WO2021017308A1 - 舌象匹配方法、电子装置、计算机设备及存储介质 - Google Patents

舌象匹配方法、电子装置、计算机设备及存储介质 Download PDF

Info

Publication number
WO2021017308A1
WO2021017308A1 PCT/CN2019/118093 CN2019118093W WO2021017308A1 WO 2021017308 A1 WO2021017308 A1 WO 2021017308A1 CN 2019118093 W CN2019118093 W CN 2019118093W WO 2021017308 A1 WO2021017308 A1 WO 2021017308A1
Authority
WO
WIPO (PCT)
Prior art keywords
tongue
tested
image
tongue image
color histogram
Prior art date
Legal status (The legal status 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 status listed.)
Ceased
Application number
PCT/CN2019/118093
Other languages
English (en)
French (fr)
Inventor
曹靖康
王义文
王健宗
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Ping An Technology Shenzhen Co Ltd
Original Assignee
Ping An Technology Shenzhen Co Ltd
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
Application filed by Ping An Technology Shenzhen Co Ltd filed Critical Ping An Technology Shenzhen Co Ltd
Publication of WO2021017308A1 publication Critical patent/WO2021017308A1/zh
Anticipated expiration legal-status Critical
Ceased legal-status Critical Current

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
    • G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
    • G06V40/16—Human faces, e.g. facial parts, sketches or expressions
    • G06V40/168—Feature extraction; Face representation
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
    • G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
    • G06V40/16—Human faces, e.g. facial parts, sketches or expressions
    • G06V40/172—Classification, e.g. identification

Definitions

  • This application relates to the field of image processing technology, and in particular to a tongue image matching method, electronic device, computer equipment and storage medium.
  • tongue is the seedling of the heart
  • Chinese medicine refers to the meaning of understanding the physical condition by observing the tongue.
  • the doctor usually checks the tongue.
  • the tongue is a barometer of health.
  • Traditional Chinese medicine generally divides the tongue into four parts, including the tip of the tongue, the middle of the tongue, the base of the tongue, and the sides of the tongue.
  • the tip of the tongue belongs to the heart and lungs
  • the edge of the tongue belongs to the liver and gallbladder
  • the spleen and stomach in the tongue and the root of the tongue belongs to the kidneys. Reflecting this, referring to the tongue coating can more comprehensively judge the biological attributes of a person, and the difference in tongue color represents the different biological attributes of the human body.
  • this application proposes a tongue image matching method, electronic device, computer equipment, and storage medium, which can effectively improve the convenience and accuracy of the user's knowledge of his physical condition through the tongue image.
  • this application proposes a tongue image matching method, which includes the steps of: establishing a database, which contains several reference tongue images; performing color feature extraction on the reference tongue images to generate a reference color histogram Figure; add biological attribute information corresponding to the reference tongue image; obtain the face image to be tested; extract the tongue image to be tested from the face image to be tested; perform color histogram smoothing processing on the tongue image to be tested ; Perform format conversion on the tongue image to be tested after the color histogram smoothing process to generate the tongue image in HSV format; match the reference tongue image closest to the tongue image to be tested from the database, based on the reference tongue
  • the biological attribute corresponding to the image describes the tongue image output result of the tongue image to be tested.
  • the present application also provides an electronic device, which includes: a database building module, suitable for building a database, the database contains a number of reference tongue images, the color feature extraction of the reference tongue image, generating Reference color histogram, and adding biological attribute information corresponding to the reference tongue image; acquisition module, suitable for acquiring the face image to be tested; extraction module, suitable for extracting the tongue image to be tested from the face image to be tested
  • the processing module is adapted to perform color histogram smoothing processing on the tongue image to be tested; the conversion module; the conversion module is adapted to perform format conversion on the tongue image to be tested after the color histogram smoothing processing is performed to generate an HSV format
  • the tongue image to be tested a matching module, adapted to match the reference tongue image closest to the tongue image to be tested from a database, and generate the tongue image of the tongue image to be tested based on the biological attribute description corresponding to the reference tongue image Output the result.
  • this application also provides a computer device, including a memory, a processor, and computer-readable instructions stored in the memory and running on the processor, and the processor executes the computer-readable instructions When implementing the steps of the above method.
  • the present application also provides a non-volatile computer-readable storage medium on which computer-readable instructions are stored, and the computer-readable instructions implement the steps of the foregoing method when executed by a processor.
  • the tongue image matching method, electronic device, computer equipment and storage medium proposed in this application can extract the tongue image to be tested from the face image to be tested, and match the reference tongue image that is closest to the tongue image to be tested.
  • the biological attribute description corresponding to the reference tongue image generates the tongue image output result of the tongue image to be tested, which can give the tongue image output result intuitively and conveniently, and effectively improve the convenience and accuracy of the user to learn his physical condition through the tongue image.
  • Fig. 1 is a schematic flowchart of a tongue image matching method according to an exemplary embodiment of the present application
  • Fig. 2 is a schematic diagram showing the tongue part corresponding to the five internal organs according to an exemplary embodiment of the present application
  • FIG. 3 is a schematic flowchart of a tongue image matching method according to an exemplary embodiment of the present application.
  • FIG. 4 is a schematic flowchart of a tongue image matching method shown in an exemplary embodiment of the present application.
  • FIG. 5 is a schematic flowchart of a tongue image matching method according to an exemplary embodiment of the present application.
  • Fig. 6 is a schematic flowchart of a tongue image matching method according to an exemplary embodiment of the present application.
  • FIG. 7 is a schematic flowchart of a tongue image matching method according to an exemplary embodiment of the present application.
  • FIG. 8 is a schematic diagram of program modules of an electronic device shown in an exemplary embodiment of the present application.
  • FIG. 9 is a schematic diagram of a hardware architecture of an electronic device shown in an exemplary embodiment of the present application.
  • FIG. 1 is a schematic flowchart of a tongue image matching method according to an embodiment of the present application. The method includes the following steps:
  • Step S110 acquiring a face image to be tested
  • Step S120 extract a tongue image to be tested from the face image to be tested.
  • Step S130 matching the reference tongue image closest to the tongue image to be tested from the database, and generating a tongue image output result of the tongue image to be tested based on the biological attribute description corresponding to the reference tongue image.
  • a tongue image with known biological attribute information can be used as a reference tongue image, and the user's physical condition can be determined by matching the user's tongue image with the reference tongue image.
  • step S110 a face image to be tested is acquired, and the person in the face image to be tested sticks out his tongue. Judging the biological attribute information of the human body based on the tongue image needs to be performed based on the full appearance of the tongue, so the person in the face image to be tested needs to stick out the tongue.
  • the face image to be tested can be taken under the guidance of a professional; it can also be taken by the user through a device with a shooting function such as a mobile phone after making corresponding actions under voice prompts.
  • the face image to be tested in this application The source of is not limited.
  • step S120 a tongue image to be tested is extracted from the face image to be tested.
  • the tongue image is based on the tongue. Therefore, it is necessary to extract the tongue image to be measured from the face image to be measured to reduce interference information and improve the accuracy of recognition.
  • the step of extracting the tongue image to be tested from the face image to be tested may include the following steps:
  • Step S301 Obtain an estimated face shape based on the pixel difference value of the relative position of the face image to be tested;
  • Step S302 establishing a normalized coordinate system for the estimated face shape to extract tongue feature points
  • step S303 an image of the tongue to be tested is captured based on the characteristic points.
  • face detection mainly uses the Cascaded Regression (ERT) algorithm, that is, the regression tree method of gradient enhancement learning.
  • ERT Cascaded Regression
  • the pixel difference of a relative position is used as the face feature point, which is used in the image
  • the face feature points are represented in the form of vectors to represent the shape of the face.
  • An embodiment of the present application also adopts an iterative method to overcome the influence of factors such as face orientation and skin on the estimated face shape.
  • a normalized coordinate system based on the current estimated face shape is used to extract the face feature points. Based on the face feature points, the update vector can be predicted, and the comparison can usually be obtained through several iterations. Precisely estimate the face shape.
  • An embodiment of the present application uses a cascade of multiple regressors.
  • the prediction of each regressor is based on the current face image to be tested and the previous prediction.
  • the calculation equation is as shown in formula (1), and the initial shape can be simply selected As the average shape of the training data, it is scaled according to the bounding box output of the universal face detector.
  • S (t) represents the current estimate of S
  • each regressor r t ( ⁇ , ⁇ ) in the cascade predicts the update vector of the image, and adds S (t) to the current shape estimate S (t)
  • t represents the cascade serial number
  • r t ( ⁇ , ⁇ ) represents the current stage regressor.
  • a regression tree of gradient enhancement learning is used to train each r t , and the least square method is used to minimize the error.
  • the input parameters of the regressor are the image I and the updated shape of the previous regressor, and the features used can be gray values or others.
  • Each regressor is composed of many trees, and the parameters of each tree are trained based on the coordinate difference between the current shape and the world coordinate system and randomly selected pixel pairs.
  • ERT is to directly store the updated value ⁇ S of the shape into the leaf node in the process of learning Tree. After the initial position S passes through all the learned trees, the average shape is added to the ⁇ S of all passed leaf nodes to obtain the final facial feature point position.
  • the estimated face shape can be obtained according to the facial feature points, a normalized coordinate system is established for the estimated face shape to extract tongue feature points, and the tongue image to be measured is intercepted based on the feature points.
  • the use of a normalized coordinate system can ensure that the intercepted tongue image to be measured is not affected by changes in the image scale, and improves the accuracy of the intercepted tongue image to be measured.
  • the step of establishing a normalized coordinate system for the estimated face shape to extract tongue feature points further includes:
  • Step S401 extracting left and right mouth corner points and tongue tip feature points from the face image to be tested based on a segmentation algorithm based on edge detection;
  • Step S402 using the connecting line of the left and right mouth corner points as the mouth corner reference line, using the mouth corner reference line as the X axis, and the vertical line of the mouth corner reference line as the Y axis, to establish a normalized coordinate system;
  • Step S403 Perform coordinate conversion on the tongue tip feature point to obtain the coordinate sequence of the tongue tip feature point in a normalized coordinate system
  • Step S404 Perform straight-line fitting on the coordinate sequence of the left and right mouth corner points and tongue feature points to extract tongue feature points.
  • the line connecting the left and right mouth corner points is at the horizontal position of the face image. Therefore, the line connecting the left and right mouth corner points can be used as the mouth corner reference line, the mouth corner reference line is taken as the X axis, and the mouth corner reference line The mid-perpendicular line is used as the Y axis to establish a normalized coordinate system.
  • the left and right corner points of the mouth and the characteristic points of the tongue tip are coordinated to obtain their coordinate sequence in the normalized coordinate system.
  • the coordinate sequence of the left and right corner points of the mouth and the characteristic points of the tongue is fitted with a straight line to extract the characteristic points of the tongue.
  • the step of extracting a tongue image to be tested from the face image to be tested may include the following steps:
  • Step S501 Perform color histogram smoothing processing on the tongue image to be tested.
  • Step S502 Perform format conversion on the tongue image to be tested after the color histogram smoothing process, to generate the tongue image to be tested in HSV format.
  • the color histogram describes the proportion of different colors in the entire image, and does not care about the spatial position of each color. Color histograms are particularly suitable for describing images that are difficult to segment automatically.
  • the values in the histogram are statistically derived, describing the quantitative characteristics of the colors in the image, which can reflect the statistical distribution and basic hue of the image colors; the histogram only contains the frequency of a certain color value in the image, which is missing The spatial location information of a certain pixel.
  • Any image can uniquely give a histogram corresponding to it, but different images may have the same color distribution, and thus have the same histogram, so the histogram and the image have a one-to-many relationship; If the image is divided into several sub-regions, the sum of the histograms of all sub-regions is equal to the histogram of the whole image.
  • a good threshold should correspond to the minimum value between the two peaks in the histogram.
  • the maximum value of the two peaks and the minimum value of the valley between them cannot be determined well, so the histogram needs to be smoothed.
  • An embodiment of the application uses Gaussian smoothing.
  • RGB red (Red), green (Green), blue (Blue) three primary colors.
  • the color mixing of the RGB color model belongs to additive color mixing. The higher the value of each primary color, the brighter the color. When R, G, and B are all 0, it is black, and when all are 255, it is white.
  • RGB is the most direct color representation method in computer design. The 24-bit true color image in the computer uses the RGB model to accurately record colors. Therefore, a certain color can be accurately obtained by using RGB values in a computer.
  • RGB represents directness, the three attributes of R, G, B and color are not directly related, and cannot reveal the relationship between colors.
  • the extracted tongue image to be tested is formatted, for example, the tongue image to be tested is converted from RGB format to HSV format to remove interference caused by image brightness , Improve the accuracy of image matching.
  • HSV refers to Hue (hue), saturation (saturation) and Value (value).
  • the three-dimensional representation of the HSV model evolved from the RGB cube, and the color information can be known more specifically.
  • the H parameter represents color information, that is, the position of the spectral color. This parameter is expressed in angles, and red, green, and blue are separated by 120 degrees. The complementary colors are 180 degrees apart.
  • V parameter represents the brightness of the color, ranging from 0 to 1.
  • Step S601 Establish a database, the database contains several reference tongue images
  • Step S602 Perform color feature extraction on the reference tongue image to generate a reference color histogram
  • Step S603 Add biological attribute information corresponding to the reference tongue image.
  • the difference in tongue color represents different biological attributes of the human body.
  • tongue images of common biological attributes and their physical conditions for reference.
  • a large number of reference tongue images of various biological attributes can be collected in advance, a database is established, the reference tongue images are imported into the database, and the corresponding relationship between the biological attributes and the reference tongue image is added.
  • Biological attribute information In order to remove the interference caused by the brightness of the picture and improve the accuracy of image matching, color feature extraction is performed on the reference tongue image to generate a reference color histogram.
  • the step of matching the reference tongue image closest to the tongue image to be tested from the database may further include the following steps:
  • Step S701 Perform color feature extraction on the tongue image to be tested to generate a color histogram to be tested.
  • Step S702 Use a histogram comparison function to match the target reference color histogram closest to the target reference color histogram from the database, and use the target reference color histogram as the reference tongue image corresponding to the target reference color histogram. The image closest to the reference tongue image.
  • the color histogram describes the proportion of different colors in the entire image, and does not care about the spatial position of each color.
  • the color histogram to be measured is matched with the reference color histogram in the database to match the target reference color histogram that is closest to the color histogram to be measured.
  • the matching is usually performed by a histogram comparison function.
  • the histogram comparison function includes: a histogram comparison compareHist function.
  • An embodiment of this application uses the histogram comparison compareHist function in opencv-python. There are four types of functions available.
  • the selected comparison distance correlation calculation, chi-square calculation, cross calculation, Bhattacharyya distance calculation.
  • the calculated result is in the interval of 0 to 1. The closer to 0, the higher the similarity between the two pictures.
  • the tongue image output result of the tongue image to be tested can be generated according to the biological attributes corresponding to the reference tongue image.
  • the generated tongue image output results indicate that the user is not a normal tongue, or cannot match the reference tongue image with high similarity, it can prompt the user whether there is any condition that affects the color of the tongue, for example, just eating warm or stimulating food , The tongue will become red or crimson; drinking milk, coffee and other foods that can stain the tongue coating. The user can be prompted to perform the test after eliminating the influence factors such as diet.
  • the tongue image matching method proposed in this application can extract the tongue image to be tested from the face image to be tested, match the reference tongue image closest to the tongue image to be tested, and generate it based on the biological attributes corresponding to the reference tongue image
  • the tongue image output result of the tongue image to be tested effectively improves the convenience and accuracy of tongue image matching, and improves the convenience and accuracy of the user to learn his own physical condition through the tongue image.
  • the application further provides an electronic device.
  • FIG. 7 is a schematic diagram of program modules of an electronic device 20 according to an exemplary embodiment of the present application.
  • the electronic device 20 includes:
  • the obtaining module 201 is adapted to obtain the face image to be tested
  • the extraction module 202 is adapted to extract the tongue image to be tested from the face image to be tested.
  • the matching module 203 is adapted to match the reference tongue image closest to the tongue image to be tested from a database, and generate a tongue image output result of the tongue image to be tested based on the biological attribute description corresponding to the reference tongue image.
  • the extraction module 202 further includes: an acquisition unit, adapted to acquire an estimated face shape based on the pixel difference of the relative position of the face image to be measured; an extraction unit, adapted to compare the estimated face The shape establishes a normalized coordinate system to extract tongue feature points; and the intercepting unit is adapted to intercept the tongue image to be measured based on the feature points.
  • the extraction unit is further adapted to extract left and right mouth corner points and tongue tip feature points from the face image to be tested based on a segmentation algorithm based on edge detection; using the connecting line of the left and right mouth corner points as the mouth corner reference line , Using the mouth corner reference line as the X axis and the vertical line of the mouth corner reference line as the Y axis to establish a normalized coordinate system; performing coordinate transformation on the tongue tip feature point to obtain the tongue tip feature point in the normalized The coordinate sequence in the coordinate system is transformed; the coordinate sequence of the left and right mouth corner points and the tongue tip feature points are linearly fitted to extract the tongue feature points.
  • the electronic device 20 further includes a processing module and a conversion module, the processing module is adapted to perform color histogram smoothing processing on the tongue image to be tested; the conversion module is adapted to perform color histogram smoothing processing The subsequent tongue image to be tested undergoes format conversion to generate the tongue image to be tested in HSV format.
  • the electronic device 20 further includes a conversion module, which is adapted to perform format conversion on the tongue image to be tested to generate the tongue image to be tested in HSV format.
  • the electronic device 20 further includes a database building module, adapted to build a database, the database contains a number of reference tongue images; color feature extraction of the reference tongue images to generate a reference color histogram; and add The biological attribute information corresponding to the reference tongue image.
  • the matching module 203 is further adapted to extract color features of the tongue image to be tested to generate a color histogram to be tested; and use a histogram comparison function to match the best color histogram from the database.
  • a target reference color histogram that is close, and a reference tongue image corresponding to the target reference color histogram is used as a reference tongue image that is closest to the tongue image to be tested.
  • the histogram comparison function includes: a histogram comparison compareHist function; the matching module 203 is further adapted to calculate the comparison distance between the color histogram to be measured and the target reference color histogram based on the histogram comparison compareHist function; select The target reference color histogram corresponding to the closest comparison distance is the target reference color histogram, and the reference tongue image corresponding to the target reference color histogram is used as the reference tongue image closest to the tongue image to be tested.
  • the electronic device 20 proposed in this application can extract a tongue image to be tested from a face image to be tested, match the reference tongue image closest to the tongue image to be tested, and generate based on the biological attribute information corresponding to the reference tongue image
  • the tongue image output result of the tongue image to be tested effectively improves the convenience and accuracy of the user to learn his own physical condition through the tongue image.
  • the present application also provides a computer device 20, including a memory 21, a processor 22, and computer-readable instructions stored on the memory 21 and running on the processor 22, and the processor 22 executes
  • the computer-readable instructions implement the steps of the above method.
  • the computer-readable instructions may be stored in the memory 24.
  • the present application also provides a non-volatile computer-readable storage medium on which computer-readable instructions are stored, and the computer-readable instructions implement the steps of the foregoing method when executed by a processor.
  • This application also provides a computer device, such as a smart phone, a tablet computer, a notebook computer, a desktop computer, a rack server, a blade server, a tower server or a cabinet server (including independent servers, or more A server cluster composed of two servers), etc.
  • the computer device in this embodiment at least includes, but is not limited to: a memory, a processor, etc. that can be communicatively connected to each other through a system bus.
  • This embodiment also provides a non-volatile computer-readable storage medium, such as flash memory, hard disk, multimedia card, card-type memory (for example, SD or DX memory, etc.), random access memory (RAM), static random access memory ( SRAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), programmable read-only memory (PROM), magnetic memory, magnetic disk, optical disk, server, App application mall, etc., on which storage There are computer-readable instructions, and the corresponding functions are realized when the program is executed by the processor.
  • the non-volatile computer-readable storage medium of this embodiment is used to store the electronic device 20, and when executed by the processor 22, implements the tongue image matching method of the present application.
  • the method of the above embodiments can be implemented by means of software plus the necessary general hardware platform. Of course, it can also be implemented by hardware, but in many cases the former is better. ⁇
  • the technical solution of this application essentially or the part that contributes to the existing technology can be embodied in the form of a software product, and the computer software product is stored in a storage medium (such as ROM/RAM, magnetic disk, The optical disc) includes several instructions to enable a terminal device (which can be a mobile phone, a computer, a server, an air conditioner, or a network device, etc.) to execute the method described in each embodiment of the present application.

Landscapes

  • Engineering & Computer Science (AREA)
  • Health & Medical Sciences (AREA)
  • Oral & Maxillofacial Surgery (AREA)
  • General Health & Medical Sciences (AREA)
  • Human Computer Interaction (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Multimedia (AREA)
  • Theoretical Computer Science (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Image Analysis (AREA)

Abstract

一种舌象匹配方法、电子装置、计算机设备及存储介质，通过建立数据库，所述数据库中包含若干基准舌图像；对所述基准舌图像进行颜色特征提取，生成基准颜色直方图；添加与所述基准舌图像对应的生物属性信息；获取待测人脸图像（S110）；从所述待测人脸图像中提取待测舌图像（S120）；对所述待测舌图像进行颜色直方图平滑处理；对进行颜色直方图平滑处理后的待测舌图像进行格式转换，生成HSV格式的待测舌图像；从数据库中匹配与所述待测舌图像最接近的基准舌图像，基于所述基准舌图像对应的生物属性描述生成所述待测舌图像的舌象输出结果（S130）。所述方法、电子装置、计算机设备及存储介质能够直观方便的给出舌象输出结果，有效提高用户通过舌象获知自己身体状况的方便性及准确性。

Description

舌象匹配方法、电子装置、计算机设备及存储介质
本申请要求于2019年7月31日提交中国专利局，专利名称为“舌象匹配方法、电子装置、计算机设备及存储介质”，申请号为201910699482.7的发明专利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
本申请涉及图像处理技术领域，尤其涉及一种舌象匹配方法、电子装置、计算机设备及存储介质。
背景技术
中医术语“舌为心之苗”，说的就是从观察舌头了解身体状况的意思。我们去看中医的时候，医生通常都会查看舌头。从中医学角度来说，舌头是健康的晴雨表。中医一般将舌头分为四部位，包括舌尖、舌中部、舌根、舌两侧，一般以舌尖属心肺、舌边属肝胆、舌中属脾胃、舌根属肾，五脏的状况都可以从这四部位反映出来，参照舌苔则可以更为全面的判断人的生物属性，而舌头颜色的不同则表征着人体不同的生物属性。
发明人发现普通用户通常不具备通过观察舌头判断自己身体状况的能力，目前基本是医生用肉眼观察患者的舌头形态和色泽等特征，通常需要患者去医院看诊，费时费力。
发明内容
有鉴于此，本申请提出一种舌象匹配方法、电子装置、计算机设备及存储介质，能够有效提高用户通过舌象获知自己身体状况的方便性及准确性。
首先，为实现上述目的，本申请提出一种舌象匹配方法，该方法包括步 骤：建立数据库，所述数据库中包含若干基准舌图像；对所述基准舌图像进行颜色特征提取，生成基准颜色直方图；添加与所述基准舌图像对应的生物属性信息；获取待测人脸图像；从所述待测人脸图像中提取待测舌图像；对所述待测舌图像进行颜色直方图平滑处理；对进行颜色直方图平滑处理后的待测舌图像进行格式转换，生成HSV格式的待测舌图像；从数据库中匹配与所述待测舌图像最接近的基准舌图像，基于所述基准舌图像对应的生物属性描述生成所述待测舌图像的舌象输出结果。
此外，为实现上述目的，本申请还提供一种电子装置，其包括：建库模块，适于建立数据库，所述数据库中包含若干基准舌图像，对所述基准舌图像进行颜色特征提取，生成基准颜色直方图，及添加与所述基准舌图像对应的生物属性信息；获取模块，适于获取待测人脸图像；提取模块，适于从所述待测人脸图像中提取待测舌图像；处理模块，适于对所述待测舌图像进行颜色直方图平滑处理；所述转换模块；转换模块，适于对进行颜色直方图平滑处理后的待测舌图像进行格式转换，生成HSV格式的待测舌图像；匹配模块，适于从数据库中匹配与所述待测舌图像最接近的基准舌图像，基于所述基准舌图像对应的生物属性描述生成所述待测舌图像的舌象输出结果。
为实现上述目的，本申请还提供一种计算机设备，包括存储器、处理器以及存储在存储器上并可在所述处理器上运行的计算机可读指令，所述处理器执行所述计算机可读指令时实现上述方法的步骤。
为实现上述目的，本申请还提供非易失性计算机可读存储介质，其上存储有计算机可读指令，所述计算机可读指令被处理器执行时实现上述方法的步骤。
本申请所提出的舌象匹配方法、电子装置、计算机设备及存储介质，能够从待测人脸图像中提取待测舌图像，匹配与所述待测舌图像最接近的基准舌图像，基于所述基准舌图像对应的生物属性描述生成所述待测舌图像的舌象输出结果，能够直观方便的给出舌象输出结果，有效提高用户通过舌象获 知自己身体状况的方便性及准确性。
附图说明
图1是本申请一示例性实施例示出的舌象匹配方法的流程示意图；
图2是本申请一示例性实施例示出的舌头部分与五脏对应的示意图；
图3是本申请一示例性实施例示出的舌象匹配方法的流程示意图；
图4是本申请一示例性实施例示出的舌象匹配方法的流程示意图；
图5是本申请一示例性实施例示出的舌象匹配方法的流程示意图；
图6是本申请一示例性实施例示出的舌象匹配方法的流程示意图；
图7是本申请一示例性实施例示出的舌象匹配方法的流程示意图；
图8是本申请一示例性实施例示出的电子装置的程序模块示意图；
图9是本申请一示例性实施例示出的电子装置的硬件架构示意图。
本申请目的的实现、功能特点及优点将结合实施例，参照附图做进一步说明。
具体实施方式
为了使本申请的目的、技术方案及优点更加清楚明白，以下结合附图及实施例，对本申请进行进一步详细说明。应当理解，此处所描述的具体实施例仅用以解释本申请，并不用于限定本申请。基于本申请中的实施例，本领域普通技术人员在没有做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。
需要说明的是，在本申请中涉及“第一”、“第二”等的描述仅用于描述目的，而不能理解为指示或暗示其相对重要性或者隐含指明所指示的技术特征的数量。由此，限定有“第一”、“第二”的特征可以明示或者隐含地包括至少一个该特征。另外，各个实施例之间的技术方案可以相互结合，但是必须是以本领域普通技术人员能够实现为基础，当技术方案的结合出现相互矛盾或无法实 现时应当认为这种技术方案的结合不存在，也不在本申请要求的保护范围之内。
参阅图1所示，是本申请一实施例之舌象匹配方法的流程示意图，所述方法包括以下步骤：
步骤S110，获取待测人脸图像；
步骤S120，从所述待测人脸图像中提取待测舌图像；及
步骤S130，从数据库中匹配与所述待测舌图像最接近的基准舌图像，基于所述基准舌图像对应的生物属性描述生成所述待测舌图像的舌象输出结果。
因此，通过观察舌头，能了解人体生理变化。可以将已知生物属性信息的舌头图像作为基准舌图像，通过将用户的舌头图像与基准舌图像进行匹配以此来判断用户的身体状况。
在步骤S110中，获取待测人脸图像，所述待测人脸图像中的人物伸出舌头。根据舌象判断人体的生物属性信息需要根据舌头的全貌进行，所以需要所述待测人脸图像中的人物伸出舌头。通常来说，拍摄舌头图像还有一些注意事项，例如，用户需要将舌自然伸出口外，充分暴露，要呈扁平形，使舌体放松，不要卷缩，也不要过分用力，以免引起颜色的改变。因此，所述待测人脸图像可以是在专业人员指导下拍摄；也可以是用户通过手机等具有拍摄功能的设备，在语音提示下做出相应动作后拍摄的，本申请对待测人脸图像的来源不作限定。
在步骤S120中，从所述待测人脸图像中提取待测舌图像。如前所述，舌象是基于舌头进行的，因此，需要从所述待测人脸图像中提取待测舌图像，以减少干扰信息，提高辨识的准确度。
如图3所示，本申请一实施例中，所述从所述待测人脸图像中提取待测舌图像的步骤可以包括以下步骤：
步骤S301，基于所述待测人脸图像的相对位置的像素差值获取预估人脸形状；
步骤S302，对所述预估人脸形状建立归一化坐标系提取舌头特征点；及
步骤S303，基于所述特征点截取待测舌图像。
本申请一实施例中，人脸检测主要使用了级联回归器(ERT)算法，即梯度提高学习的回归树方法，首先使用一个相对位置的像素差值作为人脸特征点，在图像中使用向量的形式表征所述人脸特征点来表示人脸的形状，本申请一实施例还采用迭代方法以克服人脸方向、皮肤等因素对预估人脸形状的影响。且本申请一实施例采取基于当前预估人脸形状的归一化坐标系来提取所述人脸特征点，基于所述人脸特征点可以预测更新向量，通常通过几次迭代即可获得较精确的预估人脸形状。
本申请一实施例采用多个回归器的级联，每个回归器的预测基于当前待测人脸图像和前一次的预测，其计算方程如公式(1)所示，可以简单地选择初始形状作为训练数据的平均形状，并根据通用面部检测器的边界框输出进行缩放。
Figure PCTCN2019118093-appb-000001
其中，S (t)表示对S的当前估计，级联中的每个回归量r t(·,·)预测图像的更新向量，并将S (t)添加到当前形状估计S (t)中，以改进估计，t表示级联序号，r t(·,·)表示当前级的回归器regressor。
本申请一实施例使用梯度提高学习的回归树训练每个r t，使用最小二乘法最小化误差。回归器的输入参数为图像I和上一级回归器更新后的形状，采用的特征可以是灰度值或者其它。每个回归器由很多棵树(tree)组成，每棵树参数是根据当前形状与世界坐标系的坐标差及随机挑选的像素对训练得到的。
ERT是在学习Tree的过程中，直接将形状的更新值ΔS存入叶子结点leaf node。初始位置S在通过所有学习到的Tree后，平均形状加上所有经过的叶子结点的ΔS，即可得到最终的人脸特征点位置。根据所述人脸特征点可以获取预估人脸形状，对所述预估人脸形状建立归一化坐标系提取舌头特征点，基于所述特征点截取待测舌图像。采用归一化坐标系可以保证截取的待测舌图 像不受图像尺度变化的影响，提高了截取的待测舌图像的准确性。
如图4所示，本申请一实施例中，所述对所述预估人脸形状建立归一化坐标系提取舌头特征点的步骤还包括：
步骤S401，基于边缘检测的分割算法从所述待测人脸图像中提取左、右嘴角点及舌尖特征点；
步骤S402，以所述左、右嘴角点连线作为嘴角基准线，将所述嘴角基准线作为X轴，所述嘴角基准线的中垂线作为Y轴，建立归一化坐标系；
步骤S403，对所述舌尖特征点进行坐标转换，得到所述舌尖特征点在归一化坐标系内的坐标序列；
步骤S404，对左、右嘴角点及舌尖特征点的坐标序列进行直线拟合提取舌头特征点。
由于嘴唇、舌头与皮肤颜色不同，因此，可以采用边缘检测的分割算法等算法对所述待测人脸图像进行处理，提取左、右嘴角点及舌尖特征点。通常来说，左、右嘴角点的连线位于人脸图像的水平位置，因此，可以左、右嘴角点连线作为嘴角基准线，将所述嘴角基准线作为X轴，所述嘴角基准线的中垂线作为Y轴，建立归一化坐标系。将左、右嘴角点及舌尖特征点进行坐标转换，得到其在归一化坐标系内的坐标序列。对左、右嘴角点及舌尖特征点的坐标序列进行直线拟合提取舌头特征点。
如图5所示，本申请一实施例中，所述从所述待测人脸图像中提取待测舌图像的步骤可以包括以下步骤：
步骤S501，对所述待测舌图像进行颜色直方图平滑处理。
步骤S502，对进行颜色直方图平滑处理后的待测舌图像进行格式转换，生成HSV格式的待测舌图像。
颜色直方图所描述的是不同色彩在整幅图像中所占的比例，而并不关心每种色彩所处的空间位置。颜色直方图特别适于描述那些难以进行自动分割的图像。直方图中的数值是统计而来的，描述了图像中关于颜色的数量特征， 可以反映图像颜色的统计分布和基本色调；直方图只包含了该图像中某一颜色值出现的频数，而丢失了某象素所在的空间位置信息。
任一幅图像都能唯一的给出一幅与它对应的直方图，但不同的图像可能有相同的颜色分布，从而就具有相同的直方图，因此直方图与图像是一对多的关系；如将图像划分为若干个子区域，所有子区域的直方图之和等于全图直方图。一般情况下，由于图像上的背景和前景物体颜色分布明显不同，从而在直方图上会出现双峰特性，但背景和前景颜色较为接近的图像不具有这个特性。一个好的阈值应该对应着直方图中两个峰之间的最小值。但是，由于灰度直方图中的随机波动，两个峰尖的最大值和他们之间谷底的最小值都不能被很好的确定，所以需对直方图进行平滑处理。本申请一实施例采用的是高斯平滑。
由于光照等因素会影响所拍摄的图像的亮度，而亮度会影响图像的R、G、B数值，从而降低图像匹配的准确性。所谓RGB就是：红(Red)、绿(Green)、蓝(Blue)三种色光原色。RGB色彩模型的混色属于加法混色。每种原色的数值越高，色彩越明亮。R、G、B都为0时是黑色，都为255时是白色。RGB是电脑设计中最直接的色彩表示方法。电脑中的24位真彩图像，就是采用RGB模型来精确记录色彩。所以，在电脑中利用RGB数值可以精确取得某种颜色。RGB虽然表示直接，但是R、G、B数值和色彩的三属性没有直接的联系，不能揭示色彩之间的关系。
为此，本申请一实施例中，会对提取的所述待测舌图像进行格式转换，例如，将所述待测舌图像由RGB格式转换为HSV格式，以去除由于图片亮度带来的干扰，提高图像匹配的准确性。HSV是指Hue(色相)、Saturation(饱和度)和Value(值)。
HSV模型的三维表示从RGB立方体演化而来，可以更具体的知道颜色信息。
1)、H参数表示色彩信息，即所处的光谱颜色的位置。该参数用角度量来 表示，红、绿、蓝分别相隔120度。互补色分别相差180度。
2)、S参数表示饱和度，该参数为一比例值，范围从0到1，它表示成所选颜色的纯度和该颜色最大的纯度之间的比率。S＝0时，只有灰度。
3)、V参数表示色彩的明亮程度，范围从0到1。有一点要注意：它和光强度之间并没有直接的联系。
如图6所示，本申请一实施例中，所述从数据库中匹配与所述待测舌图像最接近的基准舌图像的步骤之前，还可以包括以下步骤：
步骤S601，建立数据库，所述数据库中包含若干基准舌图像；
步骤S602，对所述基准舌图像进行颜色特征提取，生成基准颜色直方图；及
步骤S603，添加与所述基准舌图像对应的生物属性信息。
如前所述，舌头颜色的不同则表征着人体不同的生物属性，目前已经有几十种常见生物属性的舌象图及其表征的身体状况供参考。可以预先采集大量各种生物属性的基准舌象图，建立数据库，将所述基准舌象图导入所述数据库中，根据生物属性与基准舌象图的对应关系，添加与所述基准舌图像对应的生物属性信息。为了去除由于图片亮度带来的干扰，提高图像匹配的准确性，对所述基准舌图像进行颜色特征提取，生成基准颜色直方图。
如图7所示，本申请一实施例中，所述从数据库中匹配与所述待测舌图像最接近的基准舌图像的步骤还可以包括以下步骤：
步骤S701，对所述待测舌图像进行颜色特征提取，生成待测颜色直方图；及
步骤S702，通过直方图比较函数，从数据库中匹配与所述待测颜色直方图最接近的目标基准颜色直方图，以所述目标基准颜色直方图对应的基准舌图像作为与所述待测舌图像最接近的基准舌图像。
颜色直方图所描述的是不同色彩在整幅图像中所占的比例，而并不关心每种色彩所处的空间位置。对所述待测舌图像进行颜色特征提取，生成待测 颜色直方图。将所述待测颜色直方图与数据库中的基准颜色直方图进行匹配，以匹配与所述待测颜色直方图最接近的目标基准颜色直方图。通常是通过直方图比较函数进行匹配的，所述直方图比较函数包括：直方图比较compareHist函数，本申请一实施例使用的是opencv-python中的直方图比较compareHist函数，其中共有四种可供选择的比较距离：相关性计算、卡方计算、十字计算、巴氏距离计算。以巴氏距离计算为例，计算后的结果为0到1区间内，越趋近于0，表明两图的相似性越高。在匹配出最接近的基准舌图像后，可以根据该基准舌图像对应的生物属性生成所述待测舌图像的舌象输出结果。
进一步地，当生成的舌象输出结果表明用户不是正常舌，或者无法匹配相似性较高的基准舌图像时，可以提示用户是否存在影响舌头颜色的情况，例如，刚进食温热或刺激性食物，舌质会变红或绛；饮用了牛奶、咖啡等可以给舌苔染色的食物等。可以提示用户在消除饮食等影响因素后再行测试。
本申请所提出的舌象匹配方法，能够从待测人脸图像中提取待测舌图像，匹配与所述待测舌图像最接近的基准舌图像，基于所述基准舌图像对应的生物属性生成所述待测舌图像的舌象输出结果，有效提高舌象匹配的方便性及准确性，提高用户通过舌象获知自己身体状况的方便性及准确性。
本申请进一步提供一种电子装置。参阅图7，是本申请一示例性实施例示出的电子装置20的程序模块示意图。
所述电子装置20包括：
获取模块201，适于获取待测人脸图像；
提取模块202，适于从所述待测人脸图像中提取待测舌图像；及
匹配模块203，适于从数据库中匹配与所述待测舌图像最接近的基准舌图像，基于所述基准舌图像对应的生物属性描述生成所述待测舌图像的舌象输出结果。
进一步地，所述提取模块202还包括：获取单元，适于基于所述待测人脸图像的相对位置的像素差值获取预估人脸形状；提取单元，适于对所述预 估人脸形状建立归一化坐标系提取舌头特征点；及截取单元，适于基于所述特征点截取待测舌图像。
进一步地，所述提取单元还适于基于边缘检测的分割算法从所述待测人脸图像中提取左、右嘴角点及舌尖特征点；以所述左、右嘴角点连线作为嘴角基准线，将所述嘴角基准线作为X轴，所述嘴角基准线的中垂线作为Y轴，建立归一化坐标系；对所述舌尖特征点进行坐标转换，得到所述舌尖特征点在归一化坐标系内的坐标序列；对左、右嘴角点及舌尖特征点的坐标序列进行直线拟合提取舌头特征点。
进一步地，所述电子装置20还包括处理模块及转换模块，所述处理模块适于对所述待测舌图像进行颜色直方图平滑处理；所述转换模块，适于对进行颜色直方图平滑处理后的待测舌图像进行格式转换，生成HSV格式的待测舌图像。
进一步地，所述电子装置20还包括转换模块，适于对所述待测舌图像进行格式转换，生成HSV格式的待测舌图像。
进一步地，所述电子装置20还包括建库模块，适于建立数据库，所述数据库中包含若干基准舌图像；对所述基准舌图像进行颜色特征提取，生成基准颜色直方图；及添加与所述基准舌图像对应的生物属性信息。
进一步地，所述匹配模块203还适于对所述待测舌图像进行颜色特征提取，生成待测颜色直方图；及通过直方图比较函数，从数据库中匹配与所述待测颜色直方图最接近的目标基准颜色直方图，以所述目标基准颜色直方图对应的基准舌图像作为与所述待测舌图像最接近的基准舌图像。
进一步地，所述直方图比较函数包括：直方图比较compareHist函数；所述匹配模块203还适于基于直方图比较compareHist函数计算所述待测颜色直方图与目标基准颜色直方图的比较距离；选取最接近的比较距离对应的目标基准颜色直方图为目标基准颜色直方图，以所述目标基准颜色直方图对应的基准舌图像作为与所述待测舌图像最接近的基准舌图像。
本申请所提出的电子装置20，能够从待测人脸图像中提取待测舌图像，匹配与所述待测舌图像最接近的基准舌图像，基于所述基准舌图像对应的生物属性信息生成所述待测舌图像的舌象输出结果，有效提高用户通过舌象获知自己身体状况的方便性及准确性。
为实现上述目的，本申请还提供一种计算机设备20，包括存储器21、处理器22以及存储在存储器21上并可在所述处理器22上运行的计算机可读指令，所述处理器22执行所述计算机可读指令时实现上述方法的步骤。可以将所述计算机可读指令存储于内存24中。
为实现上述目的，本申请还提供非易失性计算机可读存储介质，其上存储有计算机可读指令，所述计算机可读指令被处理器执行时实现上述方法的步骤。
本申请还提供一种计算机设备，如可以执行程序的智能手机、平板电脑、笔记本电脑、台式计算机、机架式服务器、刀片式服务器、塔式服务器或机柜式服务器(包括独立的服务器，或者多个服务器所组成的服务器集群)等。本实施例的计算机设备至少包括但不限于：可通过系统总线相互通信连接的存储器、处理器等。
本实施例还提供一种非易失性计算机可读存储介质，如闪存、硬盘、多媒体卡、卡型存储器(例如，SD或DX存储器等)、随机访问存储器(RAM)、静态随机访问存储器(SRAM)、只读存储器(ROM)、电可擦除可编程只读存储器(EEPROM)、可编程只读存储器(PROM)、磁性存储器、磁盘、光盘、服务器、App应用商城等等，其上存储有计算机可读指令，程序被处理器执行时实现相应功能。本实施例的非易失性计算机可读存储介质用于存储电子装置20，被处理器22执行时实现本申请的舌象匹配方法。
通过以上的实施方式的描述，本领域的技术人员可以清楚地了解到上述实施例方法可借助软件加必需的通用硬件平台的方式来实现，当然也可以通过硬件，但很多情况下前者是更佳的实施方式。基于这样的理解，本申请的 技术方案本质上或者说对现有技术做出贡献的部分可以以软件产品的形式体现出来，该计算机软件产品存储在一个存储介质(如ROM/RAM、磁碟、光盘)中，包括若干指令用以使得一台终端设备(可以是手机，计算机，服务器，空调器，或者网络设备等)执行本申请各个实施例所述的方法。
以上仅为本申请的优选实施例，并非因此限制本申请的专利范围，凡是利用本申请说明书及附图内容所作的等效结构或等效流程变换，或直接或间接运用在其他相关的技术领域，均同理包括在本申请的专利保护范围内。

Claims (20)

  1. 一种基于图像检测的舌象匹配方法，所述方法包括步骤：
    建立数据库，所述数据库中包含若干基准舌图像；
    对所述基准舌图像进行颜色特征提取，生成基准颜色直方图；
    添加与所述基准舌图像对应的生物属性信息；
    获取待测人脸图像；
    从所述待测人脸图像中提取待测舌图像；
    对所述待测舌图像进行颜色直方图平滑处理；
    对进行颜色直方图平滑处理后的待测舌图像进行格式转换，生成HSV格式的待测舌图像；
    从数据库中匹配与所述待测舌图像最接近的基准舌图像，基于所述基准舌图像对应的生物属性描述生成所述待测舌图像的舌象输出结果。
  2. 如权利要求1所述的舌象匹配方法，所述从所述待测人脸图像中提取待测舌图像的步骤还包括：
    基于所述待测人脸图像的相对位置的像素差值获取预估人脸形状；
    对所述预估人脸形状建立归一化坐标系提取舌头特征点；及
    基于所述特征点截取待测舌图像。
  3. 如权利要求2所述的舌象匹配方法，所述对所述预估人脸形状建立归一化坐标系提取舌头特征点的步骤还包括：
    基于边缘检测的分割算法从所述待测人脸图像中提取左、右嘴角点及舌尖特征点；
    以所述左、右嘴角点连线作为嘴角基准线，将所述嘴角基准线作为X轴，所述嘴角基准线的中垂线作为Y轴，建立归一化坐标系；
    对所述舌尖特征点进行坐标转换，得到所述舌尖特征点在归一化坐标系内的坐标序列；
    对左、右嘴角点及舌尖特征点的坐标序列进行直线拟合提取舌头特征点。
  4. 如权利要求3所述的舌象匹配方法，所述从数据库中匹配与所述待测舌图像最接近的基准舌图像的步骤还包括：
    对所述待测舌图像进行颜色特征提取，生成待测颜色直方图；及
    通过直方图比较函数，从数据库中匹配与所述待测颜色直方图最接近的目标基准颜色直方图，以所述目标基准颜色直方图对应的基准舌图像作为与所述待测舌图像最接近的基准舌图像。
  5. 如权利要求4所述的舌象匹配方法，所述直方图比较函数包括：直方图比较compareHist函数；
    基于直方图比较compareHist函数计算所述待测颜色直方图与目标基准颜色直方图的比较距离；
    选取最接近的比较距离对应的目标基准颜色直方图为目标基准颜色直方图，以所述目标基准颜色直方图对应的基准舌图像作为与所述待测舌图像最接近的基准舌图像。
  6. 一种电子装置，其包括：
    建库模块，适于建立数据库，所述数据库中包含若干基准舌图像，对所述基准舌图像进行颜色特征提取，生成基准颜色直方图，及添加与所述基准舌图像对应的生物属性信息；
    获取模块，适于获取待测人脸图像；
    提取模块，适于从所述待测人脸图像中提取待测舌图像；
    处理模块，适于对所述待测舌图像进行颜色直方图平滑处理；所述转换模块；
    转换模块，适于对进行颜色直方图平滑处理后的待测舌图像进行格式转换，生成HSV格式的待测舌图像；
    匹配模块，适于从数据库中匹配与所述待测舌图像最接近的基准舌图像，基于所述基准舌图像对应的生物属性描述生成所述待测舌图像的舌象输出结果。
  7. 如权利要求6所述的电子装置，所述提取模块还包括：获取单元，适于基于所述待测人脸图像的相对位置的像素差值获取预估人脸形状；提取单元，适于对所述预估人脸形状建立归一化坐标系提取舌头特征点；及截取单元，适于基于所述特征点截取待测舌图像。
  8. 如权利要求7所述的电子装置，所述所述提取单元还适于基于边缘检测的分割算法从所述待测人脸图像中提取左、右嘴角点及舌尖特征点；以所述左、右嘴角点连线作为嘴角基准线，将所述嘴角基准线作为X轴，所述嘴角基准线的中垂线作为Y轴，建立归一化坐标系；对所述舌尖特征点进行坐标转换，得到所述舌尖特征点在归一化坐标系内的坐标序列；对左、右嘴角点及舌尖特征点的坐标序列进行直线拟合提取舌头特征点。
  9. 如权利要求8所述的电子装置，所述匹配模块还适于对所述待测舌图像进行颜色特征提取，生成待测颜色直方图；及通过直方图比较函数，从数据库中匹配与所述待测颜色直方图最接近的目标基准颜色直方图，以所述目标基准颜色直方图对应的基准舌图像作为与所述待测舌图像最接近的基准舌图像。
  10. 如权利要求9所述的电子装置，所述直方图比较函数包括：直方图比较compareHist函数，所述匹配模块还适于基于直方图比较compareHist函数计算所述待测颜色直方图与目标基准颜色直方图的比较距离，选取最接近的比较距离对应的目标基准颜色直方图为目标基准颜色直方图，以所述目标基准颜色直方图对应的基准舌图像作为与所述待测舌图像最接近的基准舌图像。
  11. 一种计算机设备，包括存储器、处理器以及存储在存储器上并可在处理器上运行的计算机可读指令，所述处理器执行所述计算机可读指令时实现、舌象匹配方法包括步骤：
    建立数据库，所述数据库中包含若干基准舌图像；
    对所述基准舌图像进行颜色特征提取，生成基准颜色直方图；
    添加与所述基准舌图像对应的生物属性信息；
    获取待测人脸图像；
    从所述待测人脸图像中提取待测舌图像；
    对所述待测舌图像进行颜色直方图平滑处理；
    对进行颜色直方图平滑处理后的待测舌图像进行格式转换，生成HSV格式的待测舌图像；
    从数据库中匹配与所述待测舌图像最接近的基准舌图像，基于所述基准舌图像对应的生物属性描述生成所述待测舌图像的舌象输出结果。
  12. 如权利要求11所述的计算机设备，所述从所述待测人脸图像中提取待测舌图像的步骤还包括：
    基于所述待测人脸图像的相对位置的像素差值获取预估人脸形状；
    对所述预估人脸形状建立归一化坐标系提取舌头特征点；
    基于所述特征点截取待测舌图像。
  13. 如权利要求12所述的计算机设备，所述对所述预估人脸形状建立归一化坐标系提取舌头特征点的步骤还包括：
    基于边缘检测的分割算法从所述待测人脸图像中提取左、右嘴角点及舌尖特征点；
    以所述左、右嘴角点连线作为嘴角基准线，将所述嘴角基准线作为X轴，所述嘴角基准线的中垂线作为Y轴，建立归一化坐标系；
    对所述舌尖特征点进行坐标转换，得到所述舌尖特征点在归一化坐标系内的坐标序列；
    对左、右嘴角点及舌尖特征点的坐标序列进行直线拟合提取舌头特征点。
  14. 如权利要求13所述的计算机设备，所述从数据库中匹配与所述待测舌图像最接近的基准舌图像的步骤还包括：
    对所述待测舌图像进行颜色特征提取，生成待测颜色直方图；及
    通过直方图比较函数，从数据库中匹配与所述待测颜色直方图最接近的目标基准颜色直方图，以所述目标基准颜色直方图对应的基准舌图像作为与 所述待测舌图像最接近的基准舌图像。
  15. 如权利要求14所述的舌象匹配方法，所述直方图比较函数包括：直方图比较compareHist函数；
    基于直方图比较compareHist函数计算所述待测颜色直方图与目标基准颜色直方图的比较距离；
    选取最接近的比较距离对应的目标基准颜色直方图为目标基准颜色直方图，以所述目标基准颜色直方图对应的基准舌图像作为与所述待测舌图像最接近的基准舌图像。
  16. 一种非易失性计算机可读存储介质，其上存储有计算机可读指令，所述计算机可读指令被处理器执行时实现舌象匹配方法包括步骤：
    建立数据库，所述数据库中包含若干基准舌图像；
    对所述基准舌图像进行颜色特征提取，生成基准颜色直方图；
    添加与所述基准舌图像对应的生物属性信息；
    获取待测人脸图像；
    从所述待测人脸图像中提取待测舌图像；
    对所述待测舌图像进行颜色直方图平滑处理；
    对进行颜色直方图平滑处理后的待测舌图像进行格式转换，生成HSV格式的待测舌图像；
    从数据库中匹配与所述待测舌图像最接近的基准舌图像，基于所述基准舌图像对应的生物属性描述生成所述待测舌图像的舌象输出结果。
  17. 如权利要求16所述的非易失性计算机可读存储介质，所述从所述待测人脸图像中提取待测舌图像的步骤还包括：
    基于所述待测人脸图像的相对位置的像素差值获取预估人脸形状；
    对所述预估人脸形状建立归一化坐标系提取舌头特征点；及
    基于所述特征点截取待测舌图像。
  18. 如权利要求17所述的非易失性计算机可读存储介质，所述对所述预估 人脸形状建立归一化坐标系提取舌头特征点的步骤还包括：
    基于边缘检测的分割算法从所述待测人脸图像中提取左、右嘴角点及舌尖特征点；
    以所述左、右嘴角点连线作为嘴角基准线，将所述嘴角基准线作为X轴，所述嘴角基准线的中垂线作为Y轴，建立归一化坐标系；
    对所述舌尖特征点进行坐标转换，得到所述舌尖特征点在归一化坐标系内的坐标序列；
    对左、右嘴角点及舌尖特征点的坐标序列进行直线拟合提取舌头特征点。
  19. 如权利要求18所述的非易失性计算机可读存储介质，所述从数据库中匹配与所述待测舌图像最接近的基准舌图像的步骤还包括：
    对所述待测舌图像进行颜色特征提取，生成待测颜色直方图；及
    通过直方图比较函数，从数据库中匹配与所述待测颜色直方图最接近的目标基准颜色直方图，以所述目标基准颜色直方图对应的基准舌图像作为与所述待测舌图像最接近的基准舌图像。
  20. 如权利要求19所述的非易失性计算机可读存储介质，所述直方图比较函数包括：直方图比较compareHist函数；
    基于直方图比较compareHist函数计算所述待测颜色直方图与目标基准颜色直方图的比较距离；
    选取最接近的比较距离对应的目标基准颜色直方图为目标基准颜色直方图，以所述目标基准颜色直方图对应的基准舌图像作为与所述待测舌图像最接近的基准舌图像。
PCT/CN2019/118093 2019-07-31 2019-11-13 舌象匹配方法、电子装置、计算机设备及存储介质 Ceased WO2021017308A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201910699482.7 2019-07-31
CN201910699482.7A CN110598533A (zh) 2019-07-31 2019-07-31 舌象匹配方法、电子装置、计算机设备及存储介质

Publications (1)

Publication Number Publication Date
WO2021017308A1 true WO2021017308A1 (zh) 2021-02-04

Family

ID=68853253

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2019/118093 Ceased WO2021017308A1 (zh) 2019-07-31 2019-11-13 舌象匹配方法、电子装置、计算机设备及存储介质

Country Status (2)

Country Link
CN (1) CN110598533A (zh)
WO (1) WO2021017308A1 (zh)

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN117094966A (zh) * 2023-08-21 2023-11-21 青岛美迪康数字工程有限公司 基于图像扩增的舌图像识别方法、装置和计算机设备
CN118135613A (zh) * 2024-05-08 2024-06-04 江西中医药大学 一种基于混合深度神经网络的舌象图像分割方法及系统
CN119068223A (zh) * 2024-09-27 2024-12-03 荣耀终端有限公司 一种对象检测方法、对象检测系统和电子设备

Families Citing this family (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111738980B (zh) * 2020-05-14 2023-08-04 北京深睿博联科技有限责任公司 一种医学影像的显示方法、计算机设备及存储介质

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20040151379A1 (en) * 2002-12-28 2004-08-05 Samsung Electronics Co., Ltd. Method of digital image analysis for isolating a region of interest within a tongue image and health monitoring method and apparatus using the tongue image
CN101576913A (zh) * 2009-06-12 2009-11-11 中国科学技术大学 基于自组织映射神经网络的舌象自动聚类、可视化和检索系统
CN102426583A (zh) * 2011-10-10 2012-04-25 北京工业大学 基于图像内容分析的中医舌象检索方法
CN103745217A (zh) * 2013-12-31 2014-04-23 北京工业大学 基于图像检索的中医舌色苔色自动分析方法

Family Cites Families (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109190535B (zh) * 2018-08-23 2022-04-01 南京邮电大学 一种基于深度学习的人脸面部肤色分析方法和系统

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20040151379A1 (en) * 2002-12-28 2004-08-05 Samsung Electronics Co., Ltd. Method of digital image analysis for isolating a region of interest within a tongue image and health monitoring method and apparatus using the tongue image
CN101576913A (zh) * 2009-06-12 2009-11-11 中国科学技术大学 基于自组织映射神经网络的舌象自动聚类、可视化和检索系统
CN102426583A (zh) * 2011-10-10 2012-04-25 北京工业大学 基于图像内容分析的中医舌象检索方法
CN103745217A (zh) * 2013-12-31 2014-04-23 北京工业大学 基于图像检索的中医舌色苔色自动分析方法

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN117094966A (zh) * 2023-08-21 2023-11-21 青岛美迪康数字工程有限公司 基于图像扩增的舌图像识别方法、装置和计算机设备
CN117094966B (zh) * 2023-08-21 2024-04-05 青岛美迪康数字工程有限公司 基于图像扩增的舌图像识别方法、装置和计算机设备
CN118135613A (zh) * 2024-05-08 2024-06-04 江西中医药大学 一种基于混合深度神经网络的舌象图像分割方法及系统
CN119068223A (zh) * 2024-09-27 2024-12-03 荣耀终端有限公司 一种对象检测方法、对象检测系统和电子设备

Also Published As

Publication number Publication date
CN110598533A (zh) 2019-12-20

Similar Documents

Publication Publication Date Title
US11900647B2 (en) Image classification method, apparatus, and device, storage medium, and medical electronic device
WO2021017308A1 (zh) 舌象匹配方法、电子装置、计算机设备及存储介质
WO2019100282A1 (zh) 一种人脸肤色识别方法、装置和智能终端
EP2545531B1 (en) System for skin treatment analysis using spectral image data to generate 3d rgb model
CN107679466B (zh) 信息输出方法和装置
WO2021218696A1 (zh) 肤色检测方法、装置、终端和存储介质
WO2019019828A1 (zh) 目标对象的遮挡检测方法及装置、电子设备及存储介质
WO2020062523A1 (zh) 注视点判断方法和装置、电子设备和计算机存储介质
CN108985210A (zh) 一种基于人眼几何特征的视线追踪方法及系统
CN110458830A (zh) 图像处理方法、装置、服务器及存储介质
CN106878695A (zh) 白平衡处理的方法、装置和计算机设备
EP2856426A1 (en) Body measurement
JP2023539483A (ja) 医用画像処理方法、装置、機器、記憶媒体及びコンピュータプログラム
CN109002846B (zh) 一种图像识别方法、装置和存储介质
WO2025200534A1 (zh) 基于改进特征描述符和knn搜索的眼底图像特征点匹配方法
US9924865B2 (en) Apparatus and method for estimating gaze from un-calibrated eye measurement points
CN109241898B (zh) 腔镜视像的目标定位方法和系统、存储介质
CN109166172B (zh) 服装模型的构建方法、装置、服务器和存储介质
US10176612B2 (en) System and method for retrieval of similar findings from a hybrid image dataset
WO2022161235A1 (zh) 身份识别方法、装置、设备、存储介质及计算机程序产品
CN115170460A (zh) 图像识别方法、图像识别装置、图像采集器及存储介质
CN116052209B (zh) 一种红外图像和标准3d人体模型的处理方法及存储介质
CN115100121A (zh) 中医面部脏腑分割和颜色特征的疾病筛查方法
Ma et al. Multimodal retinal image registration using edge map and feature guided Gaussian mixture model
CN112785683B (zh) 一种人脸图像调整方法及装置

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: 19939306

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: 19939306

Country of ref document: EP

Kind code of ref document: A1

122 Ep: pct application non-entry in european phase

Ref document number: 19939306

Country of ref document: EP

Kind code of ref document: A1