WO2019014813A1 - 一种定量检测人脸肤质参量的方法、装置和智能终端 - Google Patents
一种定量检测人脸肤质参量的方法、装置和智能终端 Download PDFInfo
- Publication number
- WO2019014813A1 WO2019014813A1 PCT/CN2017/093190 CN2017093190W WO2019014813A1 WO 2019014813 A1 WO2019014813 A1 WO 2019014813A1 CN 2017093190 W CN2017093190 W CN 2017093190W WO 2019014813 A1 WO2019014813 A1 WO 2019014813A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- image
- skin
- block set
- black pixel
- pixel blocks
- Prior art date
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/20—Image preprocessing
- G06V10/28—Quantising the image, e.g. histogram thresholding for discrimination between background and foreground patterns
-
- 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
- G06V10/44—Local feature extraction by analysis of parts of the pattern, e.g. by detecting edges, contours, loops, corners, strokes or intersections; Connectivity analysis, e.g. of connected components
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR 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
- G06V40/161—Detection; Localisation; Normalisation
Definitions
- the present application relates to the field of face recognition technology, and in particular, to a method, an apparatus, and an intelligent terminal for quantitatively detecting a human skin condition parameter.
- Face recognition technology is a technology for identity identification by analyzing and comparing facial visual feature information. Its research fields include: identity recognition, expression recognition and gender recognition. In recent years, people's attention to beauty and skin care has been increasing, and the field of face recognition technology has also proposed some methods for detecting a person's skin condition by recognizing a face in an image based on the demand. For example, to quantitatively detect the severity of the blackhead, the specific implementation manner of the method includes: first converting the original facial image block into a grayscale image, and then binarizing the facial image block by using a local adaptive threshold method. The connected domain analysis method is used to calculate the number of connected domains and their respective areas.
- the selected area is between 7 ⁇ 10 -5 times the original facial image block area Area and 1 ⁇ 10 -3 times the original facial image block area Area.
- the connected domain whose average pixel gray value is less than 180 is detected as a blackhead portion;
- the blackhead severity parameter of the single blackhead portion is (255 - the average pixel gray value of the blackhead portion) ⁇ the area of the blackhead portion, by the blackhead portion
- the inventors have found that at least the following problems exist in the prior art: in the prior art, the skin condition of a face is determined by a severity score formula, and the determination process and the result are for the person to be tested. It is not intuitive enough to be convincing enough.
- the embodiment of the present invention provides a method, a device and a smart terminal for quantitatively detecting a human skin condition parameter, which can solve the prior art process and the result of quantitatively detecting a skin condition of a human face, which is not intuitive enough for the person to be tested, and has insufficient persuasiveness. Strong question.
- an embodiment of the present application provides a method for quantitatively detecting a facial skin mass parameter, including:
- the image processing including binarization processing
- a level of the skin mass parameter is determined based on the set of valid blocks.
- the detecting, in the binary image, all the black pixel blocks corresponding to the skin quality parameter, and recording the effective pixel block includes:
- the skin quality parameter includes a blackhead
- the method for quantitatively detecting a human skin condition parameter is specifically:
- the black pixel block corresponding to the blackhead in the first binary image is detected, Recorded as the first valid block set, including:
- the determining, according to the first active block set, the level of the blackhead includes:
- the rank of the blackhead is determined based on the number.
- the determining, according to the first active block set, the level of the blackhead includes:
- the image processing further comprises an expansion etching process.
- an embodiment of the present application provides a device for quantitatively detecting a facial skin mass parameter, including:
- An image acquisition unit configured to acquire a face image
- An intercepting unit configured to intercept an image of the region corresponding to the skin parameter to be detected in the face image, and record the image as an original image
- An image processing unit configured to perform image processing on the original image to obtain a binary image, where the image processing includes a binarization process
- An edge detecting unit configured to detect all black pixel blocks corresponding to the skin parameters in the binary image, and record them as a valid block set
- a determining unit configured to determine a level of the skin quality parameter according to the effective block set.
- an intelligent terminal including:
- At least one processor and,
- the memory stores instructions executable by the at least one processor, the instructions being executed by the at least one processor to enable the at least one processor to perform quantitative detection of facial skin parameters as described above Methods.
- an embodiment of the present application provides a non-transitory computer readable storage medium, where the non-transitory computer readable storage medium stores computer executable instructions for causing a smart terminal to execute the above The method for quantitatively detecting human skin parameters.
- the embodiment of the present application further provides a computer program product, the computer program product comprising a computer program stored on a non-transitory computer readable storage medium, the computer program comprising program instructions, when the program When the instruction is executed by the smart terminal, the smart terminal is caused to perform the method of quantitatively detecting the facial skin mass parameter as described above.
- the method, device, and intelligent terminal for quantitatively detecting human skin parameters provided by the embodiments of the present application intercept the skin to be detected in the face image when acquiring the face image An image of a region corresponding to the qualitative parameter, and recorded as an original image; then performing image processing including binarization processing on the original image to obtain a binary image; and detecting all of the binary image corresponding to the skin parameter
- the black pixel block is recorded as a set of valid blocks; finally, the level of the skin parameter is determined according to the set of valid blocks, and the skin parameters of the face to be detected can be more intuitively and more targeted. Severity, enhance the persuasiveness of test results, and enhance the user experience.
- FIG. 1 is a schematic flow chart of a method for quantitatively detecting a facial skin mass parameter provided by an embodiment of the present application
- FIG. 2 is an illustration of a binary image for quantitatively detecting a face-lifting head image provided by an embodiment of the present application.
- FIG. 3 is a schematic flow chart of a method for quantitatively detecting a black face of a human face according to an embodiment of the present application
- FIG. 4 is a schematic diagram showing an example of a first original image of a face image provided by an embodiment of the present application
- FIG. 5 is a schematic diagram showing an example of a first binary image of the first original image shown in FIG. 4;
- FIG. 6 is a schematic structural diagram of an apparatus for quantitatively detecting a human skin condition parameter according to an embodiment of the present application
- FIG. 7 is a schematic structural diagram of an intelligent terminal according to an embodiment of the present application.
- the field of face recognition technology has also proposed some methods for detecting a person's skin condition by recognizing a face in an image.
- the existing method for detecting facial skin is mainly used for comprehensively determining the type of skin on the face, such as oily skin, dry skin, neutral skin, mixed skin, etc., through the severity score formula of each skin condition parameter.
- the judgment process and results are not intuitive enough for the person to be tested, and the persuasive power is not strong enough.
- the embodiments of the present application provide a method, a device, and a method for quantitatively detecting a human skin condition parameter.
- Intelligent Terminal wherein the method for quantitatively detecting a human skin condition parameter first extracts an image of a region corresponding to a skin condition parameter to be detected from a face image, and then performs a series of image processing on the image of the region to extract the image.
- the texture characteristics of the skin parameters to be detected, and thus the quantitative determination of the severity of the skin parameters can more accurately and more specifically give the severity of the facial skin parameters.
- the “skin parameter” may be any reference amount related to the skin condition of the face, such as: head-up pattern, crow's feet, laugh lines, acne, freckles, blackheads, pores, melanin precipitation, etc.
- one or more skin parameters can be detected according to different needs of the user.
- the method, device and intelligent terminal for quantitatively detecting facial skin parameters provided by the embodiments of the present application are applicable to any technical field related to face recognition, such as: beauty shooting, especially suitable for the skin care field.
- the beauty application can be developed based on the inventive concept of the method for quantitatively detecting human skin parameters provided by the embodiments of the present application, so that the user can conveniently identify one or more by real-time self-timer or uploading a face image.
- the severity of the skin mass parameter can also set the detection process to be visualized to make the test results more convincing.
- the application can also recommend the most suitable skin care method for the user according to different detection results to enhance the user experience.
- the method provided by the embodiment of the present application can be applied to any smart terminal having an image processing function.
- the smart terminal includes, but is not limited to, a beauty identification device, a personal computer, a tablet computer, a smart phone, and the like.
- the smart terminal can include any suitable type of storage medium for storing data, such as a magnetic disk, a compact disc (CD-ROM), a read-only memory or a random access memory.
- the smart terminal may also include one or more logical computing modules that perform any suitable type of function or operation in parallel, such as viewing a database, image processing, etc., in a single thread or multiple threads.
- the logic operation module may be any suitable type of electronic circuit or chip-type electronic device capable of performing logical operation operations, such as a single core processor, a multi-core processor, a graphics processing unit (GPU), or the like.
- FIG. 1 is a schematic flow chart of a method for quantitatively detecting a facial skin mass parameter provided by an embodiment of the present application. Referring to FIG. 1, the method includes:
- the "face image” refers to an image including the face of the detected person, by which all face features of the detected person can be acquired.
- the face image of the detected person is acquired.
- the specific implementation manner of acquiring the face image may be: collecting the positive face image of the detected person in real time; or: directly capturing the existing image including the positive face of the detected person directly in the smart terminal or in the cloud. .
- different ways of acquiring the face image may be selected.
- the skin quality parameter may be any reference amount related to the skin condition of the face, such as: head-up pattern, crow's feet, laugh lines, acne, freckles, blackheads, pores, melanin precipitation, and the like.
- the “region corresponding to the skin mass parameter to be detected” refers to a region in which the skin parameter is present in the face image, that is, the target detection region. For different skin parameters, different regions can be intercepted as target detection regions. For example, the target detection region corresponding to the headline is the forehead, the target detection region corresponding to the crow's feet is the corner of the eye, and the target detection region corresponding to the acne is the full face.
- the target detection area corresponding to the black head is a nose, and the target detection area corresponding to the pore is a cheek.
- the "original image” refers to a target image to be subjected to image processing, that is, the “original image” is the basis for subsequent image processing.
- the skin image to be detected is first intercepted in the face image.
- the region corresponding to the parameter is used as the target region of the skin parameter detection, and the intercepted image is recorded as the original image to reduce the amount of data processing in the subsequent image processing.
- the range of the area corresponding to the skin parameter to be detected may be different in different face images.
- the specific implementation manner of intercepting the image of the region corresponding to the skin quality parameter to be detected in the face image may also be different.
- the “region corresponding to the skin parameter to be detected” is closely related to the face image to be detected, and the actual image of the region is intercepted.
- the method may be: firstly, all the key points related to the target detection area in the human face are detected by using the method of face key point localization, and then the image corresponding to the area surrounded by the key points is intercepted and recorded as an original image.
- the area shape of the cut original image may be different for different face images.
- the “area corresponding to the skin parameter to be detected” may also be an area range set according to the empirical value, and does not change with the change of the face image.
- the specific embodiment of capturing an image of the region corresponding to the skin parameter to be detected in the face image may be: first acquiring a key point in the face image related to the region The coordinate parameter is then determined based on the coordinate parameter; finally, the image corresponding to the intercepted region is intercepted and recorded as an original image.
- the method for obtaining the coordinate parameters of the key points in the face image may be: first, using a third-party toolkit, such as: dlib, face++, etc., to perform face key points on the face image (eg, eyebrow key points) Positioning the eye key points, facial contour key points, mouth key points, etc., and then selecting the preset key points and determining the coordinate parameters thereof, wherein the preset key points may include one or more; or, The key points preset in the face image may be directly extracted, and the coordinate parameters thereof may be acquired.
- a third-party toolkit such as: dlib, face++, etc.
- the “binarization processing” refers to setting the gray value of the pixel on the image to 0 or 255 according to a preset rule, so that the entire image presents a distinct black and white vision. Effect; the "binary image” refers to a binarized image of the original image obtained after performing image processing including binarization processing on the original image.
- a pixel block having a texture feature of a skin parameter to be measured in the binary image is represented as a black pixel block
- a pixel block having a texture feature having a skin parameter to be measured is represented as a white pixel region.
- the skin parameter to be measured is a head-up pattern
- the binary image is as shown in FIG. 2, then the position of the black pixel block in the binary image is the face image thereof. The position where the headline is located.
- the texture feature of the skin condition parameter to be detected in the original image is mapped by binarizing the original image.
- binarization of an image may be performed according to a certain rule. Then, the image is divided into N windows, and then the pixels in the window are divided into two parts according to a unified threshold T for each of the N windows, and binarization processing is performed.
- the original image is binarized by adaptive threshold binarization processing, which is based on the pixels of the neighborhood block of the pixel. The value distribution determines the binarization threshold at that pixel location, resulting in a better segmentation effect.
- the adaptive Threshold function of the OpenCV may be invoked to perform adaptive threshold binarization processing on the original image, wherein the calculated neighborhood block may be set according to the characteristics of the skin parameters to be detected and the empirical parameters.
- the size and offset value adjustment may be invoked to perform adaptive threshold binarization processing on the original image, wherein the calculated neighborhood block may be set according to the characteristics of the skin parameters to be detected and the empirical parameters. The size and offset value adjustment.
- the image processing described above includes an expansion etching process in addition to the binarization processing. That is, after the original image is binarized, an expansion etching process is also performed.
- the "expansion corrosion treatment” is one of two basic operations of image morphology, which has the function of filling small holes in the object, connecting adjacent objects and smoothing the boundary.
- the “black pixel block corresponding to the skin parameter” may refer to the binary image. Any black pixel block, and the set of all black pixel blocks in the binary image is the active block set.
- a specific implementation manner of detecting all black pixel blocks corresponding to the skin parameters in the binary image may be: directly detecting all black pixel regions in the binary image by using an image edge detection algorithm.
- a block, denoted as a set of blocks, is a set of valid blocks. The size of each black pixel block may be different.
- a noise block having no texture feature of the skin parameter to be detected for example, if the skin parameter to be detected is a head-up pattern, hair may exist in the above-mentioned block set
- a large noise block such as a acne or a scar. Therefore, in order to improve the accuracy of the quantitative detection, in some embodiments, all the black pixel blocks corresponding to the skin parameters in the binary image are detected.
- the specific implementation manner may further be: determining a reference width threshold u based on the binary image; detecting all black pixel blocks in the binary image, as a block set; filtering the according to the reference width threshold u A noise block in the block set obtains all black pixel block sets corresponding to the skin parameters, and is recorded as a valid block set.
- the "reference width threshold” is a criterion for judging a noise block in a set of blocks, and the value thereof is related to the length or width of the binary image itself, and the characteristics of the skin parameters to be detected, for example: for the headline
- the reference width threshold u can be set to 1/40 or 1/80 of the total length W (or the total width in the vertical direction) of the binary image in the horizontal direction. This is mainly because the pixels of the photographing terminal or the distance of the photographing distance are different, and the face pixels which are large in different binary images are small, and it is necessary to give different reference width thresholds for each binary image. It is convenient to subsequently calculate the length of the black pixel block in the horizontal direction and the width in the vertical direction, thereby defining the noise block in the block set.
- the “level of skin quality parameter” refers to the severity of the skin condition parameter of the detected person, which can be divided into multiple levels such as mild, moderate, and severe, and each level corresponds to There is a corresponding range of reference values.
- the "reference value” is a parameter R that can determine the severity of the skin condition parameter
- the division criterion of the reference value range corresponding to each level can be set by experimenting and observing a large number of face images. For example, a batch of face images with statistically significant wrinkles appears to have a parameter R greater than or equal to a certain value. For example, R ⁇ a, the headline level of the face image satisfying R ⁇ a is determined to be severe.
- Wrinkle level statistics of a group of face images whose blackhead severity looks general, whose parameters R belong to a certain range of values, such as: b2>R ⁇ b1, which will satisfy the blackhead of the face image of b2>R ⁇ b1
- the rating is determined to be a moderate blackhead rating.
- the formula can vary depending on the setting of the different "reference values".
- the setting of the "reference value” can be determined by the characteristics of the skin parameters to be detected. For example, for headstrip detection, its “reference value” can be set to the total length of all black pixel blocks in the active block set in the horizontal direction; then, in actual application, all the valid block sets are calculated. The total length of the black pixel block in the horizontal direction and compare it with the empirical value to obtain the heading level.
- the method for quantitatively detecting the facial skin parameters provided by the embodiments of the present application can be used for quantitative detection of any skin type parameter.
- the embodiment of the present application further provides a method for quantitatively detecting a black face of a human face, to further describe a method for quantitatively detecting a human skin condition parameter provided by an embodiment of the present application.
- FIG. 3 is a schematic flow chart of a method for quantitatively detecting a black face of a human face according to an embodiment of the present application. Referring to FIG. 3, the method includes:
- the step 310 has the same technical features as the step 110 shown in FIG. 1, and the specific embodiment is also applicable to the embodiment, and therefore, it will not be described in detail in this embodiment.
- the nose region is used as the target detection region corresponding to the blackhead.
- the specific implementation manner of intercepting the image of the nose region in the face image may be: first, using the third-party toolkit to perform the face key point positioning on the acquired face image, and then selecting the preset nose key point. And determining its coordinate parameter, and then determining the nose region in the face image based on the coordinate parameter.
- the third-party toolkit dlib is used to locate the key points of the nose and obtain the key points of the nose including the key points 28-36; it can be pre-specified: the value of the x-axis coordinate of the key point 32 is taken as x1, and the key point 36 The value of the x-axis coordinate is taken as x2, the y-axis coordinate value of the key point 29 is taken as y1, and the intermediate value of the y-axis coordinate value of the key point 31 and the key point 34 is taken as y2, and the coordinate points (x1, y1), (x1) , the area enclosed by y2), (x2, y1), and (x2, y2) is the "nose area"; then, after obtaining the face image, the key points 29, 31, 32 in the face image are first acquired.
- the image corresponding to the domain is recorded as the first original image.
- 4 is an exemplary schematic diagram of a first original image of a face image obtained by the above manner.
- coordinate parameters of other nose key points may also be selected as criteria for dividing the forehead area; or, other division methods may be used (eg, specifying key points 36)
- the y-axis coordinate is used as the y2) to divide the nose region; as long as the same key point and the division manner are used to locate the nose region for different facial images, the embodiment of the present application does not specifically limit this.
- the nose region of the face image can be determined only by acquiring the coordinate parameters of the nose key point, which improves the efficiency of intercepting the first original image.
- the Adaptive Threshold function of OpenCV can be invoked to perform adaptive threshold binarization processing on the first original image as shown in FIG. 4, wherein the calculated neighborhood block can be set according to the characteristics of the blackhead and the empirical parameters.
- the size (“block size”) is 9; and the offset value adjustment (parameter "C") is 5.
- the first binary image obtained by the above processing is as shown in Fig. 5(a). Further, in order to remove the fine noise points, the obtained first binary image highlights the shape of the blackhead.
- the image processing may further include an expansion etching process, that is, performing an expansion etching process after performing adaptive threshold binarization processing on the first original image.
- the first binary image obtained after performing image processing including adaptive threshold binarization processing and expansion etching processing on the first original image as shown in FIG. 4 is as shown in FIG. 5(b).
- the first binary image in the process of performing quantitative detection of the serious condition of the face blackhead, may also be presented to the subject at the same time, so that the detected person clearly sees the specificity of the blackhead. Distribution, which makes the test results more intuitive and more convincing.
- the specificity of detecting the black pixel block corresponding to the blackhead in the first binary image is: determining a first reference width threshold u1 based on the first binary image; detecting the All black pixel blocks in the first binary image are recorded as a first block set; the first noise block in the first block set is filtered according to the first reference width threshold u1, and all and the The black pixel block corresponding to the blackhead is recorded as the first valid block set.
- the first reference width threshold u1 may be set to 1/100 or other suitable multiple of the total length W1 (or the total width in the vertical direction) of the first binary image in the horizontal direction. According to the characteristics of the blackhead (the blackhead is generally similar to a circle), it is possible to define a black pixel region in which the length w1 in the horizontal direction is smaller than u1 or w1 is larger than 4*u1, the width h1 in the vertical direction is larger than 3*w1, or h1 is smaller than w1/3.
- the block is a first noise block, whereby black pixels with w1 less than u1 and/or w1 greater than 4*u1 and/or h1 greater than 3*w1 and/or h1 less than w1/3 are filtered out in the first block set.
- the block can obtain the first set of valid blocks.
- the first noise block defining the blackhead here is that the black pixel block with w1 less than u1 and/or w1 greater than 4*u1 and/or h1 greater than 3*w1 and/or h1 less than w1/3 is only
- the embodiments of the present application are not intended to limit the embodiments of the present application. In other practical applications, the first noise block of the blackhead may be defined by other conditions.
- the "reference value" for determining the severity of the blackhead can be set to the number N of black pixel blocks in the first active block set
- the specific implementation manner of determining the level of the blackhead according to the first valid block set may be: calculating the number N of black pixel blocks in the first valid block set; determining, according to the number N The level of blackheads.
- N ⁇ 2 when N ⁇ 2, it is divided into no blackhead level; when 20 ⁇ N>2, it is divided into mild blackhead level; when 50 ⁇ N> At 20 o'clock, it is divided into medium blackhead level; when N>50, it is divided into severe blackhead level. Then, if the number N of calculations for the first binary image of a certain subject is 70, it can be determined that the blackhead level of the subject is a severe blackhead level.
- the ratio of the total area of all the black pixel blocks in the first active block set to the total area of the first binary image may be used as the “reference value”.
- the specific implementation manner of determining the level of the blackhead by the first active block set may also be: Describe a total area of all black pixel blocks in the first active block set, and record it as a first total area; determine the blackhead according to a ratio of the first total area to a total area of the first binarized image The level.
- the ratio of the number of all black pixel blocks in the first active block set to the total area of the first binary image may be used as a “reference value”, and then determined according to the “reference value”.
- the blackhead level is not listed here.
- the method for quantitatively detecting human skin parameters can also quantitatively detect other skin parameters such as pores, crow's feet, and acne. It differs from the quantitative detection of blackheads or heads up in that different skin parameters correspond to different target detection areas, and the parameters used in binarization and the definition of noise blocks are also based on The properties of the detected skin parameters vary accordingly.
- the target detection area is the cheeks.
- the parameter “block Size” may be selected.
- the method of the present invention provides a method for quantitatively detecting a human face skin parameter provided by the embodiment of the present application.
- the face image is acquired, the face image is intercepted and the image to be detected is intercepted.
- An image corresponding to the skin condition parameter and recorded as an original image; then performing image processing including binarization processing on the original image to obtain a binary image; and detecting all of the binary image corresponding to the skin parameter
- the black pixel block is recorded as a valid block set; finally, the level of the skin parameter is determined according to the effective block set, which can more accurately and more specifically give a serious face parameter to be detected.
- FIG. 6 is a schematic structural diagram of an apparatus for quantitatively detecting a human skin condition parameter according to an embodiment of the present application.
- the apparatus 6 includes:
- An image obtaining unit 61 configured to acquire a face image
- the intercepting unit 62 is configured to intercept an image of the region corresponding to the skin parameter to be detected in the face image, and record it as an original image;
- An image processing unit 63 configured to perform image processing on the original image to obtain a binary image, wherein the image processing includes a binarization process, or the image processing includes a binarization process and an erosion process;
- the edge detecting unit 64 is configured to detect all black pixel blocks corresponding to the skin parameters in the binary image, and record them as a valid block set;
- the determining unit 65 is configured to determine a level of the skin quality parameter according to the effective block set.
- the image acquisition unit 61 acquires the face image
- the image of the region corresponding to the skin parameter to be detected in the face image is first intercepted in the clipping unit 62, and recorded as an original image.
- the original image is subjected to image processing by the image processing unit 63 to obtain a binary image, and all the black pixel blocks corresponding to the skin parameters in the binary image are detected by the edge detecting unit 84, and are marked as effective.
- the block set; finally the determining unit 84 determines the heading level of the face image according to the valid block set.
- the edge detecting unit 64 includes a reference width threshold determining module 641, a detecting module 642, and a noise removing module 643.
- the reference width threshold determining module 641 determines a reference width threshold based on the binary image
- the detecting module 642 detects all black pixel blocks in the binary image, and records it as a block set.
- the noise block in the block set is filtered according to the reference width threshold by the noise removing module 643, and all black pixel blocks corresponding to the skin parameter are obtained, which are recorded as a valid block set.
- the intercepting unit 62 is specifically configured to: intercept an image of a nose region in the face image, and record it as a first original image.
- the image processing unit 63 is specifically configured to: perform image processing on the first original image to obtain a first binary image; and the edge detecting unit 64 is configured to: detect all blackheads in the first binary image.
- the black pixel block is recorded as the first active block set; the determining unit 65 is specifically configured to: determine the level of the blackhead according to the first valid block set.
- the reference width threshold determining module 641 determines based on the first binary image a first reference width threshold
- the detection module 642 detects all black pixel blocks in the first binary image, and records them as a first block set; and then filters the first reference width threshold according to the first noise reduction threshold by the denoising module 643.
- the first noise block in the first block set obtains all black pixel blocks corresponding to the black head, and is recorded as the first valid block set.
- the determining unit 65 is specifically configured to: calculate a number of black pixel blocks in the first valid block set; determine a level of the blackhead according to the number; or calculate the first valid block set
- the total area of all black pixel blocks is recorded as the first total area; the level of the blackhead is determined according to the ratio of the first total area to the total area of the first binarized image.
- the device for quantitatively detecting the facial skin parameters intercepts the intercepting unit 62 by acquiring the facial image when the image acquiring unit 61 acquires the facial image.
- An image of a region corresponding to the skin condition parameter to be detected in the face image is recorded as an original image; and then the image processing unit 63 performs image processing including binarization processing to obtain a binary image;
- the edge detecting unit 64 detects all the black pixel blocks corresponding to the skin parameters in the binary image, and records them as a valid block set; finally, the determining unit 65 determines the skin quality according to the effective block set.
- the level of the parameter can more accurately and more specifically give the severity of the skin parameters of the face to be detected, enhance the persuasiveness of the test results, and enhance the user experience.
- FIG. 7 is a schematic structural diagram of an intelligent terminal according to an embodiment of the present disclosure.
- the smart terminal 700 can be any type of electronic device, such as a mobile phone, a tablet computer, a beauty authentication device, etc. Referring to FIG. 7, the smart terminal 700 is shown in FIG. include:
- processors 710 and memory 720 one processor 710 is taken as an example in FIG.
- the processor 710 and the memory 720 may be connected by a bus or other means, as exemplified by a bus connection in FIG.
- the memory 720 is a non-transitory computer readable storage medium for storing non-transitory software a program, a non-transitory computer executable program, and a module, such as the program instruction/module corresponding to the method for quantitatively detecting a human skin parameter in the embodiment of the present application (for example, the image acquisition unit 61 shown in FIG. 6 , intercepting Unit 62, image processing unit 63, edge detection unit 64, and decision unit 65).
- the processor 710 performs various functional applications and data processing of the apparatus for quantitatively detecting facial skin parameters by running non-transitory software programs, instructions, and modules stored in the memory 720, that is, implementing any of the above method embodiments.
- a method for quantitatively detecting facial skin parameters is performed by running non-transitory software programs, instructions, and modules stored in the memory 720, that is, implementing any of the above method embodiments.
- the memory 720 can include a storage program area and a storage data area, wherein the storage program area can store an operating system, an application required for at least one function; the storage data area can be stored according to the use of the device for quantitatively detecting a human skin parameter. Data, etc.
- memory 720 can include high speed random access memory, and can also include non-transitory memory, such as at least one magnetic disk storage device, flash memory device, or other non-transitory solid state storage device.
- the memory 720 can optionally include a memory remotely located relative to the processor 710 that can be connected via a network to a device that quantitatively detects human skin parameters. Examples of such networks include, but are not limited to, the Internet, intranets, local area networks, mobile communication networks, and combinations thereof.
- the one or more modules are stored in the memory 720, and when executed by the one or more processors 710, perform a method of quantitatively detecting a human face skin parameter in any of the above method embodiments, for example, performing The method steps 110 to 150 of FIG. 1 described above, and the method steps 310 to 350 of FIG. 3, implement the functions of the units 61-65 of FIG.
- the embodiment of the present application further provides a non-transitory computer readable storage medium storing computer executable instructions executed by one or more processors, for example,
- a processor 710 of FIG. 7 may be configured to cause the one or more processors to perform the method of quantitatively detecting a human skin condition parameter in any of the above method embodiments, for example, performing the method step 110 in FIG. 1 described above to Step 150, method step 310 to step 350 in FIG. 3, implements the functions of units 61-65 in FIG.
- the device embodiments described above are merely illustrative, wherein the units described as separate components may or may not be physically separate, and the components displayed as units may or may not be physical units, ie may be located A place, or it can be distributed to multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of the embodiment.
- the storage medium may be a magnetic disk, an optical disk, a read-only memory (ROM), or a random access memory (RAM).
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Multimedia (AREA)
- Theoretical Computer Science (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Oral & Maxillofacial Surgery (AREA)
- Human Computer Interaction (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Measuring And Recording Apparatus For Diagnosis (AREA)
- Measurement Of The Respiration, Hearing Ability, Form, And Blood Characteristics Of Living Organisms (AREA)
Abstract
本申请实施例提供了一种定量检测人脸肤质参量的方法、装置和智能终端。其中,所述方法包括:获取人脸图像;截取所述人脸图像中与待检测的肤质参量对应的区域的图像,并记为原始图像;对所述原始图像进行图像处理获得二值图像,所述图像处理包括二值化处理;检测出所述二值图像中所有与所述肤质参量对应的黑色像素区块,记为有效区块集合;根据所述有效区块集合确定所述肤质参量的等级。通过上述技术方案,本申请实施例能够更加直观并且更加有针对性地给出待检测的人脸肤质参量的严重程度,增强检测结果的说服力,提升用户体验。
Description
本申请涉及人脸识别技术领域,尤其涉及一种定量检测人脸肤质参量的方法、装置和智能终端。
人脸识别技术是一种通过分析比较人脸视觉特征信息进行身份鉴定的技术,其研究领域包括:身份识别、表情识别以及性别识别等。近年来,人们对美容护肤的关注度日益增强,人脸识别技术领域目前也基于该需求提出了一些通过识别图像中的人脸检测出一个人的皮肤肤质情况的方法。其中,以定量检测黑头的严重程度为例,该方法的具体实施方式包括:首先将原始面部图像区块转化为灰度图,然后使用局部自适应阈值法将面部图像区块中二值化,用连通域分析方法计算得到其中连通域数量和各自的面积,筛选出面积介于7×10-5倍原始面部图像区块面积Area与1×10-3倍原始面部图像区块面积Area之间且平均像素灰度值小于180的连通域检测为黑头部分;单个黑头部分的黑头严重程度参量为(255-该黑头部分的平均像素灰度值)×该黑头部分的面积,由各个黑头部分的黑头严重程度参量相加再除以原始面部图像区块面积Area得到总黑头严重程度参量h,再计算总黑头严重程度分值Blackhead=log1.4(2.5×h+0.4)。
然而,在实现本申请过程中,发明人发现现有技术中至少存在如下问题:现有技术中通过严重程度分值公式来判定人脸的肤质状况,该判定过程及结果对于待检测者来说不够直观,说服力不够强。
发明内容
本申请实施例提供一种定量检测人脸肤质参量的方法、装置和智能终端,能够解决现有技术定量检测人脸肤质状况的过程及结果对于待检测者来说不够直观,说服力不够强的问题。
第一方面,本申请实施例提供了一种定量检测人脸肤质参量的方法,包括:
获取人脸图像;
截取所述人脸图像中与待检测的肤质参量对应区域的图像,并记为原始图像;
对所述原始图像进行图像处理获得二值图像,所述图像处理包括二值化处理;
检测出所述二值图像中所有与所述肤质参量对应的黑色像素区块,记为有效区块集合;
根据所述有效区块集合确定所述肤质参量的等级。
可选地,所述检测出所述二值图像中所有与所述肤质参量对应的黑色像素区块,记为有效区块集合,包括:
基于所述二值图像确定基准宽度阈值;
检测出所述二值图像中所有黑色像素区块,记为区块集合;
根据所述基准宽度阈值过滤所述区块集合中的噪音区块,获得所有与所述肤质参量对应的黑色像素区块,记为有效区块集合。
可选地,所述肤质参量包括黑头,则,所述定量检测人脸肤质参量的方法,具体为:
截取所述人脸图像中鼻子区域的图像,并记为第一原始图像;
对所述第一原始图像进行图像处理获得第一二值图像,所述图像处理包括二值化处理;
检测出所述第一二值图像中所有与黑头对应的黑色像素区块,记为第一有效区块集合;
根据所述第一有效区块集合确定所述黑头的等级。
可选地,所述检测出所述第一二值图像中所有与黑头对应的黑色像素区块,
记为第一有效区块集合,包括:
基于所述第一二值图像确定第一基准宽度阈值;
检测出所述第一二值图像中所有黑色像素区块,记为第一区块集合;
根据所述第一基准宽度阈值过滤所述第一区块集合中的第一噪音区块,获得所有与所述黑头对应的黑色像素区块,记为第一有效区块集合。
可选地,所述根据所述第一有效区块集合确定所述黑头的等级,包括:
计算所述第一有效区块集合中黑色像素区块的个数;
根据所述个数确定所述黑头的等级。
可选地,所述根据所述第一有效区块集合确定所述黑头的等级,包括:
计算所述第一有效区块集合中所有黑色像素区块的总面积,并记为第一总面积;
根据所述第一总面积与所述第一二值化图像的总面积的比值确定所述黑头的等级。
可选地,所述图像处理还包括膨胀腐蚀处理。
第二方面,本申请实施例提供一种定量检测人脸肤质参量的装置,包括:
图像获取单元,用于获取人脸图像;
截取单元,用于截取所述人脸图像中与待检测的肤质参量对应的区域的图像,并记为原始图像;
图像处理单元,用于对所述原始图像进行图像处理获得二值图像,所述图像处理包括二值化处理;
边缘检测单元,用于检测出所述二值图像中所有与所述肤质参量对应的黑色像素区块,记为有效区块集合;
判定单元,用于根据所述有效区块集合确定所述肤质参量的等级。
第三方面,本申请实施例提供一种智能终端,包括:
至少一个处理器;以及,
与所述至少一个处理器通信连接的存储器;其中,
所述存储器存储有可被所述至少一个处理器执行的指令,所述指令被所述至少一个处理器执行,以使所述至少一个处理器能够执行如上所述的定量检测人脸肤质参量的方法。
第四方面,本申请实施例提供一种非暂态计算机可读存储介质,所述非暂态计算机可读存储介质存储有计算机可执行指令,所述计算机可执行指令用于使智能终端执行如上所述的定量检测人脸肤质参量的方法。
第五方面,本申请实施例还提供了一种计算机程序产品,所述计算机程序产品包括存储在非暂态计算机可读存储介质上的计算机程序,所述计算机程序包括程序指令,当所述程序指令被智能终端执行时,使所述智能终端执行如上所述的定量检测人脸肤质参量的方法。
本申请实施例的有益效果在于:本申请实施例提供的定量检测人脸肤质参量的方法、装置和智能终端通过在获取到人脸图像时,截取所述人脸图像中与待检测的肤质参量对应的区域的图像,并记为原始图像;然后对所述原始图像进行包括二值化处理的图像处理获得二值图像;检测出所述二值图像中所有与所述肤质参量对应的黑色像素区块,记为有效区块集合;最后根据所述有效区块集合确定所述肤质参量的等级,能够更加直观并且更加有针对性地给出待检测的人脸肤质参量的严重程度,增强检测结果的说服力,提升用户体验。
一个或多个实施例通过与之对应的附图中的图片进行示例性说明,这些示例性说明并不构成对实施例的限定,附图中具有相同参考数字标号的元件表示为类似的元件,除非有特别申明,附图中的图不构成比例限制。
图1是本申请实施例提供的一种定量检测人脸肤质参量的方法的流程示意图;
图2是本申请实施例提供的一种用于定量检测人脸抬头纹的二值图像的示
例示意图;
图3是本申请实施例提供的一种定量检测人脸黑头的方法的流程示意图;
图4是本申请实施例提供的一个人脸图像的第一原始图像的示例示意图;
图5是图4所示的第一原始图像的第一二值图像的示例示意图;
图6是本申请实施例提供的一种定量检测人脸肤质参量的装置的结构示意图;
图7是本申请实施例提供的一种智能终端的结构示意图。
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。
需要说明的是,如果不冲突,本申请实施例中的各个特征可以相互结合,均在本申请的保护范围之内。另外,虽然在装置示意图中进行了功能模块划分,在流程图中示出了逻辑顺序,但是在某些情况下,可以以不同于装置中的模块划分,或流程图中的顺序执行所示出或描述的步骤。再者,本申请所采用的“第一”“第二”“第三”“第四”等字样并不对数据和执行次序进行限定,仅是对功能和作用基本相同的相同项或相似项进行区分。
近年来,随着生活水平的提高,人们对美容护肤的关注度日益增强。然而,当代城市生活节奏较快,人们鲜有时间特意到美容院或美容机构做肤质检测以及进行相应的护肤保养。为了方便人们自行进行肤质检测,人脸识别技术领域目前也提出了一些通过识别图像中的人脸检测出一个人的皮肤肤质情况的方法。但现有的人脸肤质检测方法主要用于综合判定人脸皮肤的类型,如:油性皮肤、干性皮肤、中性皮肤、混合性皮肤等,通过各肤质参量的严重程度分值公式来判定人脸的肤质状况,其判定过程及结果对于待检测者来说不够直观,说服力不够强。
基于此,本申请实施例提供了一种定量检测人脸肤质参量的方法、装置和
智能终端。其中,该定量检测人脸肤质参量的方法通过首先从人脸图像中截取出与待检测的肤质参量对应的区域的图像,然后对该区域的图像进行一系列的图像处理以提取出该待检测的肤质参量的纹理特征,进而定量化地识别出该肤质参量的严重程度,能够更加直观并且更具针对性地给出该人脸肤质参量的严重程度。其中,在本申请实施例中,所述“肤质参量”可以是任意与人脸皮肤状况相关的参考量,如:抬头纹、鱼尾纹、笑纹、痘痘、雀斑、黑头、毛孔、黑色素沉淀等,在本申请实施例中,可以根据用户的不同需求针对其中的一种或者多种肤质参量进行检测。
本申请实施例提供的定量检测人脸肤质参量的方法、装置和智能终端适用于任意与人脸识别相关的技术领域,如:美颜拍摄,尤其适用于美容护肤领域。例如:可以基于本申请实施例提供的定量检测人脸肤质参量的方法的发明构思开发美容类的应用程序,使用户可以方便地通过实时自拍或者上传人脸图像的方法鉴定自己某一或者多种肤质参量的严重程度。其中,该应用程序也可以将检测过程设置为可视化的,以使检测结果更有说服力。进一步地,该应用程序还可以根据不同的检测结果为用户推荐最合适的护肤方法,以提升用户体验。
本申请实施例提供的方法能够应用于任意具有图像处理功能的智能终端。所述智能终端包括但不限于:美容鉴定仪器、个人电脑、平板电脑、智能手机等。该智能终端可以包括任何合适类型的,用以存储数据的存储介质,例如磁碟、光盘(CD-ROM)、只读存储记忆体或随机存储记忆体等。该智能终端还可以包括一个或者多个逻辑运算模块,单线程或者多线程并行执行任何合适类型的功能或者操作,例如查看数据库、图像处理等。所述逻辑运算模块可以是任何合适类型的,能够执行逻辑运算操作的电子电路或者贴片式电子器件,例如:单核心处理器、多核心处理器、图形处理器(GPU)等。
具体地,下面结合附图,对本申请实施例作进一步阐述。
图1是本申请实施例提供的一种定量检测人脸肤质参量的方法的流程示意图,请参阅图1,该方法包括:
110、获取人脸图像。
在本实施例中,所述“人脸图像”是指包括被检测人的正脸的图像,通过该人脸图像能够获取到该被检测人的所有面部特征。
在本实施例中,当接收到人脸肤质参量检测命令时(比如:用户点击检测黑头或者抬头纹或者笑纹或者毛孔),获取被检测人的人脸图像。其中,获取人脸图像的具体实施方式可以是:实时采集被检测人的正脸图像;或者,也可以是:直接在智能终端本地或云端调取已有的包括被检测人的正脸的图像。针对不同的应用场景或者被检测人的选择,可以选择不同的获取人脸图像的方式。
120、截取所述人脸图像中与待检测的肤质参量对应的区域的图像,并记为原始图像。
在本实施例中,所述“肤质参量”可以是任意与人脸皮肤状况相关的参考量,如:抬头纹、鱼尾纹、笑纹、痘痘、雀斑、黑头、毛孔、黑色素沉淀等等。所述“与待检测的肤质参量对应的区域”是指在人脸图像中存在该肤质参量的区域,也即:目标检测区域。针对不同的肤质参量可以截取不同的区域作为目标检测区域,例如,与抬头纹对应的目标检测区域为额头,与鱼尾纹对应的目标检测区域为眼角,与痘痘对应的目标检测区域为全脸,与黑头对应的目标检测区域为鼻子,与毛孔对应的目标检测区域为脸颊等。所述“原始图像”是指待进行图像处理的目标图像,也就是说,所述的“原始图像”是进行后续图像处理的基础。
由于在实际应用中,每一肤质参量都有其特定的目标检测区域,因此,在本实施例中,在获取到人脸图像之后,首先截取所述人脸图像中与待检测的肤质参量对应的区域作为该肤质参量检测的目标区域,并将截取到的图像记为原始图像,以减少后续图像处理过程中的数据处理量。其中,由于每个人的脸型有可能会有所差别,因此,在不同的人脸图像中,与待检测的肤质参量对应的区域的范围有可能会不同。而对应于不同的实际应用需求,截取所述人脸图像中与待检测的肤质参量对应的区域的图像的具体实施方式也会有所差异。
在一些实施例中,为了获得更加精确的检测结果,所述“与待检测的肤质参量对应的区域”与待检测的人脸图像密切相关,截取该区域的图像的具体实
施方式可以是:首先利用人脸关键点定位的方法检测出人脸中与该目标检测区域相关的全部关键点,然后截取出这些关键点所包围的区域对应的图像,并记为原始图像,在该实施例中,对于不同的人脸图像,截取出的原始图像的区域形状有可能会有所差异。
而在另一些实施例中,为了提升获取原始图像的效率,所述“与待检测的肤质参量对应的区域”也可以是一根据经验值设置的区域范围,不随人脸图像的改变而变化。则,在该实施例中,截取所述人脸图像中与待检测的肤质参量对应的区域的图像的具体实施方式可以是:首先获取所述人脸图像中与该区域相关的关键点的坐标参数;然后基于该坐标参数确定截取区域;最后截取该截取区域对应的图像,并记为原始图像。其中,获取所述人脸图像中的关键点的坐标参数的方式可以是:首先利用第三方工具包,如:dlib、face++等,对该人脸图像进行人脸关键点(如:眉毛关键点、眼睛关键点、面部轮廓关键点、嘴巴关键点等)定位,然后选出预设的关键点并确定其坐标参数,其中,所述预设的关键点可以包括一个或者多个;或者,也可以是直接提取所述人脸图像中预设的关键点,并获取其坐标参数。
130、对所述原始图像进行图像处理获得二值图像,所述图像处理包括二值化处理。
在本实施例中,所述“二值化处理”是指按照预设的规则将图像上的像素点的灰度值设置为0或255,使整个图像呈现出明显的只有黑和白的视觉效果;所述“二值图像”是指对原始图像进行包括二值化处理的图像处理之后获得的该原始图像的二值化图像。在理想状态下,该二值图像中具有待测肤质参量的纹理特征的像素区块表现为黑色像素区块,而不具备待测肤质参量的纹理特征的像素区块表现为白色像素区块。比如:在一实施例中,待测的肤质参量为抬头纹,其二值图像如图2所示,则,在该二值图像中的黑色像素区块所在的位置即其人脸图像的抬头纹所在的位置。
在本实施例中,通过对原始图像进行二值化处理映射出原始图像中待检测的肤质参量的纹理特征。一般地,对图像进行二值化处理可以是按照一定的规
则将该图像划分为N个窗口,然后对这N个窗口中的每一个窗口再按照一个统一的阈值T将该窗口内的像素划分为两部分,进行二值化处理。然而,仅仅通过设定固定阈值很难达到理想的分割效果,因此,在本实施例中,采用自适应阈值二值化处理对原始图像进行二值化处理,其根据像素的邻域块的像素值分布来确定该像素位置上的二值化阈值,能够获得更好的分割效果。具体地,在本实施例中,可以调用OpenCV的adaptive Threshold函数对原始图像进行自适应阈值二值化处理,其中,可以根据待检测的肤质参量的特点以及经验参数设置所计算的邻域块的大小以及偏移值调整量。
此外,由于人脸的皮肤从细微的角度看是不均衡的,所以对原始图像进行二值化处理之后会留下一些形状比较小且分布较为离散,与待检测的肤质参量的形状有较大区别的细小噪音点。因此,为了去除这些由二值化处理后留下的细小的噪音点,提高识别精度,在一些实施例中,上述图像处理除了包括二值化处理外,还包括膨胀腐蚀处理。即,在对原始图像进行二值化处理后,还进行膨胀腐蚀处理。其中,所述“膨胀腐蚀处理”是图像形态学的两个基本操作之一,其具有填充物体内细小空洞,连接邻近物体和平滑边界的作用。
此外,应当理解的是,在本申请实施例中仅对原始图像中的图像内容进行后续的图像处理,而不改变其本身的形状和大小,因此,在本申请实施例中原始图像与其二值图像具有相同的尺寸。
140、检测出所述二值图像中所有与所述肤质参量对应的黑色像素区块,记为有效区块集合。
由于在二值图像中以黑色像素区块表征人脸图像中的肤质参量,因此,在本实施例中,“与所述肤质参量对应的黑色像素区块”可以是指二值图像中任意一个黑色像素区块,而二值图像中所有的黑色像素区块的集合即有效区块集合。在该情况下,检测出所述二值图像中所有与所述肤质参量对应的黑色像素区块的具体实施方式可以是:直接利用图像边缘检测算法检测出二值图像中所有的黑色像素区块,记为区块集合,该区块集合即有效区块集合。其中,每一黑色像素区块的大小可以不一样。
然而,在上述的区块集合中有可能存在不具有待检测的肤质参量的纹理特征的噪音区块(比如,待检测的肤质参量为抬头纹,则在上述区块集合中可能存在头发、痘痘、伤疤等较大的噪音区块),因此,为了提高定量检测的精度,在一些实施例中,检测出所述二值图像中所有与所述肤质参量对应的黑色像素区块的具体实施方式还可以是:基于所述二值图像确定基准宽度阈值u;检测出所述二值图像中所有黑色像素区块,记为区块集合;根据所述基准宽度阈值u过滤所述区块集合中的噪音区块,获得所有与所述肤质参量对应的黑色像素区块集合,记为有效区块集合。
其中,所述“基准宽度阈值”是判断区块集合中的噪音区块的标准,其值与二值图像本身的长度或者宽度,以及待检测的肤质参量的特性相关,例如:对于抬头纹的检测,该基准宽度阈值u可以设置为二值图像在水平方向上的总长W(或者垂直方向上的总宽)的1/40或者1/80。这主要是因为拍摄终端的像素或者拍摄距离的远近不同导致在不同的二值图像中有的脸像素大有的脸像素小,需要针对每一张二值图像给出不同的基准宽度阈值,以方便后续计算黑色像素区块在水平方向上的长度和垂直方向上的宽度,进而定义区块集合中的噪音区块。
150、根据所述有效区块集合确定所述肤质参量的等级。
在本实施例中,所述“肤质参量的等级”是指被检测人该肤质参量的严重程度,其可以分为轻度、中度和重度等多个等级,而每一等级都对应有相应的参考值范围。其中,所述“参考值”是一个可以判定该肤质参量的严重程度的参数R,而每一等级对应的参考值范围的划分标准可以通过对大量人脸图像进行实验和观察来设定。比如:统计皱纹程度看起来非常严重的一批人脸图像,其参数R都大于或等于某一数值,如:R≥a,则将满足R≥a的人脸图像的抬头纹等级确定为重度皱纹等级;统计黑头严重程度看起来一般的一批人脸图像,其参数R都属于某一数值范围内,如:b2>R≥b1,则将满足b2>R≥b1的人脸图像的黑头等级确定为中度黑头等级。
其中,所述根据所述有效区块集合确定所述肤质参量的等级的具体实施方
式可以根据不同的“参考值”的设定而有所不同。而该“参考值”的设定可以由待检测的肤质参量的特点决定。例如:对于抬头纹检测,可以将其“参考值”设定为有效区块集合中所有黑色像素区块在水平方向上的总长度;则,在实际应用时,计算出有效区块集合中所有黑色像素区块在水平方向上的总长度,并将其与经验值进行比较,即可得到其抬头纹等级。
在本申请实施例中,只要选用合适的参数,应用本申请实施例提供的定量检测人脸肤质参量的方法可以对任意一种肤质参量进行定量化检测。具体地,本申请实施例还提供了一种定量检测人脸黑头的方法,以进一步对本申请实施例提供的定量检测人脸肤质参量的方法进行详细说明。
图3是本申请实施例提供的一种定量检测人脸黑头的方法的流程示意图,请参阅图3,该方法包括:
310、获取人脸图像。
在本实施例中,步骤310与图1中所示的步骤110具有相同的技术特征,其具体实施方式同样适用于本实施例,因此,在本实施例中便不再赘述。
320、截取所述人脸图像中鼻子区域的图像,并记为第一原始图像。
由于人脸的黑头多集中于鼻子的区域,因此,在本实施例中,以鼻子区域作为与黑头对应的目标检测区域。具体地,截取所述人脸图像中鼻子区域的图像的具体实施方式可以是:首先利用第三方工具包对获取到的人脸图像进行人脸关键点定位,然后选出预设的鼻子关键点并确定其坐标参数,进而基于该坐标参数确定人脸图像中的鼻子区域。
举例说明:假设利用第三方工具包dlib进行人脸关键点定位之后得到鼻子关键点包括关键点28-36;可以预先规定:以关键点32的x轴坐标的值作为x1,以关键点36的x轴坐标的值作为x2,以关键点29的y轴坐标值作为y1,以关键点31和关键点34的y轴坐标值的中间值作为y2,而坐标点(x1,y1)、(x1,y2)、(x2,y1)和(x2,y2)所围成的区域即所述“鼻子区域”;则,当获得人脸图像之后,首先获取人脸图像中关键点29、31、32、34和36的坐标参数,然后基于这些坐标参数确定该人脸图像的鼻子区域,最后截取出该鼻子区
域对应的图像,并记为第一原始图像。其中,图4即通过上述方式获取到的一个人脸图像的第一原始图像的示例示意图。
应当理解的是,在实际应用中,也可以选用其他鼻子关键点(如:33、35等)的坐标参数作为划分额头区域的标准;或者,还可以采用其他划分方式(如:规定关键点36的y轴坐标作为y2)划分鼻子区域;只要对不同的人脸图像采用相同的关键点以及划分方式来定位鼻子区域即可,本申请实施例对此不作具体限定。在该实施例中,只需获取鼻子关键点的坐标参数即可确定该人脸图像的鼻子区域,提升了截取第一原始图像的效率。
330、对所述第一原始图像进行图像处理获得第一二值图像,所述图像处理包括二值化处理。
在本实施例中,可以调用OpenCV的adaptive Threshold函数对如图4所示的第一原始图像进行自适应阈值二值化处理,其中,可以根据黑头的特点以及经验参数设置所计算的邻域块的大小(“block size”)为9;以及偏移值调整量(参数“C”)为5。通过上述处理获得的第一二值图像如图5(a)所示。进一步地,为了去掉细小的噪音点,使得获得的第一二值图像凸显出黑头的形状。在本实施例中,所述图像处理还可以包括膨胀腐蚀处理,即,在对上述第一原始图像进行自适应阈值二值化处理之后还进行膨胀腐蚀处理。对如图4所示的第一原始图像进行包括自适应阈值二值化处理和膨胀腐蚀处理的图像处理之后获得的第一二值图像如图5(b)所示。
此外,在一些实施例中,在进行定量检测人脸黑头的严重情况的过程中,还可以同时将该第一二值图像呈现给被检测者,以使被检测者清楚看到其黑头的具体分布情况,进而使检测结果更加直观以及更具说服力。
340、检测出所述第一二值图像中所有与黑头对应的黑色像素区块,记为第一有效区块集合。
在理想状态下,第一二值图像中所有的黑色像素区块均为鼻子区域中黑头的映射,然而,在实际情况下,在人脸鼻子区域中可能还会存在细小的斑点、残留痘印、细纹等不属于黑头的噪音点,因此,在本实施例中,为了提高黑头
检测的准确度,检测出所述第一二值图像中所有与黑头对应的黑色像素区块的具体实施方式为:基于所述第一二值图像确定第一基准宽度阈值u1;检测出所述第一二值图像中所有黑色像素区块,记为第一区块集合;根据所述第一基准宽度阈值u1过滤所述第一区块集合中的第一噪音区块,获得所有与所述黑头对应的黑色像素区块,记为第一有效区块集合。其中,该第一基准宽度阈值u1可以设置为第一二值图像在水平方向上的总长W1(或者垂直方向上的总宽)的1/100或者其他合适的倍数。而根据黑头的特性(黑头一般类似圆形)可以定义在水平方向上的长度w1小于u1或者w1大于4*u1,垂直方向上的宽度h1大于3*w1或者h1小于w1/3的黑色像素区块为第一噪音区块,由此,在第一区块集合中过滤掉w1小于u1和/或w1大于4*u1和/或h1大于3*w1和/或h1小于w1/3的黑色像素区块即可获得第一有效区块集合。应当理解的是,此处定义黑头的第一噪音区块为w1小于u1和/或w1大于4*u1和/或h1大于3*w1和/或h1小于w1/3的黑色像素区块仅为了解释本申请实施例并不用于限定本申请实施例,在其他的实际应用中,还可以以其他的条件来定义黑头的第一噪音区块。
350、根据所述第一有效区块集合确定所述黑头的等级。
在本实施例中,由于黑头具有离散分布的特性,因此,可以将用于判定黑头的严重程度的“参考值”设置为:第一有效区块集合中黑色像素区块的个数N,则,根据所述第一有效区块集合确定所述黑头的等级的具体实施方式可以是:计算所述第一有效区块集合中黑色像素区块的个数N;根据所述个数N确定所述黑头的的等级。例如:通过对大量人脸图像进行实验和观察,可以预先设定:当N≤2时,划分为无黑头级别;当20≥N>2时,划分为轻度黑头级别;当50≥N>20时,划分为中度黑头级别;当N>50时,划分为重度黑头级别。则,若针对某一被测者的第一二值图像计算得到的个数N为70个,则可以确定该被测者的黑头等级为重度黑头级别。
或者,在另一些实施例中,也可以以第一有效区块集合中所有黑色像素区块的总面积与该第一二值图像的总面积的比值作为“参考值”,则,所述根据所述第一有效区块集合确定所述黑头的等级的具体实施方式还可以是:计算所
述第一有效区块集合中所有黑色像素区块的总面积,并记为第一总面积;根据所述第一总面积与所述第一二值化图像的总面积的比值确定所述黑头的等级。
当然,在实际应用中,还可以以第一有效区块集合中所有黑色像素区块的个数与第一二值图像的总面积的比值作为“参考值”,进而根据该“参考值”确定黑头等级,此处便不一一列举。
此外,应当理解的是:应用本申请实施例提供的定量检测人脸肤质参量的方法还可以对毛孔、鱼尾纹、痘痘等其他肤质参量进行定量化检测。其与对黑头或者抬头纹的定量检测的不同之处在于:不同的肤质参量对应有不同的目标检测区域,并且在进行二值化处理时采用的参数以及对噪音区块的定义也根据待检测的肤质参量的特性相应变化。例如:对于毛孔的定量检测,其目标检测区域为两颊,在对两颊区域对应的第二原始图像进行自适应阈值二值化处理获取其第二二值图像时,可以选用参数“block Size”=7;参数“C”=2;定义其第二基准宽度阈值u2=1/100W2(其中,W2为该第二二值图像在水平方向上的总长);其第二噪音区块为w2小于u2/2,w2大于3*u2,w2大于3*h2,h2大于3*w2的黑色区块;其“参考值”可以设置为第二有效区块集合中的黑色像素区块的个数。基于实际获取到的人脸图像以及根据上述预设的参数执行上述步骤110至步骤150,即可得到被检测人脸的毛孔的等级。
通过上述技术方案可知,本申请实施例的有益效果在于:本申请实施例提供的定量检测人脸肤质参量的方法通过在获取到人脸图像时,截取所述人脸图像中与待检测的肤质参量对应的图像,并记为原始图像;然后对所述原始图像进行包括二值化处理的图像处理获得二值图像;检测出所述二值图像中所有与所述肤质参量对应的黑色像素区块,记为有效区块集合;最后根据所述有效区块集合确定所述肤质参量的等级,能够更加直观并且更加有针对性地给出待检测的人脸肤质参量的严重程度,增强检测结果的说服力,提升用户体验。
图6是本申请实施例提供的一种定量检测人脸肤质参量的装置的结构示意图,请参阅图6,该装置6包括:
图像获取单元61,用于获取人脸图像;
截取单元62,用于截取所述人脸图像中与待检测的肤质参量对应的区域的图像,并记为原始图像;
图像处理单元63,用于对所述原始图像进行图像处理获得二值图像,其中,所述图像处理包括二值化处理,或者,所述图像处理包括二值化处理和腐蚀处理;
边缘检测单元64,用于检测出所述二值图像中所有与所述肤质参量对应的黑色像素区块,记为有效区块集合;以及,
判定单元65,用于根据所述有效区块集合确定所述肤质参量的等级。
在本申请实施例中,当图像获取单元61获取到人脸图像时,首先在截取单元62中截取所述人脸图像中与待检测的肤质参量对应的区域的图像,并记为原始图像,然后通过图像处理单元63对所述原始图像进行图像处理获得二值图像,通过边缘检测单元84检测出所述二值图像中所有与所述肤质参量对应的黑色像素区块,记为有效区块集合;最后利用判定单元84根据所述有效区块集合确定所述人脸图像的抬头纹等级。
其中,在一些实施例中,边缘检测单元64包括基准宽度阈值确定模块641、检测模块642以及除噪模块643。具体地,当获取到二值图像时,基准宽度阈值确定模块641基于所述二值图像确定基准宽度阈值,检测模块642检测出所述二值图像中所有黑色像素区块,记为区块集合;然后通过除噪模块643根据所述基准宽度阈值过滤所述区块集合中的噪音区块,获得所有与所述肤质参量对应的黑色像素区块,记为有效区块集合。
具体地,在一些实施例中,如:将该装置6应用于定量检测黑头的等级时,截取单元62具体用于:截取所述人脸图像中鼻子区域的图像,并记为第一原始图像;图像处理单元63具体用于:对所述第一原始图像进行上述图像处理获得第一二值图像;边缘检测单元64具体用于:检测出所述第一二值图像中所有与黑头对应的黑色像素区块,记为第一有效区块集合;判定单元65具体用于:根据所述第一有效区块集合确定所述黑头的等级。其中,在边缘检测单元64中,当获取到二值图像时,基准宽度阈值确定模块641基于所述第一二值图像确定
第一基准宽度阈值,检测模块642检测出所述第一二值图像中所有黑色像素区块,记为第一区块集合;然后通过除噪模块643根据所述第一基准宽度阈值过滤所述第一区块集合中的第一噪音区块,获得所有与所述黑头对应的黑色像素区块,记为第一有效区块集合。而判定单元65具体用于:计算所述第一有效区块集合中黑色像素区块的个数;根据所述个数确定所述黑头的等级;或者,计算所述第一有效区块集合中所有黑色像素区块的总面积,并记为第一总面积;根据所述第一总面积与所述第一二值化图像的总面积的比值确定所述黑头的等级。
需要说明的是,由于所述定量检测人脸肤质参量的装置与上述方法实施例中的定量检测人脸肤质参量的方法基于相同的发明构思,因此,上述方法实施例的相应内容同样适用于本装置实施例,此处不再详述。
通过上述技术方案可知,本申请实施例的有益效果在于:本申请实施例提供的定量检测人脸肤质参量的装置通过在图像获取单元61获取到人脸图像时,在截取单元62中截取所述人脸图像中与待检测的肤质参量对应的区域的图像,并记为原始图像;然后通过图像处理单元63对所述原始图像进行包括二值化处理的图像处理获得二值图像;通过边缘检测单元64检测出所述二值图像中所有与所述肤质参量对应的黑色像素区块,记为有效区块集合;最后利用判定单元65根据所述有效区块集合确定所述肤质参量的等级,能够更加直观并且更加有针对性地给出待检测的人脸肤质参量的严重程度,增强检测结果的说服力,提升用户体验。
图7是本申请实施例提供的一种智能终端的结构示意图,该智能终端700可以是任意类型的电子设备,如:手机、平板电脑、美容鉴定仪器等,请参阅图7,该智能终端700包括:
一个或多个处理器710以及存储器720,图7中以一个处理器710为例。
处理器710和存储器720可以通过总线或者其他方式连接,图7中以通过总线连接为例。
存储器720作为一种非暂态计算机可读存储介质,可用于存储非暂态软件
程序、非暂态性计算机可执行程序以及模块,如本申请实施例中的定量检测人脸肤质参量的方法对应的程序指令/模块(例如,附图6所示的图像获取单元61、截取单元62、图像处理单元63、边缘检测单元64以及判定单元65)。处理器710通过运行存储在存储器720中的非暂态软件程序、指令以及模块,从而执行定量检测人脸肤质参量的装置的各种功能应用以及数据处理,即实现上述任一方法实施例的定量检测人脸肤质参量的方法。
存储器720可以包括存储程序区和存储数据区,其中,存储程序区可存储操作***、至少一个功能所需要的应用程序;存储数据区可存储根据定量检测人脸肤质参量的装置的使用所创建的数据等。此外,存储器720可以包括高速随机存取存储器,还可以包括非暂态存储器,例如至少一个磁盘存储器件、闪存器件、或其他非暂态固态存储器件。在一些实施例中,存储器720可选包括相对于处理器710远程设置的存储器,这些远程存储器可以通过网络连接至定量检测人脸肤质参量的装置。上述网络的实例包括但不限于互联网、企业内部网、局域网、移动通信网及其组合。
所述一个或者多个模块存储在所述存储器720中,当被所述一个或者多个处理器710执行时,执行上述任意方法实施例中的定量检测人脸肤质参量的方法,例如,执行以上描述的图1中的方法步骤110至步骤150,图3中的方法步骤310至步骤350,实现图6中的单元61-65的功能。
本申请实施例还提供了一种非暂态计算机可读存储介质,所述非暂态计算机可读存储介质存储有计算机可执行指令,该计算机可执行指令被一个或多个处理器执行,例如图7中的一个处理器710,可使得上述一个或多个处理器执行上述任意方法实施例中的定量检测人脸肤质参量的方法,例如,执行以上描述的图1中的方法步骤110至步骤150,图3中的方法步骤310至步骤350,实现图6中的单元61-65的功能。
以上所描述的装置实施例仅仅是示意性的,其中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模块来实现本实施例方案的目的。
通过以上的实施方式的描述,本领域普通技术人员可以清楚地了解到各实施方式可借助软件加通用硬件平台的方式来实现,当然也可以通过硬件。本领域普通技术人员可以理解实现上述实施例方法中的全部或部分流程是可以通过计算机程序来指令相关的硬件来完成,所述的程序可存储于一计算机可读取存储介质中,该程序在执行时,可包括如上述各方法的实施例的流程。其中,所述的存储介质可为磁碟、光盘、只读存储记忆体(Read-Only Memory,ROM)或随机存储记忆体(Random Access Memory,RAM)等。
上述产品可执行本申请实施例所提供的方法,具备执行方法相应的功能模块和有益效果。未在本实施例中详尽描述的技术细节,可参见本申请实施例所提供的方法。
最后应说明的是:以上实施例仅用以说明本申请的技术方案,而非对其限制;在本申请的思路下,以上实施例或者不同实施例中的技术特征之间也可以进行组合,步骤可以以任意顺序实现,并存在如上所述的本申请的不同方面的许多其它变化,为了简明,它们没有在细节中提供;尽管参照前述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 (10)
- 一种定量检测人脸肤质参量的方法,其特征在于,包括:获取人脸图像;截取所述人脸图像中与待检测的肤质参量对应区域的图像,并记为原始图像;对所述原始图像进行图像处理获得二值图像,所述图像处理包括二值化处理;检测出所述二值图像中所有与所述肤质参量对应的黑色像素区块,记为有效区块集合;根据所述有效区块集合确定所述肤质参量的等级。
- 根据权利要求1所述的方法,其特征在于,所述检测出所述二值图像中所有与所述肤质参量对应的黑色像素区块,记为有效区块集合,包括:基于所述二值图像确定基准宽度阈值;检测出所述二值图像中所有黑色像素区块,记为区块集合;根据所述基准宽度阈值过滤所述区块集合中的噪音区块,获得所有与所述肤质参量对应的黑色像素区块,记为有效区块集合。
- 根据权利要求1所述的方法,其特征在于,所述肤质参量包括黑头,则,所述定量检测人脸肤质参量的方法具体为:截取所述人脸图像中鼻子区域的图像,并记为第一原始图像;对所述第一原始图像进行图像处理获得第一二值图像,所述图像处理包括二值化处理;检测出所述第一二值图像中所有与黑头对应的黑色像素区块,记为第一有效区块集合;根据所述第一有效区块集合确定所述黑头的等级。
- 根据权利要求3所述的方法,其特征在于,所述检测出所述第一二值图像中所有与黑头对应的黑色像素区块,记为第一有效区块集合,包括:基于所述第一二值图像确定第一基准宽度阈值;检测出所述第一二值图像中所有黑色像素区块,记为第一区块集合;根据所述第一基准宽度阈值过滤所述第一区块集合中的第一噪音区块,获得所有与所述黑头对应的黑色像素区块,记为第一有效区块集合。
- 根据权利要求3所述的方法,其特征在于,所述根据所述第一有效区块集合确定所述黑头的等级,包括:计算所述第一有效区块集合中黑色像素区块的个数;根据所述个数确定所述黑头的等级。
- 根据权利要求3所述的方法,其特征在于,所述根据所述第一有效区块集合确定所述黑头的等级,包括:计算所述第一有效区块集合中所有黑色像素区块的总面积,并记为第一总面积;根据所述第一总面积与所述第一二值化图像的总面积的比值确定所述黑头的等级。
- 根据权利要求1-6任一项所述的方法,其特征在于,所述图像处理还包括膨胀腐蚀处理。
- 一种定量检测人脸肤质参量的装置,其特征在于,包括:图像获取单元,用于获取人脸图像;截取单元,用于截取所述人脸图像中与待检测的肤质参量对应区域的图像,并记为原始图像;图像处理单元,用于对所述原始图像进行图像处理获得二值图像,所述图像处理包括二值化处理;边缘检测单元,用于检测出所述二值图像中所有与所述肤质参量对应的黑色像素区块,记为有效区块集合;判定单元,用于根据所述有效区块集合确定所述肤质参量的等级。
- 一种智能终端,其特征在于,包括:至少一个处理器;以及,与所述至少一个处理器通信连接的存储器;其中,所述存储器存储有可被所述至少一个处理器执行的指令,所述指令被所述至少一个处理器执行,以使所述至少一个处理器能够执行如权利要求1-7任一项所述的方法。
- 一种非暂态计算机可读存储介质,其特征在于,所述非暂态计算机可读存储介质存储有计算机可执行指令,所述计算机可执行指令用于使智能终端执行如权利要求1-7任一项所述的方法。
Priority Applications (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201780003821.2A CN108323203A (zh) | 2017-07-17 | 2017-07-17 | 一种定量检测人脸肤质参量的方法、装置和智能终端 |
PCT/CN2017/093190 WO2019014813A1 (zh) | 2017-07-17 | 2017-07-17 | 一种定量检测人脸肤质参量的方法、装置和智能终端 |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
PCT/CN2017/093190 WO2019014813A1 (zh) | 2017-07-17 | 2017-07-17 | 一种定量检测人脸肤质参量的方法、装置和智能终端 |
Publications (1)
Publication Number | Publication Date |
---|---|
WO2019014813A1 true WO2019014813A1 (zh) | 2019-01-24 |
Family
ID=62894884
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
PCT/CN2017/093190 WO2019014813A1 (zh) | 2017-07-17 | 2017-07-17 | 一种定量检测人脸肤质参量的方法、装置和智能终端 |
Country Status (2)
Country | Link |
---|---|
CN (1) | CN108323203A (zh) |
WO (1) | WO2019014813A1 (zh) |
Cited By (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109948476A (zh) * | 2019-03-06 | 2019-06-28 | 南京七奇智能科技有限公司 | 一种基于计算机视觉的人脸皮肤检测***及其实现方法 |
CN113499036A (zh) * | 2021-07-23 | 2021-10-15 | 厦门美图之家科技有限公司 | 皮肤监测方法、装置、电子设备及计算机可读存储介质 |
CN114119598A (zh) * | 2021-12-08 | 2022-03-01 | 林丹柯 | 皮肤血管瘤非接触检测方法、***、计算机设备、及介质 |
CN114973487A (zh) * | 2022-05-13 | 2022-08-30 | 杭州魔点科技有限公司 | 一种基于动态磨皮的人脸检测方法、***、装置和介质 |
Families Citing this family (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109300131A (zh) * | 2018-10-18 | 2019-02-01 | 广州智颜科技有限公司 | 一种图像处理方法、装置、计算机设备和存储介质 |
CN109165700B (zh) * | 2018-10-18 | 2022-09-06 | 广州智颜科技有限公司 | 一种美容液的挤出控制方法、装置及*** |
CN109544515B (zh) * | 2018-11-05 | 2021-08-31 | 深圳和而泰数据资源与云技术有限公司 | 一种趋势确定方法及装置 |
CN109614900B (zh) * | 2018-11-29 | 2021-04-02 | 深圳和而泰数据资源与云技术有限公司 | 图像检测方法及装置 |
CN110334229A (zh) * | 2019-04-30 | 2019-10-15 | 王松年 | 可视化显示方法、设备、***及计算机可读存储介质 |
CN110310252A (zh) * | 2019-04-30 | 2019-10-08 | 深圳市四季宏胜科技有限公司 | 黑头吸除方法、装置及计算机可读存储介质 |
CN110378304B (zh) * | 2019-07-24 | 2021-11-02 | 厦门美图之家科技有限公司 | 皮肤状态检测方法、装置、设备及存储介质 |
CN113255463A (zh) * | 2021-04-30 | 2021-08-13 | 深圳数联天下智能科技有限公司 | 一种显示黑头的方法及相关装置 |
CN113576421A (zh) * | 2021-09-10 | 2021-11-02 | 广州帕克西软件开发有限公司 | 一种智能测肤方法及装置 |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103425985A (zh) * | 2013-08-28 | 2013-12-04 | 山东大学 | 一种人脸抬头纹检测方法 |
CN104299011A (zh) * | 2014-10-13 | 2015-01-21 | 吴亮 | 一种基于人脸图像识别的肤质与皮肤问题识别检测方法 |
CN104732214A (zh) * | 2015-03-24 | 2015-06-24 | 吴亮 | 一种基于人脸图像识别的定量化肤质检测方法 |
CN106388781A (zh) * | 2016-09-29 | 2017-02-15 | 深圳可思美科技有限公司 | 一种皮肤肤色及其色素沉淀情况的检测方法 |
CN106875391A (zh) * | 2017-03-02 | 2017-06-20 | 深圳可思美科技有限公司 | 皮肤图像的识别方法及电子设备 |
Family Cites Families (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP5450560B2 (ja) * | 2011-10-19 | 2014-03-26 | 東芝テック株式会社 | 商品データ処理装置、商品データ処理方法及び制御プログラム |
CN106469302B (zh) * | 2016-09-07 | 2019-05-28 | 成都知识视觉科技有限公司 | 一种基于人工神经网络的人脸肤质检测方法 |
-
2017
- 2017-07-17 CN CN201780003821.2A patent/CN108323203A/zh active Pending
- 2017-07-17 WO PCT/CN2017/093190 patent/WO2019014813A1/zh active Application Filing
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103425985A (zh) * | 2013-08-28 | 2013-12-04 | 山东大学 | 一种人脸抬头纹检测方法 |
CN104299011A (zh) * | 2014-10-13 | 2015-01-21 | 吴亮 | 一种基于人脸图像识别的肤质与皮肤问题识别检测方法 |
CN104732214A (zh) * | 2015-03-24 | 2015-06-24 | 吴亮 | 一种基于人脸图像识别的定量化肤质检测方法 |
CN106388781A (zh) * | 2016-09-29 | 2017-02-15 | 深圳可思美科技有限公司 | 一种皮肤肤色及其色素沉淀情况的检测方法 |
CN106875391A (zh) * | 2017-03-02 | 2017-06-20 | 深圳可思美科技有限公司 | 皮肤图像的识别方法及电子设备 |
Cited By (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109948476A (zh) * | 2019-03-06 | 2019-06-28 | 南京七奇智能科技有限公司 | 一种基于计算机视觉的人脸皮肤检测***及其实现方法 |
CN109948476B (zh) * | 2019-03-06 | 2022-12-13 | 南京七奇智能科技有限公司 | 一种基于计算机视觉的人脸皮肤检测***及其实现方法 |
CN113499036A (zh) * | 2021-07-23 | 2021-10-15 | 厦门美图之家科技有限公司 | 皮肤监测方法、装置、电子设备及计算机可读存储介质 |
CN114119598A (zh) * | 2021-12-08 | 2022-03-01 | 林丹柯 | 皮肤血管瘤非接触检测方法、***、计算机设备、及介质 |
CN114973487A (zh) * | 2022-05-13 | 2022-08-30 | 杭州魔点科技有限公司 | 一种基于动态磨皮的人脸检测方法、***、装置和介质 |
CN114973487B (zh) * | 2022-05-13 | 2024-04-30 | 杭州魔点科技有限公司 | 一种基于动态磨皮的人脸检测方法、***、装置和介质 |
Also Published As
Publication number | Publication date |
---|---|
CN108323203A (zh) | 2018-07-24 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
WO2019014813A1 (zh) | 一种定量检测人脸肤质参量的方法、装置和智能终端 | |
CN108323204B (zh) | 一种检测人脸瑕疵点的方法和智能终端 | |
RU2711050C2 (ru) | Качество изображения и признака, улучшение изображения и выделение признаков для распознавания по сосудам глаза и лицам и объединение информации о сосудах глаза с информацией о лицах и/или частях лиц для биометрических систем | |
CN108369644B (zh) | 一种定量检测人脸抬头纹的方法、智能终端和存储介质 | |
US20160283780A1 (en) | Positioning feature points of human face edge | |
JP2017531883A (ja) | 画像の主要被写体を抽出する方法とシステム | |
CN111144293A (zh) | 带交互式活体检测的人脸身份认证***及其方法 | |
KR20130084437A (ko) | 실시간 얼굴 인식 장치 | |
CN110969046B (zh) | 人脸识别方法、设备及计算机可读存储介质 | |
WO2019000543A1 (zh) | 动态手势识别方法和装置 | |
CN106446921A (zh) | 一种高压输电线障碍物识别方法及装置 | |
CN108121946B (zh) | 一种指纹图像预处理方法及装置 | |
CN113139439A (zh) | 一种基于人脸识别的在线学习专注度评价方法及装置 | |
CN113496215B (zh) | 一种活体人脸检测的方法、装置及电子设备 | |
CN108647605B (zh) | 一种结合全局颜色与局部结构特征的人眼凝视点提取方法 | |
CN111814738A (zh) | 基于人工智能的人脸识别方法、装置、计算机设备及介质 | |
Lin et al. | Face detection based on skin color segmentation and SVM classification | |
WO2022156441A1 (zh) | 活体检测方法及装置、存储介质、终端 | |
CN111612712B (zh) | 一种人脸端正度的确定方法、装置、设备及介质 | |
CN114663930A (zh) | 一种活体检测方法、装置、终端设备及存储介质 | |
CN113837018A (zh) | 一种化妆进度检测方法、装置、设备及存储介质 | |
CN113128372A (zh) | 基于图像处理的黑头识别方法、黑头识别装置及终端设备 | |
CN111898464A (zh) | 一种基于5g网络的人脸识别方法 | |
CN117593679B (zh) | 伪造视频检测方法、装置、电子设备及存储介质 | |
CN117495855B (zh) | 基于图像处理的皮肤缺陷评价方法及*** |
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: 17918211 Country of ref document: EP Kind code of ref document: A1 |
|
NENP | Non-entry into the national phase |
Ref country code: DE |
|
32PN | Ep: public notification in the ep bulletin as address of the adressee cannot be established |
Free format text: NOTING OF LOSS OF RIGHTS PURSUANT TO RULE 112(1) EPC (EPO FORM 1205A DATED 25.05.2020) |
|
122 | Ep: pct application non-entry in european phase |
Ref document number: 17918211 Country of ref document: EP Kind code of ref document: A1 |