the image contour path.
  • the user can use a sliding operation to outline the contour of the laundry on the contour input interface of the terminal, and the terminal can obtain the path of the user's sliding operation, that is, the image contour path of the laundry. According to whether the path of the user's sliding operation is closed, it is judged whether the image contour input by the user is qualified.
  • determining the image contour information according to the image contour path includes:
  • Step 12021 Obtain a preset amount of position information from the image contour path as the image contour information.
  • the terminal may obtain a certain amount of position information on the image contour path input by the user as the image contour information, so as to reduce the amount of data processing and increase the processing speed. Or, all the position information on the image contour path may be used as the image contour information.
  • the specific preset number can be set according to actual needs.
  • FIG. 15 is a schematic diagram of a work flow provided in this embodiment. Specifically:
  • Step 121 The user places the clothes to be washed (that is, the clothes to be washed) in a certain position, opens the APP client (that is, the client application APP corresponding to the washing machine installed in the terminal), presses the color recognition button, and the APP will actively call
  • the phone's built-in camera enters the snapshot mode, and the user chooses to capture a picture and display it in the APP.
  • Step 122 The user selects the rough outline of the clothing on the APP client interface, and after clicking confirm, the APP will determine whether the outline is qualified.
  • Step 123 The eligibility criterion of the contour is whether it is closed or not, the user is prompted to reselect the contour if it is unqualified, and the next step is performed if it is qualified.
  • Step 124 The qualified clothing contour images are uploaded to the cloud (ie, the server) for analysis and saved in the database.
  • Step 125 After the cloud algorithm performs image analysis, the recognition result is applied to the washing machine according to the prescribed logic, and the washing machine performs a corresponding washing procedure.
  • Step 126 The processing result of the cloud algorithm (ie, the recognition result) is returned to the APP client for display at the same time.
  • this embodiment is an embodiment of the method on the terminal side that interacts with the above-mentioned server-side embodiment.
  • the specific execution process has been described in detail in the interactive description of the above-mentioned embodiment.
  • the above-mentioned embodiment relates to the terminal side.
  • the executed part can be used to explain the above steps of this embodiment, and will not be repeated here.
  • the original image data can be sent to the server through network transmission, and the captured original image can be displayed on the terminal for the user to outline the original image.
  • the terminal can obtain the image contour path outlined by the user, and determine the image contour information according to the image contour path, and the terminal can send the image contour information to the server.
  • the clothing color recognition processing device on the server can obtain the original image data and the determined image contour information, and determine the segmented image data of the laundry based on the original image data and the image contour information, and use the clothing color recognition algorithm based on the HSV space Perform color recognition on the segmented image data to obtain the recognition result; send the recognition result to the terminal and the washing machine. After the terminal receives the recognition result, it will display the recognition result. After the washing machine receives the recognition result, it will control according to the recognition result and preset control logic. The washing machine carries out the corresponding treatment.
  • the user terminal realizes image shooting without installing a camera on the washing machine, which reduces the cost of the washing machine.
  • Yet another embodiment of the present application provides a processing device for clothing color recognition, which is used to execute the method provided by the foregoing server-side embodiment.
  • the processing device 150 for clothing color recognition includes a first receiving module 151, a first determining module 152, a processing module 153, and a first sending module 154.
  • the first receiving module is used to receive the photographed original image data and the image contour information of the laundry sent by the terminal; the first determining module is used to determine the segmented image of the laundry according to the original image data and the image contour information Data; processing module, used to use the HSV space-based clothing color recognition algorithm to perform color recognition on the segmented image data to obtain the recognition result; the first sending module is used to send the recognition result to the washing machine, so that the washing machine can respond according to the recognition result Processing.
  • the first sending module is further configured to send the recognition result to the terminal for display.
  • the processing device for clothing color recognition after the clothing scene image (or original image) is captured through the terminal, the original image data can be sent to the server through network transmission, and the captured original image can be displayed on the terminal for the user to outline For the contour of the laundry in the original image, the terminal can obtain the image contour path outlined by the user, and determine the image contour information according to the image contour path, and the terminal can send the image contour information to the server.
  • the clothing color recognition processing device on the server can obtain the original image data and the determined image contour information, and determine the segmented image data of the laundry based on the original image data and the image contour information, and use the clothing color recognition algorithm based on the HSV space Perform color recognition on the segmented image data to obtain the recognition result; send the recognition result to the terminal and the washing machine. After the terminal receives the recognition result, it will display the recognition result. After the washing machine receives the recognition result, it will control according to the recognition result and preset control logic. The washing machine carries out the corresponding treatment.
  • the user terminal realizes image shooting without installing a camera on the washing machine, which reduces the cost of the washing machine.
  • Another embodiment of the present application further supplements the device provided in the foregoing embodiment.
  • the image contour information includes position information of the contour of the laundry on the original image; the first determining module is specifically used for:
  • the segmented image data of the laundry is segmented from the original image data.
  • the processing module is specifically used for:
  • the color composition of the laundry is determined as the recognition result.
  • processing module is also used to:
  • the processing device for clothing color recognition after the clothing scene image (or original image) is captured through the terminal, the original image data can be sent to the server through network transmission, and the captured original image can be displayed on the terminal for the user to outline For the contour of the laundry in the original image, the terminal can obtain the image contour path outlined by the user, and determine the image contour information according to the image contour path, and the terminal can send the image contour information to the server.
  • the clothing color recognition processing device on the server can obtain the original image data and the determined image contour information, and determine the segmented image data of the laundry based on the original image data and the image contour information, and use the clothing color recognition algorithm based on the HSV space Perform color recognition on the segmented image data to obtain the recognition result; send the recognition result to the terminal and the washing machine. After the terminal receives the recognition result, it will display the recognition result. After the washing machine receives the recognition result, it will control according to the recognition result and preset control logic. The washing machine carries out the corresponding treatment.
  • the user terminal realizes image shooting without installing a camera on the washing machine, which reduces the cost of the washing machine.
  • Yet another embodiment of the present application provides a processing device for clothing color recognition, which is used to execute the method provided by the foregoing terminal-side embodiment.
  • the processing device 170 for clothing color recognition includes an acquiring module 171, a second sending module 172, a display module 173, a second determining module 174, and a second receiving module 175.
  • the acquisition module is used to obtain the original image data taken; the second sending module is used to send the original image data to the cloud server; the display module is used to display the corresponding original image on the contour input interface according to the original image data; The module is also used to obtain the image contour path of the laundry input by the user on the contour input interface; the second determining module is used to determine the image contour information according to the image contour path; the second sending module is also used to send the image contour information To the cloud server; the second receiving module is used for the recognition result of the color of the laundry returned by the cloud server; the display module is also used for displaying the recognition result of the color of the laundry returned by the cloud server.
  • the processing device for clothing color recognition after the clothing scene image (or original image) is captured through the terminal, the original image data can be sent to the server through network transmission, and the captured original image can be displayed on the terminal for the user to outline For the contour of the laundry in the original image, the terminal can obtain the image contour path outlined by the user, and determine the image contour information according to the image contour path, and the terminal can send the image contour information to the server.
  • the clothing color recognition processing device on the server can obtain the original image data and the determined image contour information, and determine the segmented image data of the laundry based on the original image data and the image contour information, and use the clothing color recognition algorithm based on the HSV space Perform color recognition on the segmented image data to obtain the recognition result; send the recognition result to the terminal and the washing machine. After the terminal receives the recognition result, it will display the recognition result. After the washing machine receives the recognition result, it will control according to the recognition result and preset control logic. The washing machine carries out the corresponding treatment.
  • the user terminal realizes image shooting without installing a camera on the washing machine, which reduces the cost of the washing machine.
  • Another embodiment of the present application further supplements the device provided in the foregoing embodiment.
  • the second determining module is further used for:
  • the second determining module is specifically used for:
  • the image contour information is determined according to the image contour path.
  • the second determining module is specifically used for:
  • a preset amount of position information is obtained from the image contour path as the image contour information.
  • the processing device for clothing color recognition after the clothing scene image (or original image) is captured through the terminal, the original image data can be sent to the server through network transmission, and the captured original image can be displayed on the terminal for the user to outline For the contour of the laundry in the original image, the terminal can obtain the image contour path outlined by the user, and determine the image contour information according to the image contour path, and the terminal can send the image contour information to the server.
  • the clothing color recognition processing device on the server can obtain the original image data and the determined image contour information, and determine the segmented image data of the laundry based on the original image data and the image contour information, and use the clothing color recognition algorithm based on the HSV space Perform color recognition on the segmented image data to obtain the recognition result; send the recognition result to the terminal and the washing machine. After the terminal receives the recognition result, it will display the recognition result. After the washing machine receives the recognition result, it will control according to the recognition result and preset control logic. The washing machine carries out the corresponding treatment.
  • the user terminal realizes image shooting without installing a camera on the washing machine, which reduces the cost of the washing machine.
  • Yet another embodiment of the present application provides a server for executing the method provided by the foregoing server-side embodiment.
  • the server 180 includes: at least one processor 181 and a memory 182;
  • the memory is used to store computer-executable instructions, so that at least one processor executes the computer-executable instructions to implement the methods provided in the foregoing embodiments.
  • the original image data can be sent to the server through network transmission, and the captured original image can be displayed on the terminal for the user to outline the laundry in the original image
  • the terminal can obtain the image contour path outlined by the user, and determine the image contour information according to the image contour path, and the terminal can send the image contour information to the server.
  • the clothing color recognition processing device on the server can obtain the original image data and the determined image contour information, and determine the segmented image data of the laundry based on the original image data and the image contour information, and use the clothing color recognition algorithm based on the HSV space Perform color recognition on the segmented image data to obtain the recognition result; send the recognition result to the terminal and the washing machine.
  • the terminal After the terminal receives the recognition result, it will display the recognition result. After the washing machine receives the recognition result, it will control according to the recognition result and preset control logic.
  • the washing machine carries out the corresponding treatment. Through the user's outline of clothing, the accurate segmentation of the image background and clothing area is realized, and the clothing color recognition algorithm based on HSV space is used to recognize the segmented image data, which effectively improves the accuracy of clothing color recognition, thereby providing users with accurate The laundry service improves the user experience.
  • the user terminal realizes image shooting without installing a camera on the washing machine, which reduces the cost of the washing machine.
  • Another embodiment of the present application provides a terminal for executing the method provided in the foregoing terminal-side embodiment.
  • the terminal 190 includes: at least one processor 191 and a memory 192;
  • the memory is used to store computer-executable instructions, so that at least one processor executes the computer-executable instructions to implement the method provided by the foregoing washing machine-side embodiment.
  • the original image data can be sent to the server through network transmission, and the captured original image can be displayed on the terminal for the user to outline the laundry in the original image
  • the terminal can obtain the image contour path outlined by the user, and determine the image contour information according to the image contour path, and the terminal can send the image contour information to the server.
  • the clothing color recognition processing device on the server can obtain the original image data and the determined image contour information, and determine the segmented image data of the laundry based on the original image data and the image contour information, and use the clothing color recognition algorithm based on the HSV space Perform color recognition on the segmented image data to obtain the recognition result; send the recognition result to the terminal and the washing machine.
  • the terminal After the terminal receives the recognition result, it will display the recognition result. After the washing machine receives the recognition result, it will control according to the recognition result and preset control logic.
  • the washing machine carries out the corresponding treatment. Through the user's outline of clothing, the accurate segmentation of the image background and clothing area is realized, and the clothing color recognition algorithm based on HSV space is used to recognize the segmented image data, which effectively improves the accuracy of clothing color recognition, thereby providing users with accurate The laundry service improves the user experience.
  • the user terminal realizes image shooting without installing a camera on the washing machine, which reduces the cost of the washing machine.
  • the computer-readable storage medium stores computer-executable instructions.
  • the computer-executable instructions are executed by a processor, they are used to implement the server-side execution method provided by any of the foregoing embodiments. .
  • a clothing scene image (ie, an original image) is captured through a terminal camera, and a flash can also be used to fill light, so as to help capture a clear and analyzable image.
  • the terminal After the terminal captures the clothing scene image (or original image), it can send the original image data to the server through network transmission, and display the captured original image on the terminal. Specifically, it can be displayed on the contour input interface for the user to outline the original image.
  • the contour of the laundry the user can outline the contour of the laundry in the original image on the original image of the contour input interface, the terminal can obtain the image contour path outlined by the user, and determine the image contour information according to the image contour path.
  • the terminal may send the image contour information to the server.
  • the clothing color recognition processing device on the server can obtain the original image data and the determined image contour information.
  • the original image data and image contour information determine the segmented image data of the laundry.
  • the segmented image data refers to the segmented image data that only includes the image of the laundry. Image data; use the clothing color recognition algorithm based on HSV space to perform color recognition on the segmented image data to obtain the recognition result; send the recognition result to the terminal and the washing machine.
  • the recognition result is displayed, and the washing machine receives the recognition After the result, the washing machine is controlled to perform corresponding processing according to the recognition result and the preset control logic.
  • the clothing color recognition algorithm based on HSV space is used to recognize the segmented image data, which effectively improves the accuracy of clothing color recognition, thereby providing users with accurate
  • the laundry service improves the user experience.
  • the user terminal realizes image shooting without installing a camera on the washing machine, which reduces the cost of the washing machine.
  • Another embodiment of the present application provides a computer-readable storage medium in which computer-executable instructions are stored.
  • the computer-executable instructions are executed by a processor, the computer-executable instructions are used to implement any one of the above-mentioned embodiments to provide a washing machine-side execution method.
  • the clothing scene image (ie, the original image) is captured through the terminal camera, and the flash can also be used to fill light to help capture clear and analyzable images.
  • the terminal After the terminal captures the clothing scene image (or original image), it can send the original image data to the server through network transmission, and display the captured original image on the terminal. Specifically, it can be displayed on the contour input interface for the user to outline the original image.
  • the contour of the laundry the user can outline the contour of the laundry in the original image on the original image of the contour input interface, the terminal can obtain the image contour path outlined by the user, and determine the image contour information according to the image contour path.
  • the terminal may send the image contour information to the server.
  • the clothing color recognition processing device on the server can obtain the original image data and the determined image contour information.
  • the original image data and image contour information determine the segmented image data of the laundry.
  • the segmented image data refers to the segmented image data that only includes the image of the laundry.
  • Image data use the clothing color recognition algorithm based on HSV space to perform color recognition on the segmented image data to obtain the recognition result; send the recognition result to the terminal and the washing machine, after the terminal receives the recognition result, the recognition result is displayed, and the washing machine receives the recognition After the result, the washing machine is controlled to perform corresponding processing according to the recognition result and the preset control logic.
  • the clothing color recognition algorithm based on HSV space is used to recognize the segmented image data, which effectively improves the accuracy of clothing color recognition, thereby providing users with accurate
  • the laundry service improves the user experience.
  • the user terminal realizes image shooting without installing a camera on the washing machine, which reduces the cost of the washing machine.
  • the disclosed device and method can be implemented in other ways.
  • the device embodiments described above are merely illustrative.
  • the division of the units is only a logical function division, and there may be other divisions in actual implementation, for example, multiple units or components may be combined or It can be integrated into another system, or some features can be ignored or not implemented.
  • the displayed or discussed mutual coupling or direct coupling or communication connection may be indirect coupling or communication connection through some interfaces, devices or units, and may be in electrical, mechanical or other forms.
  • the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the objectives of the solutions of the embodiments.
  • the functional units in the various embodiments of the present application may be integrated into one processing unit, or each unit may exist alone physically, or two or more units may be integrated into one unit.
  • the above-mentioned integrated unit may be implemented in the form of hardware, or may be implemented in the form of hardware plus software functional units.

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Multimedia (AREA)
  • Theoretical Computer Science (AREA)
  • Textile Engineering (AREA)
  • Image Analysis (AREA)

Abstract

一种衣物颜色识别的处理方法、装置、设备及存储介质，其方法通过拍摄衣物的原始图像数据，确定待洗衣物的分割图像数据，采用基于HSV空间的衣物颜色识别算法对待洗衣物的分割图像数据进行颜色识别，获得颜色识别结果，有效提高了衣物颜色识别的准确性，从而将识别结果发送给洗衣机（90），以使洗衣机（90）根据识别结果进行相应的处理，能够基于颜色识别为用户提供准确的洗衣服务，提高用户体验。

Description

衣物颜色识别的处理方法、装置、设备及存储介质
本申请要求于2019年11月22日提交中国专利局、申请号为201911152480.2、申请名称为“衣物颜色识别的处理方法、装置、设备及存储介质”的中国专利申请，以及于2019年11月22日提交中国专利局、申请号为201911152484.0、申请名称为“衣物颜色识别的处理方法、装置、设备及存储介质”的中国专利申请的优先权，其全部内容通过引用结合在本申请中。
技术领域
本申请涉及智能设备技术领域，尤其涉及一种衣物颜色识别的处理方法、装置、设备及存储介质。
背景技术
衣物的颜色千差万别，除了常见的红、橙、黄、绿、青、蓝、紫色之外，还有各种组合色，对于什么颜色的衣物可以混合在一起洗涤，现有技术通常是靠们根据经验来判断，判断结果不够准确，容易造成将不能混合洗涤的衣物一起洗涤，从而造成串色，导致用户体验较差。
因此，如何有效准确地判断衣物是否可以混合洗涤，成为亟需解决的技术问题。
发明内容
为了解决现有技术中的上述问题，即为了解决现有用户不能准确判断衣物是否可以混洗的技术问题，本申请提供了一种衣物颜色识别的处理方法、装置、设备及存储介质，以提高判断的准确性。
本申请第一个方面提供一种衣物颜色识别的处理方法，包括：
接收洗衣机发送的筒内的原始图像数据；
根据所述原始图像数据，采用预先训练好的分类网络模型，确定待洗衣物的分割图像数据；
采用基于HSV空间的衣物颜色识别算法对所述分割图像数据进行颜色识别，获得识别结果；
将所述识别结果发送给洗衣机，以使所述洗衣机根据所述识别结果进行相应的处理。
在上述方法的优选技术方案中，所述根据所述原始图像数据，采用预先训练好的分类网络模型，确定待洗衣物的分割图像数据，包括：
根据所述原始图像数据，采用预先训练好的基于深度学习的二分类神经网络模型，确定待洗衣物的分割图像数据。
在上述方法的优选技术方案中，所述采用基于HSV空间的衣物颜色识别算法对所述分割图像数据进行颜色识别，获得识别结果，包括：
将所述分割图像数据从RGB空间转换到HSV空间，获得转换后的分割图像数据；
获取衣物颜色表征范围；
基于转换后的分割图像数据及所述衣物颜色表征范围，确定待洗衣物的颜色组成，作为所述识别结果。
在上述方法的优选技术方案中，在基于转换后的分割图像数据及预先获得的衣物颜色表征范围，确定待洗衣物的颜色组成之后，所述方法还包括：
对所述待洗衣物的颜色组成进行过滤，获得所述待洗衣物的主色调颜色组成；
将所述待洗衣物的主色调颜色组成作为所述识别结果。
本申请第二个方面提供一种衣物颜色识别的处理方法，包括：
采集洗衣机筒内的原始图像数据；
将所述原始图像数据发送给服务器；
接收所述服务器返回的待洗衣物的颜色的识别结果；
根据所述识别结果，确定对应的目标控制程序；
根据所述目标控制程序控制所述洗衣机进行相应的处理。
在上述方法的优选技术方案中，所述根据所述识别结果，确定对应的目标控制程序，包括：
根据所述识别结果及预设控制逻辑规则，确定对应的目标控制程序，所述预设控制逻辑规则包括颜色与控制程序的对应关系。
在上述方法的优选技术方案中，所述根据所述识别结果及预设控制逻辑规则，确定对应的目标控制程序，包括：
若根据所述识别结果确定待洗衣物具有串色风险，则确定目标控制程序为告警提示程序；
所述根据所述目标控制程序控制所述洗衣机进行相应的处理，包括：
根据所述告警提示程序控制所述洗衣机发出告警提示。
本申请第三个方面提供一种衣物颜色识别的处理装置，包括：
第一接收模块，用于接收洗衣机发送的筒内的原始图像数据；
第一确定模块，用于根据所述原始图像数据，采用预先训练好的分类网络模型，确定待洗衣物的分割图像数据；
处理模块，用于采用基于HSV空间的衣物颜色识别算法对所述分割图像数据进行颜色识别，获得识别结果；
第一发送模块，用于将所述识别结果发送给洗衣机，以使所述洗衣机根据所述识别结果进行相应的处理。
在上述装置的优选技术方案中，所述第一确定模块，具体用于：
根据所述原始图像数据，采用预先训练好的基于深度学习的二分类神经网络模型，确定待洗衣物的分割图像数据。
在上述装置的优选技术方案中，所述处理模块，具体用于：
将所述分割图像数据从RGB空间转换到HSV空间，获得转换后的分割图像数据；
获取衣物颜色表征范围；
基于转换后的分割图像数据及所述衣物颜色表征范围，确定待洗衣物的颜色组成，作为所述识别结果。
在上述装置的优选技术方案中，所述处理模块，还用于：
对所述待洗衣物的颜色组成进行过滤，获得所述待洗衣物的主色调颜色组成；
将所述待洗衣物的主色调颜色组成作为所述识别结果。
本申请第四个方面提供一种衣物颜色识别的处理装置，包括：
获取模块，用于采集洗衣机筒内的原始图像数据；
第二发送模块，用于将所述原始图像数据发送给服务器；
第二接收模块，用于接收所述服务器返回的待洗衣物的颜色的识别结果；
第二确定模块，用于根据所述识别结果，确定对应的目标控制程序；
控制模块，用于根据所述目标控制程序控制所述洗衣机进行相应的处理。
在上述装置的优选技术方案中，所述第二确定模块，具体用于：
根据所述识别结果及预设控制逻辑规则，确定对应的目标控制程序，所述预设控制逻辑规则包括颜色与控制程序的对应关系。
在上述装置的优选技术方案中，所述第二确定模块，具体用于：
若根据所述识别结果确定待洗衣物具有串色风险，则确定目标控制程序为告警提示程序；
所述控制模块，具体用于：
根据所述告警提示程序控制所述洗衣机发出告警提示。
本申请第五个方面提供一种服务器，包括：至少一个处理器和存储器；
所述存储器用于存储计算机可执行指令，以使所述至少一个处理器执行所述计算机可执行指令实现第一个方面提供的方法。
本申请第六个方面提供一种洗衣机，包括：至少一个处理器和存储器；
所述存储器用于存储计算机可执行指令，以使所述至少一个处理器执行所述计算机可执行指令实现第二个方面提供的方法。
本申请第七个方面提供一种计算机可读存储介质，所述计算机可读存储介质中存储有计算机执行指令，所述计算机执行指令被处理器执行时用于实现第一个方面提供的方法。
本申请第八个方面提供一种计算机可读存储介质，所述计算机可读存储介质中存储有计算机执行指令，所述计算机执行指令被处理器执行时用于实现第二个方面提供的方法。
本申请第九个方面提供一种计算机程序，包括程序代码，当计算机运行所述计算机程序时，所述程序代码执行如第一个方面提供的方法。
本申请第十个方面提供一种计算机程序，包括程序代码，当计算机运行所述计算机程序时，所述程序代码执行如第二个方面提供的方法。
本申请第十一个方面提供一种衣物颜色识别的处理方法，包括：
接收终端发送的拍摄的原始图像数据及待洗衣物的图像轮廓信息；
根据所述原始图像数据及所述图像轮廓信息，确定待洗衣物的分割图像数据；
采用基于HSV空间的衣物颜色识别算法对所述分割图像数据进行颜色识别，获得识别结果；
将所述识别结果发送给洗衣机，以使所述洗衣机根据所述识别结果进行相应的处 理。
在上述方法的优选技术方案中，所述图像轮廓信息包括待洗衣物在原始图像上的轮廓的位置信息；
所述根据所述原始图像数据及所述图像轮廓信息，确定待洗衣物的分割图像数据，包括：
根据所述图像轮廓信息，从所述原始图像数据中分割出所述待洗衣物的分割图像数据。
在上述方法的优选技术方案中，所述采用基于HSV空间的衣物颜色识别算法对所述分割图像数据进行颜色识别，获得识别结果，包括：
将所述分割图像数据从RGB空间转换到HSV空间，获得转换后的分割图像数据；
获取衣物颜色表征范围；
基于转换后的分割图像数据及所述衣物颜色表征范围，确定待洗衣物的颜色组成，作为所述识别结果。
在上述方法的优选技术方案中，在基于转换后的分割图像数据及所述衣物颜色表征范围，确定待洗衣物的颜色组成之后，所述方法还包括：
对所述待洗衣物的颜色组成进行过滤，获得所述待洗衣物的主色调颜色组成；
将所述待洗衣物的主色调颜色组成作为所述识别结果。
本申请第十二个方面提供一种衣物颜色识别的处理方法，包括：
获取拍摄的原始图像数据；
将所述原始图像数据发送给云服务器，并根据所述原始图像数据在轮廓输入界面展示对应的原始图像；
获取用户在所述轮廓输入界面输入的待洗衣物的图像轮廓路径；
根据所述图像轮廓路径确定图像轮廓信息；
将所述图像轮廓信息发送给所述云服务器；
接收并展示所述云服务器返回的待洗衣物的颜色的识别结果
在上述方法的优选技术方案中，在所述获取用户在所述轮廓输入界面输入的待洗衣物的图像轮廓路径之后，所述方法还包括：
根据所述图像轮廓路径判断图像轮廓是否合格；
若所述图像轮廓路径为封闭路径，则确定图像轮廓合格；
若所述图像轮廓路径为非封闭路径，则确定图像轮廓不合格；
所述根据所述图像轮廓路径确定图像轮廓信息，包括：
若图像轮廓合格，则根据所述图像轮廓路径确定图像轮廓信息。
在上述方法的优选技术方案中，所述根据所述图像轮廓路径确定图像轮廓信息，包括：
从所述图像轮廓路径中获取预设数量的位置信息作为所述图像轮廓信息。
本申请第十三个方面提供一种衣物颜色识别的处理装置，包括：
第一接收模块，用于接收终端发送的拍摄的原始图像数据及待洗衣物的图像轮廓信息；
第一确定模块，用于根据所述原始图像数据及所述图像轮廓信息，确定待洗衣物 的分割图像数据；
处理模块，用于采用基于HSV空间的衣物颜色识别算法对所述分割图像数据进行颜色识别，获得识别结果；
第一发送模块，用于将所述识别结果发送给洗衣机，以使所述洗衣机根据所述识别结果进行相应的处理
在上述装置的优选技术方案中，所述图像轮廓信息包括待洗衣物在原始图像上的轮廓的位置信息；所述第一确定模块，具体用于：
根据所述图像轮廓信息，从所述原始图像数据中分割出所述待洗衣物的分割图像数据。
在上述装置的优选技术方案中，所述处理模块，具体用于：
将所述分割图像数据从RGB空间转换到HSV空间，获得转换后的分割图像数据；
获取衣物颜色表征范围；
基于转换后的分割图像数据及所述衣物颜色表征范围，确定待洗衣物的颜色组成，作为所述识别结果。
在上述装置的优选技术方案中，所述处理模块，还用于：
对所述待洗衣物的颜色组成进行过滤，获得所述待洗衣物的主色调颜色组成；
将所述待洗衣物的主色调颜色组成作为所述识别结果。
本申请第十四个方面提供一种衣物颜色识别的处理装置，包括：
获取模块，用于获取拍摄的原始图像数据；
第二发送模块，用于将所述原始图像数据发送给云服务器；
展示模块，用于根据所述原始图像数据在轮廓输入界面展示对应的原始图像；
所述获取模块，还用于获取用户在所述轮廓输入界面输入的待洗衣物的图像轮廓路径；
第二确定模块，用于根据所述图像轮廓路径确定图像轮廓信息；
所述第二发送模块，还用于将所述图像轮廓信息发送给所述云服务器；
第二接收模块，用于所述云服务器返回的待洗衣物的颜色的识别结果；
所述展示模块，还用于展示所述云服务器返回的待洗衣物的颜色的识别结果。
在上述装置的优选技术方案中，所述第二确定模块，还用于：
根据所述图像轮廓路径判断图像轮廓是否合格；
若所述图像轮廓路径为封闭路径，则确定图像轮廓合格；
若所述图像轮廓路径为非封闭路径，则确定图像轮廓不合格；
所述第二确定模块，具体用于：
若图像轮廓合格，则根据所述图像轮廓路径确定图像轮廓信息。
在上述装置的优选技术方案中，所述第二确定模块，具体用于：
从所述图像轮廓路径中获取预设数量的位置信息作为所述图像轮廓信息。
本申请第十五个方面提供一种服务器，包括：至少一个处理器和存储器；
所述存储器用于存储计算机可执行指令，以使所述至少一个处理器执行所述计算机可执行指令实现第十一个方面提供的方法。
本申请第十六个方面提供一种终端，包括：至少一个处理器和存储器；
所述存储器用于存储计算机可执行指令，以使所述至少一个处理器执行所述计算机可执行指令实现第十二个方面提供的方法。
本申请第十七个方面提供一种计算机可读存储介质，所述计算机可读存储介质中存储有计算机执行指令，所述计算机执行指令被处理器执行时用于实现第十一个方面提供的方法。
本申请第十八个方面提供一种计算机可读存储介质，所述计算机可读存储介质中存储有计算机执行指令，所述计算机执行指令被处理器执行时用于实现第十二个方面提供的方法。
本申请第十九个方面提供一种计算机程序，包括程序代码，当计算机运行所述计算机程序时，所述程序代码执行如第十一个方面提供的方法。
本申请第二十个方面提供一种计算机程序，包括程序代码，当计算机运行所述计算机程序时，所述程序代码执行如第十二个方面提供的方法。
本领域技术人员能够理解的是，本申请的衣物颜色识别的处理方法、装置、设备及存储介质，通过拍摄衣物的原始图像数据，确定待洗衣物的分割图像数据，采用基于HSV空间的衣物颜色识别算法对待洗衣物的分割图像数据进行颜色识别，获得颜色识别结果，有效提高了衣物颜色识别的准确性，从而将识别结果发送给洗衣机，以使洗衣机根据识别结果进行相应的处理，能够基于颜色识别为用户提供准确的洗衣服务，提高用户体验。
附图说明
为了更清楚地说明本申请实施例或现有技术中的技术方案，下面将对实施例或现有技术描述中所需要使用的附图作一简单地介绍，显而易见地，下面描述中的附图是本申请的一些实施例，对于本领域普通技术人员来讲，在不付出创造性劳动性的前提下，还可以根据这些附图获得其他的附图。
图1为本申请一实施例提供的处理系统的架构示意图；
图2为本申请一实施例提供的衣物颜色识别的处理方法的流程示意图；
图3为本申请另一实施例提供的衣物颜色识别的处理方法的流程示意图；
图4为本申请再一实施例提供的衣物颜色识别的处理方法的流程示意图；
图5为本申请又一实施例提供的衣物颜色识别的处理方法的流程示意图；
图6为本申请一实施例提供的衣物颜色识别的处理装置的结构示意图；
图7为本申请另一实施例提供的衣物颜色识别的处理装置的结构示意图；
图8为本申请一实施例提供的服务器的结构示意图；
图9为本申请一实施例提供的洗衣机的结构示意图；
图10为本申请另一实施例提供的处理系统的架构示意图；
图11为本申请一实施例提供的衣物颜色识别的处理方法的流程示意图；
图12为本申请另一实施例提供的衣物颜色识别的处理方法的流程示意图；
图13为本申请再一实施例提供的衣物颜色识别的处理方法的流程示意图；
图14为本申请又一实施例提供的衣物颜色识别的处理方法的流程示意图；
图15为本申请一实施例提供的工作流程示意图；
图16为本申请一实施例提供的衣物颜色识别的处理装置的结构示意图；
图17为本申请另一实施例提供的衣物颜色识别的处理装置的结构示意图；
图18为本申请一实施例提供的服务器的结构示意图；
图19为本申请一实施例提供的终端的结构示意图。
通过上述附图，已示出本申请明确的实施例，后文中将有更详细的描述。这些附图和文字描述并不是为了通过任何方式限制本公开构思的范围，而是通过参考特定实施例为本领域技术人员说明本申请的概念。
具体实施方式
为使本申请实施例的目的、技术方案和优点更加清楚，下面将结合本申请实施例中的附图，对本申请实施例中的技术方案进行清楚、完整地描述，显然，所描述的实施例是本申请一部分实施例，而不是全部的实施例。基于本申请中的实施例，本领域普通技术人员在没有作出创造性劳动前提下所获得的所有其他实施例，都属于本申请保护的范围。
首先，本领域技术人员应当理解的是，这些实施方式仅仅用于解释本申请，并非旨在限制本申请的保护范围。本领域技术人员可以根据需要对其作出调整，以便适应具体的应用场合。此外，需要说明的是，术语“第一”、“第二”等仅用于描述目的，而不能理解为指示或暗示相对重要性或者隐含指明所指示的技术特征的数量。在以下各实施例的描述中，“多个”的含义是两个以上，除非另有明确具体的限定。
首先，对本申请涉及到的名词进行解释：
RGB空间：是以R(Red:红)、G(Green:绿)、B(Blue:蓝)三种基本色为基础，进行不同程度的叠加，产生丰富而广泛的颜色，所以俗称三基色模式。是用一个单位长度的立方体来表示颜色的，黑蓝绿青红紫黄白8种常见颜色分别位居立方体的8个顶点，通常将黑色置于三维直角坐标系的原点，红绿蓝分别置于3根坐标轴土，整个立方体放在第1卦限内。而其中的青色与红色、紫色(或称品红色)与绿色、黄色与蓝色是互补色。各参数的取值范围是：R：0-255；G：0-255；B：0-255。参数值也称为三色系数或基色系数或颜色值，除以255后归一到0-1之间，但不是无穷多个而是有限多个值。由于每个灰度级都定为256，所以，红绿蓝分量全部组合起来共可表示256＝2＝16777216种不同的颜色。它比人眼能分辨的颜色种数多得多。
HSV空间：是指HSV(Hue(色调)，Saturation(饱和度)，Value(明度))颜色空间，也称六角锥体模型空间。
本申请的一种实施例提供的衣物颜色识别的处理方法，适用于在用户洗衣时，自动识别衣物颜色，并根据衣物颜色选择合适的洗衣程序进行洗涤的应用场景。参阅图1，图1为本申请下述涉及洗衣机和服务器交互的实施例基于的处理系统的架构示意图。如图1所示，该处理系统可以包括洗衣机和服务器，服务器可以是云服务器，也可以是其他服务器。洗衣机可以是滚筒洗衣机，也可以是其他类型的洗衣机，以滚筒洗衣机为例，服务器和洗衣机上均可以设置相应的衣物颜色识别的处理装置。还可以在洗衣机的门窗上方安装摄像机，比如安装迷你摄像机，用于拍摄筒内衣物场景图像，还可以借助滚筒内灯光补光，以助于拍摄到清晰可分析的图像。摄像机拍摄筒内衣物场 景图像(即原始图像)后，可以通过网络传输将原始图像数据发送到服务器。或者洗衣机的衣物颜色识别的处理装置获取摄像机拍摄的原始图像数据，通过网络传输发送给服务器，服务器上的衣物颜色识别的处理装置则可以获取到原始图像数据。服务器接收到原始图像数据后，根据原始图像数据，采用预先训练好的分类网络模型，确定待洗衣物的分割图像数据，分割图像数据是指从原始图像数据中分割出来的仅包括待洗衣物图像部分的图像数据；采用基于HSV空间的衣物颜色识别算法对分割图像数据进行颜色识别，获得识别结果；将识别结果发送给洗衣机，洗衣机接收到识别结果后，根据识别结果及预设控制逻辑控制洗衣机进行相应的处理。通过神经网络模型将待洗衣物与背景准确分割，并采用基于HSV空间的衣物颜色识别算法对分割图像数据进行颜色识别，有效提高了衣物颜色识别的准确性，从而能够为用户提供准确的洗衣服务，提高用户体验。
下面这几个具体的实施例可以相互结合，对于相同或相似的概念或过程可能在某些实施例中不再赘述。下面将结合附图，对本发明的实施例进行描述。
本申请一实施例提供一种衣物颜色识别的处理方法，用于识别洗衣机筒内待洗衣物的颜色。本实施例的执行主体为衣物颜色识别的处理装置，该装置可以设置在服务器中。
首先参阅图2，图2为本实施例提供的衣物颜色识别的处理方法的流程示意图，该方法包括：
步骤101，接收洗衣机发送的筒内的原始图像数据。
具体的，可以在洗衣机的门窗上方安装摄像机，比如安装迷你摄像机，用于拍摄筒内衣物场景图像(即原始图像)，获得洗衣机筒内原始图像数据，通过洗衣机发送给服务器。
可选地，可以是当用户打开洗衣机门窗，放入衣物，并关上门窗时触发洗衣机摄像机拍摄筒内衣物场景图像，也可以是在用户关上门窗，启动洗衣机时触发，具体可以根据实际需求设置，本实施例不做限定。
步骤102，根据原始图像数据，采用预先训练好的分类网络模型，确定待洗衣物的分割图像数据。
具体的，可以预先采集筒内的原始训练图像数据并对原始训练图像数据进行标注，获得原始训练图像数据对应的标签数据；并建立分类神经网络，基于原始训练图像数据和对应的标签数据训练建立好的分类神经网络，获得分类网络模型。
在接收到原始图像数据后，则可以根据原始图像数据，采用预先训练好的分类网络模型，确定待洗衣物的分割图像数据。分割图像数据是指从原始图像数据中分割出来的仅包括待洗衣物图像部分的图像数据，也即通过分类网络模型，将原始图像数据中的衣物和背景分割开来，得到原始图像数据中哪些部分属于背景部分，哪些部分属于衣物部分，属于衣物部分的即作为待洗衣物的分割图像数据。
可选地，分类网络模型为基于深度学习的二分类神经网络模型，兼顾了处理的速率和准确率，将原始图像分成衣物和背景两类。分类网络模型的网络架构可以包括一个级联网络，级联网络分为四层，每层会对原始对象(即原始图像数据)做不同程度的卷积下采样来提取语义特征，各层的输出特征图最终会融合为一个特征图，再经过 上采样层提取细节特征得到目标特征图，最后对所得的目标特征图进行常规语义分割，得到最终的分割图像。
步骤103，采用基于HSV空间的衣物颜色识别算法对分割图像数据进行颜色识别，获得识别结果。
具体的，在确定了待洗衣物的分割图像数据后，则可以采用基于HSV空间的衣物颜色识别算法对分割图像数据进行颜色识别，获得识别结果。
识别结果包括待洗衣物的颜色组成，比如红色80％白色20％。
可选地，基于HSV空间的衣物颜色识别算法为将分割图像数据从RGB空间转换到HSV空间进行颜色识别的算法。具体来说，将分割图像数据从RGB空间转换到HSV空间，获得转换后的分割图像数据；获取衣物颜色表征范围；基于转换后的分割图像数据及衣物颜色表征范围，确定待洗衣物的颜色组成，作为识别结果。其中，衣物颜色表征范围包括了不同范围代表不同的颜色，具体结合HSV六棱锥和大量颜色各异的衣服颜色表征实现来确定衣物颜色表征范围。比如结合HSV六棱锥和1000件颜色各异的衣服颜色表征实现，确定10种常用颜色的HSV表征范围。比如，黑色的表征范围为：
(h>＝0&&h<＝360)&&(s>＝0&&s<＝100)&&(v>＝0&&v<＝25)
其中，h、s、v分别代表三个通道的值。
步骤104，将识别结果发送给洗衣机，以使洗衣机根据识别结果进行相应的处理。
具体的，在获得待洗衣物的颜色的识别结果后，则将识别结果发送给洗衣机。洗衣机接收到待洗衣物的颜色的识别结果后，可以根据识别结果，确定对应的目标控制程序，根据目标控制程序控制洗衣机进行相应的处理。
示例性的，若识别结果为红色和白色，而且颜色组成比例大于第一阈值，则确定存在串色风险，则确定目标控制程序为不启动洗衣机并告警提示的程序，则根据目标控制程序控制不启动洗衣机，并控制洗衣机发出告警提示。具体的告警提示方式可以根据实际需求设置，比如可以是通过语音播报进行告警提示，也可以在洗衣机的显示界面显示相应的告警信息并发出告警提示音等等。
示例性的，若识别结果为白色，且大于第二阈值，则确定对应的目标控制程序为护色洗程序，则可以选择护色洗程序控制洗衣机进行护色洗流程。
示例性的，若识别结果为黑色和蓝色，则确定对应的目标控制程序为混合洗程序，则可以选择混合洗程序控制洗衣机进行混合洗流程。
可选地，可以预先在洗衣机设置预设控制逻辑规则，可以包括颜色与控制程序的对应关系，当洗衣机接收到识别结果后，根据识别结果及预设控制逻辑规则来确定目标控制程序。
本实施例提供的衣物颜色识别的处理方法，通过洗衣机上摄像机拍摄筒内的原始图像，洗衣机将原始图像数据发送给服务器，服务器接收洗衣机发送的筒内的原始图像数据；根据原始图像数据，采用预先训练好的分类网络模型，确定待洗衣物的分割图像数据；采用基于HSV空间的衣物颜色识别算法对分割图像数据进行颜色识别，获得识别结果；将识别结果发送给洗衣机，以使洗衣机根据识别结果进行相应的处理，有效提高了衣物颜色识别的准确性，从而能够为用户提供准确的洗衣服务，提高用户体验。
本申请另一实施例对上述实施例提供的方法做进一步补充说明。
参阅图3，图3为本实施例提供的衣物颜色识别的处理方法的流程示意图。
作为一种可实施的方式，在上述实施例的基础上，可选地，根据原始图像数据，采用预先训练好的分类网络模型，确定待洗衣物的分割图像数据，包括：
步骤1021，根据原始图像数据，采用预先训练好的基于深度学习的二分类神经网络模型，确定待洗衣物的分割图像数据。
具体的，分类网络模型为基于深度学习的二分类神经网络模型，兼顾了处理的速率和准确率，将原始图像分成衣物和背景两类。分类网络模型的网络架构可以包括一个级联网络，级联网络分为四层，每层会对原始对象(即原始图像数据)做不同程度的卷积下采样来提取语义特征，各层的输出特征图最终会融合为一个特征图，再经过上采样层提取细节特征得到目标特征图，最后对所得的目标特征图进行常规语义分割，得到最终的分割图像。
在采用分类网络模型之前，需要通过训练数据训练获得该分类网络模型。具体来说，可以预先采集筒内的原始训练图像数据并对原始训练图像数据进行标注，获得原始训练图像数据对应的标签数据；并建立分类神经网络，基于原始训练图像数据和对应的标签数据训练建立好的分类神经网络，获得分类网络模型。
示例性的，可以借助补光拍摄大量衣物图像，经过清洗(清洗掉颜色信息不明确，环境干扰噪声过大的图像)、标注，获得原始训练图像数据及对应的标签数据，基于原始训练图像数据和对应的标签数据对预先建立的基于深度学习的二分类神经网络进行训练，获得基于深度学习的二分类神经网络模型。
作为另一种可实施的方式，在上述实施例的基础上，可选地，采用基于HSV空间的衣物颜色识别算法对分割图像数据进行颜色识别，获得识别结果，包括：
步骤1031，将分割图像数据从RGB空间转换到HSV空间，获得转换后的分割图像数据。
步骤1032，获取衣物颜色表征范围。
步骤1033，基于转换后的分割图像数据及衣物颜色表征范围，确定待洗衣物的颜色组成。
步骤1034，将待洗衣物的颜色组成作为识别结果。
具体的，在获得待洗衣物的分割图像数据后，将分割图像数据从RGB空间转换到HSV空间，获得转换后的分割图像数据；获取衣物颜色表征范围；基于转换后的分割图像数据及衣物颜色表征范围，确定待洗衣物的颜色组成，作为识别结果。其中，衣物颜色表征范围包括了不同范围代表不同的颜色，具体可以结合HSV六棱锥和大量颜色各异的衣服颜色表征实现来确定衣物颜色表征范围。比如结合HSV六棱锥和1000件颜色各异的衣服颜色表征实现，确定10种常用颜色的HSV表征范围。比如，黑色的表征范围为：
(h>＝0&&h<＝360)&&(s>＝0&&s<＝100)&&(v>＝0&&v<＝25)
其中，h、s、v分别代表三个通道的值。
可选地，在基于转换后的分割图像数据及衣物颜色表征范围，确定待洗衣物的颜色组成之后，该方法还包括：
步骤1035，对待洗衣物的颜色组成进行过滤，获得待洗衣物的主色调颜色组成。
步骤1036，将待洗衣物的主色调颜色组成作为识别结果。
具体的，在基于转换后的分割图像数据及衣物颜色表征范围，确定待洗衣物的颜色组成之后，可能获得的颜色组成包括了非常小比例的其他颜色，可以对待洗衣物的颜色组成进行过滤，获得待洗衣物的主色调颜色组成，将待洗衣物的主色调颜色组成作为识别结果发送给洗衣机。
本申请实施例提供的衣物颜色识别的处理方法，实现了衣物颜色的自动识别，结合深度学习的二分类神经网络模型和基于HSV空间颜色识别算法，有效提高了颜色识别的准确性，省去了用户凭经验判断衣物颜色，手动选取洗涤模式的环节，解放用户双手，提高洗衣环节的便捷性，并且解决了用户凭经验判断衣物颜色不选取洗涤模式不够准确的问题。
需要说明的是，本实施例中各可实施的方式可以单独实施，也可以在不冲突的情况下以任意组合方式结合实施本申请不做限定。
本实施例提供的衣物颜色识别的处理方法，通过洗衣机上摄像机拍摄筒内的原始图像，洗衣机将原始图像数据发送给服务器，服务器接收洗衣机发送的筒内的原始图像数据；根据原始图像数据，采用预先训练好的分类网络模型，确定待洗衣物的分割图像数据；采用基于HSV空间的衣物颜色识别算法对分割图像数据进行颜色识别，获得识别结果；将识别结果发送给洗衣机，以使洗衣机根据识别结果进行相应的处理，有效提高了衣物颜色识别的准确性，从而能够为用户提供准确的洗衣服务，提高用户体验。
本申请再一实施例提供一种衣物颜色识别的处理方法，用于识别洗衣机筒内待洗衣物的颜色。本实施例的执行主体为衣物颜色识别的处理装置，该装置可以设置在洗衣机中。
首先参阅图4，图4为本实施例提供的衣物颜色识别的处理方法的流程示意图，该方法包括：
步骤301，采集洗衣机筒内的原始图像数据。
步骤302，将原始图像数据发送给服务器，以使服务器根据原始图像数据获得待洗衣物的颜色识别结果。
步骤303，接收服务器返回的待洗衣物的颜色的识别结果。
步骤304，根据识别结果，确定对应的目标控制程序。
步骤305，根据目标控制程序控制洗衣机进行相应的处理。
需要说明的是，本实施例是与上述服务器端实施例交互的洗衣机端的方法实施例，其具体执行过程已在上述实施例的交互描述中进行了详细说明，上述实施例涉及到的洗衣机侧执行的部分均可以用于解释本实施例的上述步骤，在此不再赘述。
本实施例提供的衣物颜色识别的处理方法，通过洗衣机上摄像机拍摄筒内的原始图像，洗衣机将原始图像数据发送给服务器，服务器接收洗衣机发送的筒内的原始图像数据；根据原始图像数据，采用预先训练好的分类网络模型，确定待洗衣物的分割图像数据；采用基于HSV空间的衣物颜色识别算法对分割图像数据进行颜色识别，获得识别结果；将识别结果发送给洗衣机，以使洗衣机根据识别结果进行相应的处理， 有效提高了衣物颜色识别的准确性，从而能够为用户提供准确的洗衣服务，提高用户体验。
本申请又一实施例对上述实施例提供的方法做进一步补充说明。
参阅图5，图5为本实施例提供的衣物颜色识别的处理方法的流程示意图。
作为一种可实施的方式，在上述实施例的基础上，可选地，根据识别结果，确定对应的目标控制程序，包括：
步骤3041，根据识别结果及预设控制逻辑规则，确定对应的目标控制程序。
其中，预设控制逻辑规则包括颜色与控制程序的对应关系。
作为另一种可实施的方式，在上述实施例的基础上，可选地，根据识别结果及预设控制逻辑规则，确定对应的目标控制程序，包括：
步骤30411，若根据识别结果确定待洗衣物具有串色风险，则确定目标控制程序为告警提示程序。
相应的，根据目标控制程序控制洗衣机进行相应的处理，包括：
步骤3051，根据告警提示程序控制洗衣机发出告警提示。
可选地，若识别结果为红色和白色，而且颜色组成比例大于第一阈值，则确定存在串色风险，则确定目标控制程序为不启动洗衣机并告警提示的程序，则根据目标控制程序控制不启动洗衣机，并控制洗衣机发出告警提示。具体的告警提示方式可以根据实际需求设置，比如可以是通过语音播报进行告警提示，也可以在洗衣机的显示界面显示相应的告警信息并发出告警提示音等等。
可选地，若识别结果为白色，且大于第二阈值，则确定对应的目标控制程序为护色洗程序，则可以选择护色洗程序控制洗衣机进行护色洗流程。
可选地，若识别结果为黑色和蓝色，则确定对应的目标控制程序为混合洗程序，则可以选择混合洗程序控制洗衣机进行混合洗流程。
需要说明的是，本实施例是与上述服务器端实施例交互的洗衣机端的方法实施例，其具体执行过程已在上述实施例的交互描述中进行了详细说明，上述实施例涉及到的洗衣机侧执行的部分均可以用于解释本实施例的上述步骤，在此不再赘述。
还需要说明的是，本实施例中各可实施的方式可以单独实施，也可以在不冲突的情况下以任意组合方式结合实施本申请不做限定。
本实施例提供的衣物颜色识别的处理方法，通过洗衣机上摄像机拍摄筒内的原始图像，洗衣机将原始图像数据发送给服务器，服务器接收洗衣机发送的筒内的原始图像数据；根据原始图像数据，采用预先训练好的分类网络模型，确定待洗衣物的分割图像数据；采用基于HSV空间的衣物颜色识别算法对分割图像数据进行颜色识别，获得识别结果；将识别结果发送给洗衣机，以使洗衣机根据识别结果进行相应的处理，有效提高了衣物颜色识别的准确性，从而能够为用户提供准确的洗衣服务，提高用户体验。
本申请再一实施例提供一种衣物颜色识别的处理装置，用于执行上述服务器端实施例提供的方法。
参阅图6，图6为本实施例提供的衣物颜色识别的处理装置的结构示意图。该衣物颜色识别的处理装置50包括第一接收模块51、第一确定模块52、处理模块53和第 一发送模块54。
其中，第一接收模块，用于接收洗衣机发送的筒内的原始图像数据；第一确定模块，用于根据原始图像数据，采用预先训练好的分类网络模型，确定待洗衣物的分割图像数据；处理模块，用于采用基于HSV空间的衣物颜色识别算法对分割图像数据进行颜色识别，获得识别结果；第一发送模块，用于将识别结果发送给洗衣机，以使洗衣机根据识别结果进行相应的处理。
关于本实施例中的装置，其中各个模块执行操作的具体方式已经在有关该方法的实施例中进行了详细描述，此处将不做详细阐述说明。
根据本实施例提供的衣物颜色识别的处理装置，通过洗衣机上摄像机拍摄筒内的原始图像，洗衣机将原始图像数据发送给服务器，服务器接收洗衣机发送的筒内的原始图像数据；根据原始图像数据，采用预先训练好的分类网络模型，确定待洗衣物的分割图像数据；采用基于HSV空间的衣物颜色识别算法对分割图像数据进行颜色识别，获得识别结果；将识别结果发送给洗衣机，以使洗衣机根据识别结果进行相应的处理，有效提高了衣物颜色识别的准确性，从而能够为用户提供准确的洗衣服务，提高用户体验。
本申请又一实施例对上述实施例提供的装置做进一步补充说明。
作为一种可实施的方式，在上述实施例的基础上，可选地，第一确定模块，具体用于：
根据原始图像数据，采用预先训练好的基于深度学习的二分类神经网络模型，确定待洗衣物的分割图像数据。
作为另一种可实施的方式，在上述实施例的基础上，可选地，处理模块，具体用于：
将分割图像数据从RGB空间转换到HSV空间，获得转换后的分割图像数据；获取衣物颜色表征范围；基于转换后的分割图像数据及衣物颜色表征范围，确定待洗衣物的颜色组成，作为识别结果。
可选地，处理模块，还用于：
对待洗衣物的颜色组成进行过滤，获得待洗衣物的主色调颜色组成；将待洗衣物的主色调颜色组成作为识别结果。
关于本实施例中的装置，其中各个模块执行操作的具体方式已经在有关该方法的实施例中进行了详细描述，此处将不做详细阐述说明。
需要说明的是，本实施例中各可实施的方式可以单独实施，也可以在不冲突的情况下以任意组合方式结合实施本申请不做限定。
根据本实施例提供的衣物颜色识别的处理装置，通过洗衣机上摄像机拍摄筒内的原始图像，洗衣机将原始图像数据发送给服务器，服务器接收洗衣机发送的筒内的原始图像数据；根据原始图像数据，采用预先训练好的分类网络模型，确定待洗衣物的分割图像数据；采用基于HSV空间的衣物颜色识别算法对分割图像数据进行颜色识别，获得识别结果；将识别结果发送给洗衣机，以使洗衣机根据识别结果进行相应的处理，有效提高了衣物颜色识别的准确性，从而能够为用户提供准确的洗衣服务，提高用户体验。
本申请再一实施例提供一种衣物颜色识别的处理装置，用于执行上述洗衣机端实施例提供的方法。
参阅图7，图7为本实施例提供的衣物颜色识别的处理装置的结构示意图。该衣物颜色识别的处理装置70包括获取模块71、第二发送模块72、第二接收模块73、第二确定模块74和控制模块75。
其中，获取模块，用于采集洗衣机筒内的原始图像数据；第二发送模块，用于将原始图像数据发送给服务器；第二接收模块，用于接收服务器返回的待洗衣物的颜色的识别结果；第二确定模块，用于根据识别结果，确定对应的目标控制程序；控制模块，用于根据目标控制程序控制洗衣机进行相应的处理。
关于本实施例中的装置，其中各个模块执行操作的具体方式已经在有关该方法的实施例中进行了详细描述，此处将不做详细阐述说明。
根据本实施例提供的衣物颜色识别的处理装置，通过洗衣机上摄像机拍摄筒内的原始图像，洗衣机将原始图像数据发送给服务器，服务器接收洗衣机发送的筒内的原始图像数据；根据原始图像数据，采用预先训练好的分类网络模型，确定待洗衣物的分割图像数据；采用基于HSV空间的衣物颜色识别算法对分割图像数据进行颜色识别，获得识别结果；将识别结果发送给洗衣机，以使洗衣机根据识别结果进行相应的处理，有效提高了衣物颜色识别的准确性，从而能够为用户提供准确的洗衣服务，提高用户体验。
本申请又一实施例对上述实施例提供的装置做进一步补充说明。
作为一种可实施的方式，在上述实施例的基础上，可选地，第二确定模块，具体用于：
根据识别结果及预设控制逻辑规则，确定对应的目标控制程序，预设控制逻辑规则包括颜色与控制程序的对应关系。
作为另一种可实施的方式，在上述实施例的基础上，可选地，第二确定模块，具体用于：
若根据识别结果确定待洗衣物具有串色风险，则确定目标控制程序为告警提示程序；控制模块，具体用于：
根据告警提示程序控制洗衣机发出告警提示。
关于本实施例中的装置，其中各个模块执行操作的具体方式已经在有关该方法的实施例中进行了详细描述，此处将不做详细阐述说明。
需要说明的是，本实施例中各可实施的方式可以单独实施，也可以在不冲突的情况下以任意组合方式结合实施本申请不做限定。
根据本实施例提供的衣物颜色识别的处理装置，通过洗衣机上摄像机拍摄筒内的原始图像，洗衣机将原始图像数据发送给服务器，服务器接收洗衣机发送的筒内的原始图像数据；根据原始图像数据，采用预先训练好的分类网络模型，确定待洗衣物的分割图像数据；采用基于HSV空间的衣物颜色识别算法对分割图像数据进行颜色识别，获得识别结果；将识别结果发送给洗衣机，以使洗衣机根据识别结果进行相应的处理，有效提高了衣物颜色识别的准确性，从而能够为用户提供准确的洗衣服务，提高用户体验。
本申请再一实施例提供一种服务器，用于执行上述服务器端实施例提供的方法。
参阅图8，图8为本实施例提供的服务器的结构示意图。该服务器80包括：至少一个处理器81和存储器82；
存储器用于存储计算机可执行指令，以使至少一个处理器执行计算机可执行指令实现上述实施例提供的方法。
根据本实施例的服务器，通过洗衣机上摄像机拍摄筒内的原始图像，洗衣机将原始图像数据发送给服务器，服务器接收洗衣机发送的筒内的原始图像数据；根据原始图像数据，采用预先训练好的分类网络模型，确定待洗衣物的分割图像数据；采用基于HSV空间的衣物颜色识别算法对分割图像数据进行颜色识别，获得识别结果；将识别结果发送给洗衣机，以使洗衣机根据识别结果进行相应的处理，有效提高了衣物颜色识别的准确性，从而能够为用户提供准确的洗衣服务，提高用户体验。
本申请又一实施例提供一种洗衣机，用于执行上述洗衣机端实施例提供的方法。
参阅图9，图9为本实施例提供的洗衣机的结构示意图。该洗衣机90包括：至少一个处理器91和存储器92；
存储器用于存储计算机可执行指令，以使至少一个处理器执行计算机可执行指令实现上述洗衣机端实施例提供的方法。
可选地，该洗衣机还可以包括显示界面。
可选地，该洗衣机还可以包括声音播报装置。
根据本实施例的洗衣机，通过洗衣机上摄像机拍摄筒内的原始图像，洗衣机将原始图像数据发送给服务器，服务器接收洗衣机发送的筒内的原始图像数据；根据原始图像数据，采用预先训练好的分类网络模型，确定待洗衣物的分割图像数据；采用基于HSV空间的衣物颜色识别算法对分割图像数据进行颜色识别，获得识别结果；将识别结果发送给洗衣机，以使洗衣机根据识别结果进行相应的处理，有效提高了衣物颜色识别的准确性，从而能够为用户提供准确的洗衣服务，提高用户体验。
本申请再一实施例提供一种计算机可读存储介质，计算机可读存储介质中存储有计算机执行指令，计算机执行指令被处理器执行时用于实现上述任一实施例提供的服务器端执行的方法。
根据本实施例的计算机可读存储介质，通过洗衣机上摄像机拍摄筒内的原始图像，洗衣机将原始图像数据发送给服务器，服务器接收洗衣机发送的筒内的原始图像数据；根据原始图像数据，采用预先训练好的分类网络模型，确定待洗衣物的分割图像数据；采用基于HSV空间的衣物颜色识别算法对分割图像数据进行颜色识别，获得识别结果；将识别结果发送给洗衣机，以使洗衣机根据识别结果进行相应的处理，有效提高了衣物颜色识别的准确性，从而能够为用户提供准确的洗衣服务，提高用户体验。
本申请又一实施例提供一种计算机可读存储介质，计算机可读存储介质中存储有计算机执行指令，计算机执行指令被处理器执行时用于实现上述任一实施例提供洗衣机端执行的方法。
根据本实施例的计算机可读存储介质，通过洗衣机上摄像机拍摄筒内的原始图像，洗衣机将原始图像数据发送给服务器，服务器接收洗衣机发送的筒内的原始图像数据；根据原始图像数据，采用预先训练好的分类网络模型，确定待洗衣物的分割图像数据； 采用基于HSV空间的衣物颜色识别算法对分割图像数据进行颜色识别，获得识别结果；将识别结果发送给洗衣机，以使洗衣机根据识别结果进行相应的处理，有效提高了衣物颜色识别的准确性，从而能够为用户提供准确的洗衣服务，提高用户体验。
本申请另一种实施例提供的衣物颜色识别的处理方法，适用于在用户洗衣时，自动识别衣物颜色，并根据衣物颜色选择合适的洗衣程序进行洗涤的应用场景。参阅图10，图10为本申请下述实施例涉及终端、洗衣机和服务器交互的的处理系统的架构示意图。如图10所示，该处理系统可以包括终端、洗衣机和服务器，服务器可以是云服务器，也可以是其他服务器。终端上可以安装有洗衣机相应的客户端应用程序APP或者终端上设置相应的衣物颜色识别的处理装置，洗衣机可以是滚筒洗衣机，也可以是其他类型的洗衣机，服务器上可以设置相应的衣物颜色识别的处理装置。终端可以是用户的手机、平板电脑等带有摄像头的终端，通过终端摄像头拍摄衣物场景图像(即原始图像)，还可以借助闪光灯补光，以助于拍摄到清晰可分析的图像。终端拍摄衣物场景图像(即原始图像)后，可以通过网络传输将原始图像数据发送到服务器，并在终端展示拍摄的原始图像，具体可以是在展示在轮廓输入界面，供用户勾勒原始图像中待洗衣物的轮廓，用户可以在轮廓输入界面的原始图像上勾勒出原始图像中待洗衣物的轮廓，终端可以获取用户勾勒的图像轮廓路径，并根据图像轮廓路径来确定图像轮廓信息，图像轮廓信息包括确定的待洗衣物在原始图像中的轮廓的位置信息，终端可以将图像轮廓信息发送给服务器。服务器上的衣物颜色识别的处理装置则可以获取到原始图像数据及确定的图像轮廓信息。服务器接收到原始图像数据及图像轮廓信息后，根据原始图像数据及图像轮廓信息，确定待洗衣物的分割图像数据，分割图像数据是指从原始图像数据中分割出来的仅包括待洗衣物图像部分的图像数据；采用基于HSV空间的衣物颜色识别算法对分割图像数据进行颜色识别，获得识别结果；将识别结果发送给终端及洗衣机，终端接收到识别结果后，将识别结果进行展示，洗衣机接收到识别结果后，根据识别结果及预设控制逻辑控制洗衣机进行相应的处理。通过用户勾勒衣物轮廓，实现图像背景和衣物区域的准确分割，并采用基于HSV空间的衣物颜色识别算法对分割图像数据进行颜色识别，有效提高了衣物颜色识别的准确性，从而能够为用户提供准确的洗衣服务，提高用户体验。
本申请不需要额外在洗衣机内增加任何传感器，借用用户终端集成的摄像头完成图像的采集工作，这样既没有增加洗衣机本体结构的复杂度，同时也没有提高成本，同时增加了洗衣机的智能化水平和便利程度。
下面这几个具体的实施例可以相互结合，对于相同或相似的概念或过程可能在某些实施例中不再赘述。下面将结合附图，对本发明的实施例进行描述。
本申请一实施例提供一种衣物颜色识别的处理方法，用于识别待洗衣物的颜色。本实施例的执行主体为衣物颜色识别的处理装置，该装置可以设置在服务器中。
首先参阅图11，图11为本实施例提供的衣物颜色识别的处理方法的流程示意图，该方法包括：
步骤1101，接收终端发送的拍摄的原始图像数据及待洗衣物的图像轮廓信息。
具体的，终端上可以安装有洗衣机相应的客户端应用程序APP或者终端上设置相 应的衣物颜色识别的处理装置，通过终端摄像头拍摄衣物场景图像(即原始图像)，获得原始图像数据，可以通过网络传输将原始图像数据发送到服务器，并在终端展示拍摄的原始图像，具体可以是在展示在轮廓输入界面，供用户勾勒原始图像中待洗衣物的轮廓，用户可以在轮廓输入界面的原始图像上勾勒出原始图像中待洗衣物的轮廓，终端可以获取用户勾勒的图像轮廓路径，并根据图像轮廓路径来确定图像轮廓信息，图像轮廓信息包括确定的待洗衣物在原始图像中的轮廓的位置信息，终端可以将图像轮廓信息发送给服务器。可选地，终端可以是在获得原始图像数据和图像轮廓信息后将两者一起发送给服务器，也可以是在获得原始图像数据后发送一次，在获得图像轮廓信息后再发送一次，具体可以根据实际需求设置，本实施例不做限定。
服务器上的衣物颜色识别的处理装置则可以获取到原始图像数据及确定的图像轮廓信息。
步骤1102，根据原始图像数据及图像轮廓信息，确定待洗衣物的分割图像数据。
具体的，服务器接收到原始图像数据及图像轮廓信息后，根据原始图像数据及图像轮廓信息，确定待洗衣物的分割图像数据，分割图像数据是指从原始图像数据中分割出来的仅包括待洗衣物图像部分的图像数据。
可选地，图像轮廓信息包括待洗衣物在原始图像上的轮廓的位置信息；服务器可以根据图像轮廓信息，从原始图像数据中分割出待洗衣物的分割图像数据。
示例性的，图像轮廓信息可以包括待洗衣物在原始图像中的轮廓中最上方、最左方、最有方和最下方四处的位置坐标，服务器可以根据这四个位置坐标从原始图像数据中分割出包括待洗衣物的矩形区域，作为待洗衣物的分割图像数据。
示例性的，图像轮廓信息还可以包括待洗衣物在原始图像中的轮廓中的更多的位置坐标，使得分割出的待洗衣物的分割图像数据更准确。具体可以根据实际需求设置，本实施例不做限定。
步骤1103，采用基于HSV空间的衣物颜色识别算法对分割图像数据进行颜色识别，获得识别结果。
具体的，在确定了待洗衣物的分割图像数据后，则可以采用基于HSV空间的衣物颜色识别算法对分割图像数据进行颜色识别，获得识别结果。
识别结果包括待洗衣物的颜色组成，比如红色80％白色20％。
可选地，基于HSV空间的衣物颜色识别算法为将分割图像数据从RGB空间转换到HSV空间进行颜色识别的算法。具体来说，将分割图像数据从RGB空间转换到HSV空间，获得转换后的分割图像数据；获取衣物颜色表征范围；基于转换后的分割图像数据及衣物颜色表征范围，确定待洗衣物的颜色组成，作为识别结果。其中，衣物颜色表征范围包括了不同范围代表不同的颜色，具体结合HSV六棱锥和大量颜色各异的衣服颜色表征实现来确定衣物颜色表征范围。比如结合HSV六棱锥和1000件颜色各异的衣服颜色表征实现，确定10种常用颜色的HSV表征范围。比如，黑色的表征范围为：
(h>＝0&&h<＝360)&&(s>＝0&&s<＝100)&&(v>＝0&&v<＝25)
其中，h、s、v分别代表三个通道的值。
通常获得的分割图像数据是三个RGB通道值，每个通道用0-255表示，每种颜色的确定需要三个值同时变化，无法准确判断某一颜色是哪个范围，而由RGB空间转换 到HSV空间后，H代表一种颜色的色调，基本可以确定颜色的大体属性，再通过细分S和V可细化出颜色的深浅。
在判断待洗衣物的颜色组成时，可以通过遍历分割图像数据中各像素，统计归于于每个颜色表征范围的像素比例，从而确定待洗衣物的颜色组成。
步骤1104，将识别结果发送给洗衣机，以使洗衣机根据识别结果进行相应的处理。
具体的，服务器在获得待洗衣物的颜色的识别结果后，可以将识别结果发送给洗衣机，当用户将待洗衣物放入洗衣机并启动时，洗衣机可以根据识别结果选择合适的目标控制程序，根据目标控制程序来控制洗衣机进行相应的处理。
可选地，服务器还可以将识别结果发送给终端，终端接收到识别结果可以在终端进行展示，供用户查看，使用户了解其要洗的衣物的颜色组成。
可以预先在洗衣机设置预设控制逻辑规则，可以包括颜色与控制程序的对应关系，当洗衣机接收到识别结果后，根据识别结果及预设控制逻辑规则来确定目标控制程序。
示例性的，若识别结果为红色和白色，而且颜色组成比例大于第一阈值，则确定存在串色风险，则确定目标控制程序为不启动洗衣机并告警提示的程序，则根据目标控制程序控制不启动洗衣机，并控制洗衣机发出告警提示。具体的告警提示方式可以根据实际需求设置，比如可以是通过语音播报进行告警提示，也可以在洗衣机的显示界面显示相应的告警信息并发出告警提示音等等。
示例性的，若识别结果为白色，且大于第二阈值，则确定对应的目标控制程序为护色洗程序，则可以选择护色洗程序控制洗衣机进行护色洗流程。
示例性的，若识别结果为黑色和蓝色，则确定对应的目标控制程序为混合洗程序，则可以选择混合洗程序控制洗衣机进行混合洗流程。
本实施例提供的衣物颜色识别的处理方法，通过终端拍摄衣物场景图像(即原始图像)后，可以通过网络传输将原始图像数据发送到服务器，并在终端展示拍摄的原始图像，供用户勾勒原始图像中待洗衣物的轮廓，终端可以获取用户勾勒的图像轮廓路径，并根据图像轮廓路径来确定图像轮廓信息，终端可以将图像轮廓信息发送给服务器。服务器上的衣物颜色识别的处理装置则可以获取到原始图像数据及确定的图像轮廓信息，根据原始图像数据及图像轮廓信息，确定待洗衣物的分割图像数据，采用基于HSV空间的衣物颜色识别算法对分割图像数据进行颜色识别，获得识别结果；将识别结果发送给终端及洗衣机，终端接收到识别结果后，将识别结果进行展示，洗衣机接收到识别结果后，根据识别结果及预设控制逻辑控制洗衣机进行相应的处理。通过用户勾勒衣物轮廓，实现图像背景和衣物区域的准确分割，并采用基于HSV空间的衣物颜色识别算法对分割图像数据进行颜色识别，有效提高了衣物颜色识别的准确性，从而能够为用户提供准确的洗衣服务，提高用户体验。并且，借助用户终端实现图像拍摄，不必在洗衣机上安装摄像头，降低了洗衣机成本。
本申请另一实施例对上述实施例提供的方法做进一步补充说明。
参阅图12，图12为本实施例提供的衣物颜色识别的处理方法的流程示意图。
作为一种可实施的方式，在上述实施例的基础上，可选地，图像轮廓信息包括待洗衣物在原始图像上的轮廓的位置信息；根据原始图像数据及图像轮廓信息，确定待洗衣物的分割图像数据，包括：
步骤11021，根据图像轮廓信息，从原始图像数据中分割出待洗衣物的分割图像数据。
具体的，图像轮廓信息包括待洗衣物在原始图像上的轮廓的位置信息，可以是包括预设数量的位置信息，服务器可以根据图像轮廓信息，从原始图像数据中分割出待洗衣物的分割图像数据。
示例性的，图像轮廓信息可以包括待洗衣物在原始图像中的轮廓中最上方、最左方、最有方和最下方四处的位置坐标，服务器可以根据这四个位置坐标从原始图像数据中分割出包括待洗衣物的矩形区域，作为待洗衣物的分割图像数据。
示例性的，图像轮廓信息还可以包括待洗衣物在原始图像中的轮廓中的更多的位置坐标，使得分割出的待洗衣物的分割图像数据更准确。具体可以根据实际需求设置，本实施例不做限定。
作为另一种可实施的方式，在上述实施例的基础上，可选地，采用基于HSV空间的衣物颜色识别算法对分割图像数据进行颜色识别，获得识别结果，包括：
步骤11031，将分割图像数据从RGB空间转换到HSV空间，获得转换后的分割图像数据。
步骤11032，获取衣物颜色表征范围。
步骤11033，基于转换后的分割图像数据及衣物颜色表征范围，确定待洗衣物的颜色组成。
步骤11034，将待洗衣物的颜色组成作为识别结果。
具体的，在获得待洗衣物的分割图像数据后，将分割图像数据从RGB空间转换到HSV空间，获得转换后的分割图像数据；获取衣物颜色表征范围；基于转换后的分割图像数据及衣物颜色表征范围，确定待洗衣物的颜色组成，作为识别结果。其中，衣物颜色表征范围包括了不同范围代表不同的颜色，具体可以结合HSV六棱锥和大量颜色各异的衣服颜色表征实现来确定衣物颜色表征范围。比如结合HSV六棱锥和1000件颜色各异的衣服颜色表征实现，确定10种常用颜色的HSV表征范围。比如，黑色的表征范围为：
(h>＝0&&h<＝360)&&(s>＝0&&s<＝100)&&(v>＝0&&v<＝25)
其中，h、s、v分别代表三个通道的值。
可选地，在基于转换后的分割图像数据及衣物颜色表征范围，确定待洗衣物的颜色组成之后，方法还包括：
步骤11035，对待洗衣物的颜色组成进行过滤，获得待洗衣物的主色调颜色组成。
步骤11036，将待洗衣物的主色调颜色组成作为识别结果。
具体的，在基于转换后的分割图像数据及衣物颜色表征范围，确定待洗衣物的颜色组成之后，可能获得的颜色组成包括了非常小比例的其他颜色，可以对待洗衣物的颜色组成进行过滤，获得待洗衣物的主色调颜色组成，将待洗衣物的主色调颜色组成作为识别结果发送给洗衣机
需要说明的是，本实施例中各可实施的方式可以单独实施，也可以在不冲突的情况下以任意组合方式结合实施本申请不做限定。
本实施例提供的衣物颜色识别的处理方法，通过终端拍摄衣物场景图像(即原始 图像)后，可以通过网络传输将原始图像数据发送到服务器，并在终端展示拍摄的原始图像，供用户勾勒原始图像中待洗衣物的轮廓，终端可以获取用户勾勒的图像轮廓路径，并根据图像轮廓路径来确定图像轮廓信息，终端可以将图像轮廓信息发送给服务器。服务器上的衣物颜色识别的处理装置则可以获取到原始图像数据及确定的图像轮廓信息，根据原始图像数据及图像轮廓信息，确定待洗衣物的分割图像数据，采用基于HSV空间的衣物颜色识别算法对分割图像数据进行颜色识别，获得识别结果；将识别结果发送给终端及洗衣机，终端接收到识别结果后，将识别结果进行展示，洗衣机接收到识别结果后，根据识别结果及预设控制逻辑控制洗衣机进行相应的处理。通过用户勾勒衣物轮廓，实现图像背景和衣物区域的准确分割，并采用基于HSV空间的衣物颜色识别算法对分割图像数据进行颜色识别，有效提高了衣物颜色识别的准确性，从而能够为用户提供准确的洗衣服务，提高用户体验。并且，借助用户终端实现图像拍摄，不必在洗衣机上安装摄像头，降低了洗衣机成本。
本申请再一实施例提供一种衣物颜色识别的处理方法，用于识别待洗衣物的颜色。本实施例的执行主体为衣物颜色识别的处理装置，该装置可以设置在终端中。
首先参阅图13，图13为本实施例提供的衣物颜色识别的处理方法的流程示意图，该方法包括：
步骤1301，获取拍摄的原始图像数据。
步骤1302，将原始图像数据发送给云服务器，并根据原始图像数据在轮廓输入界面展示对应的原始图像。
步骤1303，获取用户在轮廓输入界面输入的待洗衣物的图像轮廓路径。
步骤1304，根据图像轮廓路径确定图像轮廓信息。
步骤1305，将图像轮廓信息发送给云服务器。
步骤1306，接收并展示云服务器返回的待洗衣物的颜色的识别结果。
需要说明的是，本实施例是与上述服务器端实施例交互的终端侧的方法实施例，其具体执行过程已在上述实施例的交互描述中进行了详细说明，上述实施例涉及到的终端侧执行的部分均可以用于解释本实施例的上述步骤，在此不再赘述。
本实施例提供的衣物颜色识别的处理方法，通过终端拍摄衣物场景图像(即原始图像)后，可以通过网络传输将原始图像数据发送到服务器，并在终端展示拍摄的原始图像，供用户勾勒原始图像中待洗衣物的轮廓，终端可以获取用户勾勒的图像轮廓路径，并根据图像轮廓路径来确定图像轮廓信息，终端可以将图像轮廓信息发送给服务器。服务器上的衣物颜色识别的处理装置则可以获取到原始图像数据及确定的图像轮廓信息，根据原始图像数据及图像轮廓信息，确定待洗衣物的分割图像数据，采用基于HSV空间的衣物颜色识别算法对分割图像数据进行颜色识别，获得识别结果；将识别结果发送给终端及洗衣机，终端接收到识别结果后，将识别结果进行展示，洗衣机接收到识别结果后，根据识别结果及预设控制逻辑控制洗衣机进行相应的处理。通过用户勾勒衣物轮廓，实现图像背景和衣物区域的准确分割，并采用基于HSV空间的衣物颜色识别算法对分割图像数据进行颜色识别，有效提高了衣物颜色识别的准确性，从而能够为用户提供准确的洗衣服务，提高用户体验。并且，借助用户终端实现图像拍摄，不必在洗衣机上安装摄像头，降低了洗衣机成本。
本申请又一实施例对上述实施例提供的方法做进一步补充说明。
参阅图14，图14为本实施例提供的衣物颜色识别的处理方法的流程示意图。
作为一种可实施的方式，在上述实施例的基础上，可选地，在获取用户在轮廓输入界面输入的待洗衣物的图像轮廓路径之后，该方法还包括：
步骤12011，根据图像轮廓路径判断图像轮廓是否合格。
若图像轮廓路径为封闭路径，则确定图像轮廓合格；若图像轮廓路径为非封闭路径，则确定图像轮廓不合格。
根据图像轮廓路径确定图像轮廓信息，包括：
步骤13041，若图像轮廓合格，则根据图像轮廓路径确定图像轮廓信息。
具体的，用户可以在终端的轮廓输入界面通过滑动操作勾勒待洗衣物的轮廓，终端则可以获取用户滑动操作的路径，也即待洗衣物的图像轮廓路径。根据用户滑动操作的路径是否封闭来判断用户输入的图像轮廓是否合格。
通过引入用户交互式选取衣物轮廓的方法，并添加衣物轮廓合格校验，提高待识别区域的准确性，同时也简化了颜色识别方法的处理难度和复杂度，使颜色识别方法的部署和实施更简单。
作为另一种可实施的方式，在上述实施例的基础上，可选地，根据图像轮廓路径确定图像轮廓信息，包括：
步骤12021，从图像轮廓路径中获取预设数量的位置信息作为图像轮廓信息。
具体的，终端可以是获取用户输入的图像轮廓路径上的一定数量的位置信息作为图像轮廓信息，以降低数据处理量，提高处理速度。或者也可以将图像轮廓路径上的所有位置信息作为图像轮廓信息。具体的预设数量可以根据实际需求设置。
作为一种示例性的实施方式，参阅图15，图15为本实施例提供的工作流程示意图。具体包括：
步骤121，用户将待洗涤的衣物(即待洗衣物)放置在一定位置，打开APP客户端(即终端中安装的洗衣机相应的客户端应用程序APP)，按动颜色识别按钮，APP会主动调用手机内置摄像头进入抓图模式，用户选择抓拍一张图并显示在APP中。
步骤122、用户在APP客户端界面勾选衣物的大致轮廓，点击确认后，APP会判断轮廓是否合格。
步骤123、轮廓的合格标准是封闭与否，不合格则提示用户重新选择轮廓，合格则进行下一步。
步骤124、合格的衣物轮廓图像上传到云端(即服务器)分析并保存到数据库内。
步骤125、云端算法进行图像分析后，将识别结果按照规定逻辑作用于洗衣机，洗衣机进行相应的洗涤程序。
步骤126、云端算法的处理结果(即识别结果)同时返回给APP客户端进行显示。
需要说明的是，本实施例是与上述服务器端实施例交互的终端侧的方法实施例，其具体执行过程已在上述实施例的交互描述中进行了详细说明，上述实施例涉及到的终端侧执行的部分均可以用于解释本实施例的上述步骤，在此不再赘述。
还需要说明的是，本实施例中各可实施的方式可以单独实施，也可以在不冲突的情况下以任意组合方式结合实施本申请不做限定。
本实施例提供的衣物颜色识别的处理方法，通过终端拍摄衣物场景图像(即原始图像)后，可以通过网络传输将原始图像数据发送到服务器，并在终端展示拍摄的原始图像，供用户勾勒原始图像中待洗衣物的轮廓，终端可以获取用户勾勒的图像轮廓路径，并根据图像轮廓路径来确定图像轮廓信息，终端可以将图像轮廓信息发送给服务器。服务器上的衣物颜色识别的处理装置则可以获取到原始图像数据及确定的图像轮廓信息，根据原始图像数据及图像轮廓信息，确定待洗衣物的分割图像数据，采用基于HSV空间的衣物颜色识别算法对分割图像数据进行颜色识别，获得识别结果；将识别结果发送给终端及洗衣机，终端接收到识别结果后，将识别结果进行展示，洗衣机接收到识别结果后，根据识别结果及预设控制逻辑控制洗衣机进行相应的处理。通过用户勾勒衣物轮廓，实现图像背景和衣物区域的准确分割，并采用基于HSV空间的衣物颜色识别算法对分割图像数据进行颜色识别，有效提高了衣物颜色识别的准确性，从而能够为用户提供准确的洗衣服务，提高用户体验。并且，借助用户终端实现图像拍摄，不必在洗衣机上安装摄像头，降低了洗衣机成本。
本申请再一实施例提供一种衣物颜色识别的处理装置，用于执行上述服务器端实施例提供的方法。
参阅图16，图16为本实施例提供的衣物颜色识别的处理装置的结构示意图。该衣物颜色识别的处理装置150包括第一接收模块151、第一确定模块152、处理模块153和第一发送模块154。
其中，第一接收模块，用于接收终端发送的拍摄的原始图像数据及待洗衣物的图像轮廓信息；第一确定模块，用于根据原始图像数据及图像轮廓信息，确定待洗衣物的分割图像数据；处理模块，用于采用基于HSV空间的衣物颜色识别算法对分割图像数据进行颜色识别，获得识别结果；第一发送模块，用于将识别结果发送给洗衣机，以使洗衣机根据识别结果进行相应的处理。
可选地，第一发送模块还用于将识别结果发送给终端进行展示。
关于本实施例中的装置，其中各个模块执行操作的具体方式已经在有关该方法的实施例中进行了详细描述，此处将不做详细阐述说明。
根据本实施例提供的衣物颜色识别的处理装置，通过终端拍摄衣物场景图像(即原始图像)后，可以通过网络传输将原始图像数据发送到服务器，并在终端展示拍摄的原始图像，供用户勾勒原始图像中待洗衣物的轮廓，终端可以获取用户勾勒的图像轮廓路径，并根据图像轮廓路径来确定图像轮廓信息，终端可以将图像轮廓信息发送给服务器。服务器上的衣物颜色识别的处理装置则可以获取到原始图像数据及确定的图像轮廓信息，根据原始图像数据及图像轮廓信息，确定待洗衣物的分割图像数据，采用基于HSV空间的衣物颜色识别算法对分割图像数据进行颜色识别，获得识别结果；将识别结果发送给终端及洗衣机，终端接收到识别结果后，将识别结果进行展示，洗衣机接收到识别结果后，根据识别结果及预设控制逻辑控制洗衣机进行相应的处理。通过用户勾勒衣物轮廓，实现图像背景和衣物区域的准确分割，并采用基于HSV空间的衣物颜色识别算法对分割图像数据进行颜色识别，有效提高了衣物颜色识别的准确性，从而能够为用户提供准确的洗衣服务，提高用户体验。并且，借助用户终端实现图像拍摄，不必在洗衣机上安装摄像头，降低了洗衣机成本。
本申请又一实施例对上述实施例提供的装置做进一步补充说明。
作为一种可实施的方式，在上述实施例的基础上，可选地，图像轮廓信息包括待洗衣物在原始图像上的轮廓的位置信息；第一确定模块，具体用于：
根据图像轮廓信息，从原始图像数据中分割出待洗衣物的分割图像数据。
作为另一种可实施的方式，在上述实施例的基础上，可选地，处理模块，具体用于：
将分割图像数据从RGB空间转换到HSV空间，获得转换后的分割图像数据；
获取衣物颜色表征范围；
基于转换后的分割图像数据及衣物颜色表征范围，确定待洗衣物的颜色组成，作为识别结果。
可选地，处理模块，还用于：
对待洗衣物的颜色组成进行过滤，获得待洗衣物的主色调颜色组成；将待洗衣物的主色调颜色组成作为识别结果。
关于本实施例中的装置，其中各个模块执行操作的具体方式已经在有关该方法的实施例中进行了详细描述，此处将不做详细阐述说明。
需要说明的是，本实施例中各可实施的方式可以单独实施，也可以在不冲突的情况下以任意组合方式结合实施本申请不做限定。
根据本实施例提供的衣物颜色识别的处理装置，通过终端拍摄衣物场景图像(即原始图像)后，可以通过网络传输将原始图像数据发送到服务器，并在终端展示拍摄的原始图像，供用户勾勒原始图像中待洗衣物的轮廓，终端可以获取用户勾勒的图像轮廓路径，并根据图像轮廓路径来确定图像轮廓信息，终端可以将图像轮廓信息发送给服务器。服务器上的衣物颜色识别的处理装置则可以获取到原始图像数据及确定的图像轮廓信息，根据原始图像数据及图像轮廓信息，确定待洗衣物的分割图像数据，采用基于HSV空间的衣物颜色识别算法对分割图像数据进行颜色识别，获得识别结果；将识别结果发送给终端及洗衣机，终端接收到识别结果后，将识别结果进行展示，洗衣机接收到识别结果后，根据识别结果及预设控制逻辑控制洗衣机进行相应的处理。通过用户勾勒衣物轮廓，实现图像背景和衣物区域的准确分割，并采用基于HSV空间的衣物颜色识别算法对分割图像数据进行颜色识别，有效提高了衣物颜色识别的准确性，从而能够为用户提供准确的洗衣服务，提高用户体验。并且，借助用户终端实现图像拍摄，不必在洗衣机上安装摄像头，降低了洗衣机成本。
本申请再一实施例提供一种衣物颜色识别的处理装置，用于执行上述终端侧实施例提供的方法。
参阅图17，图17为本实施例提供的衣物颜色识别的处理装置的结构示意图。该衣物颜色识别的处理装置170包括获取模块171、第二发送模块172、展示模块173、第二确定模块174和第二接收模块175。
其中，获取模块，用于获取拍摄的原始图像数据；第二发送模块，用于将原始图像数据发送给云服务器；展示模块，用于根据原始图像数据在轮廓输入界面展示对应的原始图像；获取模块，还用于获取用户在轮廓输入界面输入的待洗衣物的图像轮廓路径；第二确定模块，用于根据图像轮廓路径确定图像轮廓信息；第二发送模块，还 用于将图像轮廓信息发送给云服务器；第二接收模块，用于云服务器返回的待洗衣物的颜色的识别结果；展示模块，还用于展示云服务器返回的待洗衣物的颜色的识别结果。
关于本实施例中的装置，其中各个模块执行操作的具体方式已经在有关该方法的实施例中进行了详细描述，此处将不做详细阐述说明。
根据本实施例提供的衣物颜色识别的处理装置，通过终端拍摄衣物场景图像(即原始图像)后，可以通过网络传输将原始图像数据发送到服务器，并在终端展示拍摄的原始图像，供用户勾勒原始图像中待洗衣物的轮廓，终端可以获取用户勾勒的图像轮廓路径，并根据图像轮廓路径来确定图像轮廓信息，终端可以将图像轮廓信息发送给服务器。服务器上的衣物颜色识别的处理装置则可以获取到原始图像数据及确定的图像轮廓信息，根据原始图像数据及图像轮廓信息，确定待洗衣物的分割图像数据，采用基于HSV空间的衣物颜色识别算法对分割图像数据进行颜色识别，获得识别结果；将识别结果发送给终端及洗衣机，终端接收到识别结果后，将识别结果进行展示，洗衣机接收到识别结果后，根据识别结果及预设控制逻辑控制洗衣机进行相应的处理。通过用户勾勒衣物轮廓，实现图像背景和衣物区域的准确分割，并采用基于HSV空间的衣物颜色识别算法对分割图像数据进行颜色识别，有效提高了衣物颜色识别的准确性，从而能够为用户提供准确的洗衣服务，提高用户体验。并且，借助用户终端实现图像拍摄，不必在洗衣机上安装摄像头，降低了洗衣机成本。
本申请又一实施例对上述实施例提供的装置做进一步补充说明。
作为一种可实施的方式，在上述实施例的基础上，可选地，第二确定模块，还用于：
根据图像轮廓路径判断图像轮廓是否合格；若图像轮廓路径为封闭路径，则确定图像轮廓合格；若图像轮廓路径为非封闭路径，则确定图像轮廓不合格；第二确定模块，具体用于：
若图像轮廓合格，则根据图像轮廓路径确定图像轮廓信息。
作为另一种可实施的方式，在上述实施例的基础上，可选地，第二确定模块，具体用于：
从图像轮廓路径中获取预设数量的位置信息作为图像轮廓信息。
关于本实施例中的装置，其中各个模块执行操作的具体方式已经在有关该方法的实施例中进行了详细描述，此处将不做详细阐述说明。
需要说明的是，本实施例中各可实施的方式可以单独实施，也可以在不冲突的情况下以任意组合方式结合实施本申请不做限定。
根据本实施例提供的衣物颜色识别的处理装置，通过终端拍摄衣物场景图像(即原始图像)后，可以通过网络传输将原始图像数据发送到服务器，并在终端展示拍摄的原始图像，供用户勾勒原始图像中待洗衣物的轮廓，终端可以获取用户勾勒的图像轮廓路径，并根据图像轮廓路径来确定图像轮廓信息，终端可以将图像轮廓信息发送给服务器。服务器上的衣物颜色识别的处理装置则可以获取到原始图像数据及确定的图像轮廓信息，根据原始图像数据及图像轮廓信息，确定待洗衣物的分割图像数据，采用基于HSV空间的衣物颜色识别算法对分割图像数据进行颜色识别，获得识别结果； 将识别结果发送给终端及洗衣机，终端接收到识别结果后，将识别结果进行展示，洗衣机接收到识别结果后，根据识别结果及预设控制逻辑控制洗衣机进行相应的处理。通过用户勾勒衣物轮廓，实现图像背景和衣物区域的准确分割，并采用基于HSV空间的衣物颜色识别算法对分割图像数据进行颜色识别，有效提高了衣物颜色识别的准确性，从而能够为用户提供准确的洗衣服务，提高用户体验。并且，借助用户终端实现图像拍摄，不必在洗衣机上安装摄像头，降低了洗衣机成本。
本申请再一实施例提供一种服务器，用于执行上述服务器侧实施例提供的方法。
参阅图18，图18为本实施例提供的服务器的结构示意图。该服务器180包括：至少一个处理器181和存储器182；
存储器用于存储计算机可执行指令，以使至少一个处理器执行计算机可执行指令实现上述实施例提供的方法。
根据本实施例的服务器，通过终端拍摄衣物场景图像(即原始图像)后，可以通过网络传输将原始图像数据发送到服务器，并在终端展示拍摄的原始图像，供用户勾勒原始图像中待洗衣物的轮廓，终端可以获取用户勾勒的图像轮廓路径，并根据图像轮廓路径来确定图像轮廓信息，终端可以将图像轮廓信息发送给服务器。服务器上的衣物颜色识别的处理装置则可以获取到原始图像数据及确定的图像轮廓信息，根据原始图像数据及图像轮廓信息，确定待洗衣物的分割图像数据，采用基于HSV空间的衣物颜色识别算法对分割图像数据进行颜色识别，获得识别结果；将识别结果发送给终端及洗衣机，终端接收到识别结果后，将识别结果进行展示，洗衣机接收到识别结果后，根据识别结果及预设控制逻辑控制洗衣机进行相应的处理。通过用户勾勒衣物轮廓，实现图像背景和衣物区域的准确分割，并采用基于HSV空间的衣物颜色识别算法对分割图像数据进行颜色识别，有效提高了衣物颜色识别的准确性，从而能够为用户提供准确的洗衣服务，提高用户体验。并且，借助用户终端实现图像拍摄，不必在洗衣机上安装摄像头，降低了洗衣机成本。
本申请又一实施例提供一种终端，用于执行上述终端侧实施例提供的方法。
参阅图19，图19为本实施例提供的终端的结构示意图。该终端190包括：至少一个处理器191和存储器192；
存储器用于存储计算机可执行指令，以使至少一个处理器执行计算机可执行指令实现上述洗衣机端实施例提供的方法。
根据本实施例的终端，通过终端拍摄衣物场景图像(即原始图像)后，可以通过网络传输将原始图像数据发送到服务器，并在终端展示拍摄的原始图像，供用户勾勒原始图像中待洗衣物的轮廓，终端可以获取用户勾勒的图像轮廓路径，并根据图像轮廓路径来确定图像轮廓信息，终端可以将图像轮廓信息发送给服务器。服务器上的衣物颜色识别的处理装置则可以获取到原始图像数据及确定的图像轮廓信息，根据原始图像数据及图像轮廓信息，确定待洗衣物的分割图像数据，采用基于HSV空间的衣物颜色识别算法对分割图像数据进行颜色识别，获得识别结果；将识别结果发送给终端及洗衣机，终端接收到识别结果后，将识别结果进行展示，洗衣机接收到识别结果后，根据识别结果及预设控制逻辑控制洗衣机进行相应的处理。通过用户勾勒衣物轮廓，实现图像背景和衣物区域的准确分割，并采用基于HSV空间的衣物颜色识别算法对分 割图像数据进行颜色识别，有效提高了衣物颜色识别的准确性，从而能够为用户提供准确的洗衣服务，提高用户体验。并且，借助用户终端实现图像拍摄，不必在洗衣机上安装摄像头，降低了洗衣机成本。
本申请再一实施例提供一种计算机可读存储介质，计算机可读存储介质中存储有计算机执行指令，计算机执行指令被处理器执行时用于实现上述任一实施例提供的服务器端执行的方法。
根据本实施例的计算机可读存储介质，通过终端摄像头拍摄衣物场景图像(即原始图像)，还可以借助闪光灯补光，以助于拍摄到清晰可分析的图像。终端拍摄衣物场景图像(即原始图像)后，可以通过网络传输将原始图像数据发送到服务器，并在终端展示拍摄的原始图像，具体可以是在展示在轮廓输入界面，供用户勾勒原始图像中待洗衣物的轮廓，用户可以在轮廓输入界面的原始图像上勾勒出原始图像中待洗衣物的轮廓，终端可以获取用户勾勒的图像轮廓路径，并根据图像轮廓路径来确定图像轮廓信息，图像轮廓信息包括确定的待洗衣物在原始图像中的轮廓的位置信息，终端可以将图像轮廓信息发送给服务器。服务器上的衣物颜色识别的处理装置则可以获取到原始图像数据及确定的图像轮廓信息。服务器接收到原始图像数据及图像轮廓信息后，原始图像数据及图像轮廓信息，确定待洗衣物的分割图像数据，分割图像数据是指从原始图像数据中分割出来的仅包括待洗衣物图像部分的图像数据；采用基于HSV空间的衣物颜色识别算法对分割图像数据进行颜色识别，获得识别结果；将识别结果发送给终端及洗衣机，终端接收到识别结果后，将识别结果进行展示，洗衣机接收到识别结果后，根据识别结果及预设控制逻辑控制洗衣机进行相应的处理。通过用户勾勒衣物轮廓，实现图像背景和衣物区域的准确分割，并采用基于HSV空间的衣物颜色识别算法对分割图像数据进行颜色识别，有效提高了衣物颜色识别的准确性，从而能够为用户提供准确的洗衣服务，提高用户体验。并且，借助用户终端实现图像拍摄，不必在洗衣机上安装摄像头，降低了洗衣机成本。
本申请又一实施例提供一种计算机可读存储介质，计算机可读存储介质中存储有计算机执行指令，计算机执行指令被处理器执行时用于实现上述任一实施例提供洗衣机端执行的方法。
根据本实施例的计算机可读存储介质，通过终端摄像头拍摄衣物场景图像(即原始图像)，还可以借助闪光灯补光，以助于拍摄到清晰可分析的图像。终端拍摄衣物场景图像(即原始图像)后，可以通过网络传输将原始图像数据发送到服务器，并在终端展示拍摄的原始图像，具体可以是在展示在轮廓输入界面，供用户勾勒原始图像中待洗衣物的轮廓，用户可以在轮廓输入界面的原始图像上勾勒出原始图像中待洗衣物的轮廓，终端可以获取用户勾勒的图像轮廓路径，并根据图像轮廓路径来确定图像轮廓信息，图像轮廓信息包括确定的待洗衣物在原始图像中的轮廓的位置信息，终端可以将图像轮廓信息发送给服务器。服务器上的衣物颜色识别的处理装置则可以获取到原始图像数据及确定的图像轮廓信息。服务器接收到原始图像数据及图像轮廓信息后，原始图像数据及图像轮廓信息，确定待洗衣物的分割图像数据，分割图像数据是指从原始图像数据中分割出来的仅包括待洗衣物图像部分的图像数据；采用基于HSV空间的衣物颜色识别算法对分割图像数据进行颜色识别，获得识别结果；将识别结果 发送给终端及洗衣机，终端接收到识别结果后，将识别结果进行展示，洗衣机接收到识别结果后，根据识别结果及预设控制逻辑控制洗衣机进行相应的处理。通过用户勾勒衣物轮廓，实现图像背景和衣物区域的准确分割，并采用基于HSV空间的衣物颜色识别算法对分割图像数据进行颜色识别，有效提高了衣物颜色识别的准确性，从而能够为用户提供准确的洗衣服务，提高用户体验。并且，借助用户终端实现图像拍摄，不必在洗衣机上安装摄像头，降低了洗衣机成本。
在本申请所提供的几个实施例中，应该理解到，所揭露的装置和方法，可以通过其它的方式实现。例如，以上所描述的装置实施例仅仅是示意性的，例如，所述单元的划分，仅仅为一种逻辑功能划分，实际实现时可以有另外的划分方式，例如多个单元或组件可以结合或者可以集成到另一个系统，或一些特征可以忽略，或不执行。另一点，所显示或讨论的相互之间的耦合或直接耦合或通信连接可以是通过一些接口，装置或单元的间接耦合或通信连接，可以是电性，机械或其它的形式。
所述作为分离部件说明的单元可以是或者也可以不是物理上分开的，作为单元显示的部件可以是或者也可以不是物理单元，即可以位于一个地方，或者也可以分布到多个网络单元上。可以根据实际的需要选择其中的部分或者全部单元来实现本实施例方案的目的。
另外，在本申请各个实施例中的各功能单元可以集成在一个处理单元中，也可以是各个单元单独物理存在，也可以两个或两个以上单元集成在一个单元中。上述集成的单元既可以采用硬件的形式实现，也可以采用硬件加软件功能单元的形式实现。
上述以软件功能单元的形式实现的集成的单元，可以存储在一个计算机可读取存储介质中。上述软件功能单元存储在一个存储介质中，包括若干指令用以使得一台计算机设备(可以是个人计算机，服务器，或者网络设备等)或处理器(processor)执行本申请各个实施例所述方法的部分步骤。而前述的存储介质包括：U盘、移动硬盘、只读存储器(Read-Only Memory，ROM)、随机存取存储器(Random Access Memory，RAM)、磁碟或者光盘等各种可以存储程序代码的介质。
本领域技术人员可以清楚地了解到，为描述的方便和简洁，仅以上述各功能模块的划分进行举例说明，实际应用中，可以根据需要而将上述功能分配由不同的功能模块完成，即将装置的内部结构划分成不同的功能模块，以完成以上描述的全部或者部分功能。上述描述的装置的具体工作过程，可以参考前述方法实施例中的对应过程，在此不再赘述。
最后应说明的是：以上各实施例仅用以说明本申请的技术方案，而非对其限制；尽管参照前述各实施例对本申请进行了详细的说明，本领域的普通技术人员应当理解：其依然可以对前述各实施例所记载的技术方案进行修改，或者对其中部分或者全部技术特征进行等同替换；而这些修改或者替换，并不使相应技术方案的本质脱离本申请各实施例技术方案的范围。

Claims (30)

  1. 一种衣物颜色识别的处理方法，其特征在于，包括：
    接收洗衣机发送的筒内的原始图像数据；
    根据所述原始图像数据，采用预先训练好的分类网络模型，确定待洗衣物的分割图像数据；
    采用基于HSV空间的衣物颜色识别算法对所述分割图像数据进行颜色识别，获得识别结果；
    将所述识别结果发送给洗衣机，以使所述洗衣机根据所述识别结果进行相应的处理。
  2. 根据权利要求1所述的方法，其特征在于，所述根据所述原始图像数据，采用预先训练好的分类网络模型，确定待洗衣物的分割图像数据，包括：
    根据所述原始图像数据，采用预先训练好的基于深度学习的二分类神经网络模型，确定待洗衣物的分割图像数据。
  3. 根据权利要求1或2所述的方法，其特征在于，所述采用基于HSV空间的衣物颜色识别算法对所述分割图像数据进行颜色识别，获得识别结果，包括：
    将所述分割图像数据从RGB空间转换到HSV空间，获得转换后的分割图像数据；
    获取衣物颜色表征范围；
    基于转换后的分割图像数据及所述衣物颜色表征范围，确定待洗衣物的颜色组成，作为所述识别结果。
  4. 根据权利要求3所述的方法，其特征在于，在基于转换后的分割图像数据及预先获得的衣物颜色表征范围，确定待洗衣物的颜色组成之后，所述方法还包括：
    对所述待洗衣物的颜色组成进行过滤，获得所述待洗衣物的主色调颜色组成；
    将所述待洗衣物的主色调颜色组成作为所述识别结果。
  5. 一种衣物颜色识别的处理方法，其特征在于，包括：
    采集洗衣机筒内的原始图像数据；
    将所述原始图像数据发送给服务器；
    接收所述服务器返回的待洗衣物的颜色的识别结果；
    根据所述识别结果，确定对应的目标控制程序；
    根据所述目标控制程序控制所述洗衣机进行相应的处理。
  6. 根据权利要求5所述的方法，其特征在于，所述根据所述识别结果，确定对应的目标控制程序，包括：
    根据所述识别结果及预设控制逻辑规则，确定对应的目标控制程序，所述预设控制逻辑规则包括颜色与控制程序的对应关系。
  7. 根据权利要求6所述的方法，其特征在于，所述根据所述识别结果及预设控制逻辑规则，确定对应的目标控制程序，包括：
    若根据所述识别结果确定待洗衣物具有串色风险，则确定目标控制程序为告警提示程序；
    所述根据所述目标控制程序控制所述洗衣机进行相应的处理，包括：
    根据所述告警提示程序控制所述洗衣机发出告警提示。
  8. 一种衣物颜色识别的处理装置，其特征在于，包括：
    第一接收模块，用于接收洗衣机发送的筒内的原始图像数据；
    第一确定模块，用于根据所述原始图像数据，采用预先训练好的分类网络模型，确定待洗衣物的分割图像数据；
    处理模块，用于采用基于HSV空间的衣物颜色识别算法对所述分割图像数据进行颜色识别，获得识别结果；
    第一发送模块，用于将所述识别结果发送给洗衣机，以使所述洗衣机根据所述识别结果进行相应的处理。
  9. 一种衣物颜色识别的处理装置，包括：
    获取模块，用于采集洗衣机筒内的原始图像数据；
    第二发送模块，用于将所述原始图像数据发送给服务器；
    第二接收模块，用于接收所述服务器返回的待洗衣物的颜色的识别结果；
    第二确定模块，用于根据所述识别结果，确定对应的目标控制程序；
    控制模块，用于根据所述目标控制程序控制所述洗衣机进行相应的处理。
  10. 一种服务器，其特征在于，包括：至少一个处理器和存储器；
    所述存储器用于存储计算机可执行指令，以使所述至少一个处理器执行所述计算机可执行指令实现权利要求1-4中任一项所述的方法。
  11. 一种洗衣机，其特征在于，包括：至少一个处理器和存储器；
    所述存储器用于存储计算机可执行指令，以使所述至少一个处理器执行所述计算机可执行指令实现权利要求5-7中任一项所述的方法。
  12. 一种计算机可读介质，其特征在于，其上存储有计算机程序，其中，所述程序被处理器执行时实现如权利要求1-4中任一所述的方法。
  13. 一种计算机可读介质，其特征在于，其上存储有计算机程序，其中，所述程序被处理器执行时实现如权利要求5-7中任一所述的方法。
  14. 一种计算机程序，其特征在于，包括程序代码，当计算机运行所述计算机程序时，所述程序代码执行如权利要求1-4中任一所述的方法。
  15. 一种计算机程序，其特征在于，包括程序代码，当计算机运行所述计算机程序时，所述程序代码执行如权利要求5-7中任一所述的方法。
  16. 一种衣物颜色识别的处理方法，其特征在于，包括：
    接收终端发送的拍摄的原始图像数据及待洗衣物的图像轮廓信息；
    根据所述原始图像数据及所述图像轮廓信息，确定待洗衣物的分割图像数据；
    采用基于HSV空间的衣物颜色识别算法对所述分割图像数据进行颜色识别，获得识别结果；
    将所述识别结果发送给洗衣机，以使所述洗衣机根据所述识别结果进行相应的处理。
  17. 根据权利要求16所述的方法，其特征在于，所述图像轮廓信息包括待洗衣物在原始图像上的轮廓的位置信息；
    所述根据所述原始图像数据及所述图像轮廓信息，确定待洗衣物的分割图像数据，包括：
    根据所述图像轮廓信息，从所述原始图像数据中分割出所述待洗衣物的分割图像数据。
  18. 根据权利要求16或17所述的方法，其特征在于，所述采用基于HSV空间的衣物颜色识别算法对所述分割图像数据进行颜色识别，获得识别结果，包括：
    将所述分割图像数据从RGB空间转换到HSV空间，获得转换后的分割图像数据；
    获取衣物颜色表征范围；
    基于转换后的分割图像数据及所述衣物颜色表征范围，确定待洗衣物的颜色组成，作为所述识别结果。
  19. 根据权利要求18所述的方法，其特征在于，在基于转换后的分割图像数据及所述衣物颜色表征范围，确定待洗衣物的颜色组成之后，所述方法还包括：
    对所述待洗衣物的颜色组成进行过滤，获得所述待洗衣物的主色调颜色组成；
    将所述待洗衣物的主色调颜色组成作为所述识别结果。
  20. 一种衣物颜色识别的处理方法，其特征在于，包括：
    获取拍摄的原始图像数据；
    将所述原始图像数据发送给云服务器，并根据所述原始图像数据在轮廓输入界面展示对应的原始图像；
    获取用户在所述轮廓输入界面输入的待洗衣物的图像轮廓路径；
    根据所述图像轮廓路径确定图像轮廓信息；
    将所述图像轮廓信息发送给所述云服务器；
    接收并展示所述云服务器返回的待洗衣物的颜色的识别结果。
  21. 根据权利要求20所述的方法，其特征在于，在所述获取用户在所述轮廓输入界面输入的待洗衣物的图像轮廓路径之后，所述方法还包括：
    根据所述图像轮廓路径判断图像轮廓是否合格；
    若所述图像轮廓路径为封闭路径，则确定图像轮廓合格；
    若所述图像轮廓路径为非封闭路径，则确定图像轮廓不合格；
    所述根据所述图像轮廓路径确定图像轮廓信息，包括：
    若图像轮廓合格，则根据所述图像轮廓路径确定图像轮廓信息。
  22. 根据权利要求20或21所述的方法，其特征在于，所述根据所述图像轮廓路径确定图像轮廓信息，包括：
    从所述图像轮廓路径中获取预设数量的位置信息作为所述图像轮廓信息。
  23. 一种衣物颜色识别的处理装置，其特征在于，包括：
    第一接收模块，用于接收终端发送的拍摄的原始图像数据及待洗衣物的图像轮廓信息；
    第一确定模块，用于根据所述原始图像数据及所述图像轮廓信息，确定待洗衣物的分割图像数据；
    处理模块，用于采用基于HSV空间的衣物颜色识别算法对所述分割图像数据进行颜色识别，获得识别结果；
    第一发送模块，用于将所述识别结果发送给洗衣机，以使所述洗衣机根据所述识别结果进行相应的处理。
  24. 一种衣物颜色识别的处理装置，其特征在于，包括：
    获取模块，用于获取拍摄的原始图像数据；
    第二发送模块，用于将所述原始图像数据发送给云服务器；
    展示模块，用于根据所述原始图像数据在轮廓输入界面展示对应的原始图像；
    所述获取模块，还用于获取用户在所述轮廓输入界面输入的待洗衣物的图像轮廓路径；
    第二确定模块，用于根据所述图像轮廓路径确定图像轮廓信息；
    所述第二发送模块，还用于将所述图像轮廓信息发送给所述云服务器；
    第二接收模块，用于所述云服务器返回的待洗衣物的颜色的识别结果；
    所述展示模块，还用于展示所述云服务器返回的待洗衣物的颜色的识别结果。
  25. 一种服务器，其特征在于，包括：至少一个处理器和存储器；
    所述存储器存储计算机程序；所述至少一个处理器执行所述存储器存储的计算机程序，以实现权利要求16-19中任一项所述的方法。
  26. 一种终端，其特征在于，包括：至少一个处理器和存储器；
    所述存储器存储计算机程序；所述至少一个处理器执行所述存储器存储的计算机程序，以实现权利要求20-22中任一项所述的方法。
  27. 一种计算机可读介质，其特征在于，其上存储有计算机程序，其中，所述程序被处理器执行时实现如权利要求16-19中任一所述的方法。
  28. 一种计算机可读介质，其特征在于，其上存储有计算机程序，其中，所述程序被处理器执行时实现如权利要求20-22中任一所述的方法。
  29. 一种计算机程序，其特征在于，包括程序代码，当计算机运行所述计算机程序时，所述程序代码执行如权利要求16-19中任一所述的方法。
  30. 一种计算机程序，其特征在于，包括程序代码，当计算机运行所述计算机程序时，所述程序代码执行如权利要求20-22中任一所述的方法。
PCT/CN2020/125427 2019-11-22 2020-10-30 衣物颜色识别的处理方法、装置、设备及存储介质 WO2021098486A1 (zh)

Applications Claiming Priority (4)

Application Number Priority Date Filing Date Title
CN201911152484.0A CN112836706A (zh) 2019-11-22 2019-11-22 衣物颜色识别的处理方法、装置、设备及存储介质
CN201911152480.2A CN112831982A (zh) 2019-11-22 2019-11-22 衣物颜色识别的处理方法、装置、设备及存储介质
CN201911152480.2 2019-11-22
CN201911152484.0 2019-11-22

Publications (1)

Publication Number Publication Date
WO2021098486A1 true WO2021098486A1 (zh) 2021-05-27

Family

ID=75981238

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2020/125427 WO2021098486A1 (zh) 2019-11-22 2020-10-30 衣物颜色识别的处理方法、装置、设备及存储介质

Country Status (1)

Country Link
WO (1) WO2021098486A1 (zh)

Cited By (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113361201A (zh) * 2021-06-10 2021-09-07 南京大学 一种基于噪声标签学习的众包获取标签数据清洗方法
CN113848736A (zh) * 2021-09-13 2021-12-28 青岛海尔科技有限公司 基于智能衣柜的衣物信息处理方法及设备
CN113876988A (zh) * 2021-09-18 2022-01-04 青岛海尔空调器有限总公司 用于衣物消毒的方法及装置、空调设备
CN114638026A (zh) * 2022-02-17 2022-06-17 广东时谛智能科技有限公司 鞋体设计过程中的区块划分方法和装置
CN115861451A (zh) * 2022-12-27 2023-03-28 东莞市楷德精密机械有限公司 一种基于机器视觉的多功能图像处理方法及系统

Citations (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20070299545A1 (en) * 2006-06-22 2007-12-27 Kavita Agrawal Control of washing machines using color sensors
CN102207966A (zh) * 2011-06-01 2011-10-05 华南理工大学 基于对象标签的视频内容快速检索方法
CN105839332A (zh) * 2016-05-19 2016-08-10 无锡小天鹅股份有限公司 波轮洗衣机
CN108950983A (zh) * 2017-05-27 2018-12-07 青岛海尔洗衣机有限公司 智能检测模块、洗衣机和洗衣机的洗涤程序推荐方法
CN109137388A (zh) * 2018-10-26 2019-01-04 无锡小天鹅股份有限公司 衣物洗涤方法、装置和衣物处理装置
CN109750450A (zh) * 2017-11-01 2019-05-14 青岛海尔智能技术研发有限公司 一种识别衣物材质的智能模块及智能洗衣机
CN110443137A (zh) * 2019-07-03 2019-11-12 平安科技（深圳）有限公司 多维度身份信息识别方法、装置、计算机设备及存储介质
CN110438746A (zh) * 2019-05-10 2019-11-12 青岛海尔智能技术研发有限公司 材质识别装置、方法、洗衣机及其控制方法

Patent Citations (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20070299545A1 (en) * 2006-06-22 2007-12-27 Kavita Agrawal Control of washing machines using color sensors
CN102207966A (zh) * 2011-06-01 2011-10-05 华南理工大学 基于对象标签的视频内容快速检索方法
CN105839332A (zh) * 2016-05-19 2016-08-10 无锡小天鹅股份有限公司 波轮洗衣机
CN108950983A (zh) * 2017-05-27 2018-12-07 青岛海尔洗衣机有限公司 智能检测模块、洗衣机和洗衣机的洗涤程序推荐方法
CN109750450A (zh) * 2017-11-01 2019-05-14 青岛海尔智能技术研发有限公司 一种识别衣物材质的智能模块及智能洗衣机
CN109137388A (zh) * 2018-10-26 2019-01-04 无锡小天鹅股份有限公司 衣物洗涤方法、装置和衣物处理装置
CN110438746A (zh) * 2019-05-10 2019-11-12 青岛海尔智能技术研发有限公司 材质识别装置、方法、洗衣机及其控制方法
CN110443137A (zh) * 2019-07-03 2019-11-12 平安科技（深圳）有限公司 多维度身份信息识别方法、装置、计算机设备及存储介质

Cited By (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113361201A (zh) * 2021-06-10 2021-09-07 南京大学 一种基于噪声标签学习的众包获取标签数据清洗方法
CN113361201B (zh) * 2021-06-10 2023-08-25 南京大学 一种基于噪声标签学习的众包获取标签数据清洗方法
CN113848736A (zh) * 2021-09-13 2021-12-28 青岛海尔科技有限公司 基于智能衣柜的衣物信息处理方法及设备
CN113876988A (zh) * 2021-09-18 2022-01-04 青岛海尔空调器有限总公司 用于衣物消毒的方法及装置、空调设备
CN113876988B (zh) * 2021-09-18 2023-06-23 青岛海尔空调器有限总公司 用于衣物消毒的方法及装置、空调设备
CN114638026A (zh) * 2022-02-17 2022-06-17 广东时谛智能科技有限公司 鞋体设计过程中的区块划分方法和装置
CN115861451A (zh) * 2022-12-27 2023-03-28 东莞市楷德精密机械有限公司 一种基于机器视觉的多功能图像处理方法及系统
CN115861451B (zh) * 2022-12-27 2023-06-30 东莞市楷德精密机械有限公司 一种基于机器视觉的多功能图像处理方法及系统

Similar Documents

Publication Publication Date Title
WO2021098486A1 (zh) 衣物颜色识别的处理方法、装置、设备及存储介质
KR102195922B1 (ko) 인터넷에 기반하는 얼굴 미화 시스템
US20220237811A1 (en) Method for Testing Skin Texture, Method for Classifying Skin Texture and Device for Testing Skin Texture
US9864901B2 (en) Feature detection and masking in images based on color distributions
CN110163810B (zh) 一种图像处理方法、装置以及终端
CN104636726B (zh) 一种图像色彩识别方法、装置及终端
CN105359162B (zh) 用于图像中的与脸部有关的选择和处理的图像掩模
CN110263681A (zh) 面部表情的识别方法和装置、存储介质、电子装置
CN110335312A (zh) 一种基于神经网络的目标空间定位方法及装置
CN109871845A (zh) 证件图像提取方法及终端设备
WO2010078586A1 (en) Detecting skin tone in images
CN111275645A (zh) 基于人工智能的图像去雾方法、装置、设备及存储介质
CN106886992A (zh) 一种基于饱和度的彩色多曝光融合图像的质量评价方法
CN110298893A (zh) 一种行人衣着颜色识别模型的生成方法及装置
CN113781408A (zh) 一种图像拍摄智能指导系统及方法
CN115100240A (zh) 视频中的物体追踪方法及装置、电子设备、存储介质
CN106097261A (zh) 图像处理方法和装置
CN111832464A (zh) 一种基于近红外摄像头的活体检测方法及装置
CN112831982A (zh) 衣物颜色识别的处理方法、装置、设备及存储介质
CN112836706A (zh) 衣物颜色识别的处理方法、装置、设备及存储介质
CN114257730A (zh) 图像数据的处理方法、装置、存储介质及计算机设备
CN106408617A (zh) 一种基于yuv颜色空间的交互式单幅图像材质获取系统和方法
EP4040347A1 (en) Device and method for learning data augmentation-based space analysis model
CN115760879A (zh) 图像处理方法、图像处理系统、装置、设备及介质
CN113837018A (zh) 一种化妆进度检测方法、装置、设备及存储介质

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 20889316

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 20889316

Country of ref document: EP

Kind code of ref document: A1