WO2019014814A1

WO2019014814A1 - Method for quantitatively detecting forehead wrinkles on human face, and intelligent terminal

Info

Publication number: WO2019014814A1
Application number: PCT/CN2017/093191
Authority: WO
Inventors: 林丽梅
Original assignee: 深圳和而泰智能控制股份有限公司
Priority date: 2017-07-17
Filing date: 2017-07-17
Publication date: 2019-01-24
Also published as: CN108369644A; CN108369644B

Abstract

Provided are a method for quantitatively detecting forehead wrinkles on a human face, and an intelligent terminal. The method comprises: acquiring a facial image; capturing an image, corresponding to a forehead area, in the facial image, and marking the image as a first image; processing the first image to obtain a second image, wherein the image processing comprises binarization processing; acquiring a black pixel block set, corresponding to forehead wrinkles, in the second image, and marking the black pixel block set as a first valid block set; and according to the first valid block set, determining the level of the forehead wrinkles in the facial image. By means of the above-mentioned technical solution, the embodiments of the present application can quantitatively detect the severity of forehead wrinkles in a facial image, and the quantitative detection method is fast and convenient.

Description

Method for quantitatively detecting face headline and intelligent terminal

Technical field

The present application relates to the field of face recognition technologies, and in particular, to a method and a smart terminal for quantitatively detecting face heads up.

Background technique

Face recognition technology is a technology for identifying and comparing facial visual feature information. Currently, the research fields of face recognition technology include: identity recognition, expression recognition and gender recognition. Among them, in the face recognition technology, the face-lifting pattern is often used as an important facial visual feature information because of its different characteristics depending on the age or facial expression of the person.

Currently, face wrinkles can be detected by machine learning classification algorithms such as Bayesian algorithm and neural network algorithm. The specific implementation process of the method for detecting facial wrinkles is: for a given input image that may include a human face, initially determining the approximate position of the face, extracting the outline of the inner face; and then using a sliding small window inside the face Slide on, and use the machine learning model to identify and classify the pictures in the window to determine whether they belong to wrinkles; if they belong to wrinkles, extract them; finally, combine the extracted pictures to obtain grayscale images of wrinkles.

However, in the process of implementing the present application, the inventors have found that at least the following problems exist in the prior art: First, the inner face extracted in the prior art is a portion below the eyebrows, and does not include a forehead portion. Secondly, training machine learning models requires a large amount of tag data, and collecting tag data requires a lot of manpower and financial resources. Moreover, the prior art can only determine whether the block on the face is wrinkles and cannot quantify the severity of wrinkles. Therefore, even if the existing wrinkle detection technology is applied to the detection of face lift, it is not easy and convenient to quantitatively detect the severity of the raised lines in the face image.

Summary of the invention

The embodiment of the present application provides a method for quantitatively detecting a face raising pattern and an intelligent terminal, which can solve the problem that the existing wrinkle detecting technology cannot simply and conveniently perform the severity of the raised head in the face image. Quantitative detection of problems.

In a first aspect, an embodiment of the present application provides a method for quantitatively detecting a face raising pattern, including:

Obtaining a face image;

Intercepting an image corresponding to the forehead region in the face image, and recording the image as a first image;

Performing image processing on the first image to obtain a second image, the image processing including binarization processing;

Obtaining a black pixel block set corresponding to the raised head line in the second image, and recording the first effective block set;

Determining a headline level of the face image according to the first valid block set.

Optionally, the acquiring the black pixel block set corresponding to the raised head in the second image is recorded as the first valid block set, and includes:

Determining a first reference width threshold based on the second image;

Detecting all black pixel blocks in the second image, and recording them as a first block set;

Filtering the noise blocks in the first block set according to the first reference width threshold, and obtaining a black pixel block set corresponding to the headstrip, which is recorded as the first valid block set.

Optionally, the determining, according to the first valid block set, the lookup level of the face image, including:

Calculating a total length of all black pixel blocks in the first active block set in a horizontal direction, and recording the first total length;

And determining a head-up level of the face image according to a ratio of the first total length to a total area of the second image.

Calculating a total area of all black pixel blocks in the first active block set, and recording the first total area;

And determining a headline level of the face image according to a ratio of the first total area to a total area of the second image.

Optionally, the method further includes:

Performing a filtering process on the first image to obtain a third image;

Performing the image processing on the third image to obtain a fourth image;

Obtaining a black pixel block set corresponding to the header in the fourth image, and recording the second effective block set;

And determining, according to the first valid block set, the heading level of the face image, including:

Determining a headline level of the face image according to the first valid block set and the second valid block set.

Optionally, the acquiring the black pixel block set corresponding to the raised head in the fourth image, and recording the second active block set, includes:

Determining a second reference width threshold based on the fourth image;

Detecting all black pixel blocks in the fourth image, and recording them as a second block set;

Filtering the noise blocks in the second block set according to the second reference width threshold to obtain a black pixel block set corresponding to the headstrip, and recording the second effective block set.

Optionally, the determining, according to the first valid block set and the second valid block set, the headline level of the face image, including:

Calculating a total length of all black pixel blocks in the second active block set in the horizontal direction, and recording the second total length;

And determining a head-up level of the face image according to a ratio of the first total length to a total area of the second image and a ratio of the second total length to a total area of the fourth image.

Calculating a total area of all black pixel blocks in the second active block set, and recording it as a second total area;

And determining a headline level of the face image according to a ratio of the first total area to a total area of the second image and a ratio of the second total area to a total area of the fourth image.

Optionally, the image processing further includes: an etching process.

In a second aspect, an embodiment of the present application provides an intelligent terminal, including:

At least one processor; and,

a memory communicatively coupled to the at least one processor; wherein

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 face lift as described above method.

In a third aspect, 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 a face raising pattern.

In a fourth aspect, 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 face lift pattern as described above.

An advantageous effect of the embodiment of the present application is that the method for quantitatively detecting a face-lifting head and the smart terminal provided by the embodiment of the present application intercepts an image corresponding to the forehead region of the face image by acquiring a face image, and recording a first image; then performing image processing including binarization processing on the first image to obtain a second image; and then acquiring a black pixel block set corresponding to the raised pattern in the second image, which is recorded as the first effective Block collection; finally determining the headline level of the face image according to the first effective block set, so that not only accurate recognition of the face image of the face image but also quantification of the severity thereof can be achieved. At the same time, the detection and quantitative detection methods are quick and convenient.

DRAWINGS

The one or more embodiments are exemplified by the accompanying drawings in the accompanying drawings, and FIG. The figures in the drawings do not constitute a scale limitation unless otherwise stated.

1 is a schematic flow chart of a method for quantitatively detecting a face-lifting head line provided by an embodiment of the present application;

2 is a schematic diagram of an example of performing face key point positioning on a face image according to an embodiment of the present application;

3 is a schematic diagram of an example of a first image provided by an embodiment of the present application;

[Correct according to Rule 91 20.09.2017]
4(a) is a schematic diagram showing an example of a second image provided by an embodiment of the present application;
FIG. 4(b) is a schematic diagram showing another example of a second image provided by an embodiment of the present application;

FIG. 5 is a schematic flow chart of another method for quantitatively detecting a face-lifting head according to an embodiment of the present application; FIG.

6 is a schematic diagram of an example of a third image provided by an embodiment of the present application;

[Correct according to Rule 91 20.09.2017]
FIG. 7(a) is a schematic diagram showing an example of a fourth image provided by an embodiment of the present application;
FIG. 7(b) is a schematic diagram showing another example of a fourth image provided by an embodiment of the present application;

FIG. 8 is a schematic structural diagram of an apparatus for quantitatively detecting a face raising pattern according to an embodiment of the present application; FIG.

FIG. 9 is a schematic structural diagram of an intelligent terminal according to an embodiment of the present application.

Detailed ways

In order to make the objects, technical solutions, and advantages of the present application more comprehensible, the present application will be further described in detail below with reference to the accompanying drawings and embodiments. It is understood that the specific embodiments described herein are merely illustrative of the application and are not intended to be limiting.

It should be noted that, if there is no conflict, the various features in the embodiments of the present application may be combined with each other, and are all within the protection scope of the present application. In addition, although the functional module partitioning is performed in the device schematic, the logical sequence is shown in the flowchart, but in some cases, the illustrated may be performed in a different manner from the modules in the device, or in the order in the flowchart. Or the steps described. Furthermore, the words "first", "second", "third", "fourth" and the like used in the present application do not limit the data and the order of execution, but only the same or similar items having substantially the same function and function. distinguish.

The face-lifting pattern refers to the wrinkles on the forehead of the face. In people's ordinary facial expressions, it will not help Autonomously raises the eyebrows. In the long run, it will reduce and damage the muscles of the forehead. The elasticity of the subcutaneous fibrous tissue will gradually decrease, and the eyebrows will be habitually left behind when pressed into the forehead skin. Forming a face-lifting pattern, as the number of times the eyebrows are squeezed to the forehead skin increases, the number of face-lifting lines formed gradually increases, and gradually tends to become stubborn true wrinkles. At present, with the development of face recognition technology, the detection of face headline is also put forward higher requirements, and the existing techniques for detecting face wrinkles mostly use machine learning classification algorithm to detect wrinkles, and the detection method is more It is cumbersome, costly, and it is impossible to quantify the severity of the raised lines.

Based on this, the embodiment of the present application provides a method and a smart terminal for quantitatively detecting a face raising pattern. The method for quantitatively detecting a face raising pattern first extracts an image of a forehead area of a face from a face image, and then performs a series of image processing on the image of the forehead area to extract a texture feature of the raised head, and further Quantitatively recognizing the headline level of the face image, it is possible to train the machine learning model without a large number of wrinkled label pictures, reduce the complexity of the head-up detection, and quantify the severity of the head-up pattern. Identification.

The method for quantitatively detecting face headline and the intelligent terminal provided by the embodiments of the present application are applicable to any technical field related to face recognition, and are particularly suitable for age identification and plastic surgery.

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 authentication machine, a personal computer, a tablet computer, a smart phone, a terminal server, 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.

Specifically, the embodiments of the present application are further described below in conjunction with the accompanying drawings.

Embodiment 1

1 is a schematic flow chart of a method for quantitatively detecting a face-lifting head line according to an embodiment of the present application. Referring to FIG. 1, the method includes:

110. Obtain a face image.

In the present embodiment, 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.

In the present embodiment, when the face facet detection command is received, 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. . For different application scenarios, different ways of acquiring face images can be selected. For example, for the application scenario of the beauty appraisal, the manner of collecting the face image of the detected person in real time may be selected to determine the current level of the raised face of the detected person; and for the application scenario of the age identification, for the acquired person The real-time requirements of the face image are not high, so it is possible to directly use the existing face image for detection.

120. Intercept an image corresponding to the forehead region in the face image, and record the image as a first image.

In the present embodiment, the "forehead region" refers to one of the portions of the face below the hairline above the hairline, which may be the entire region of the portion or a portion of the portion.

Since the head-up pattern is a wrinkle of the forehead portion of the face, in the embodiment, after the face image is acquired, the forehead region in the face image is first intercepted as the target region for the head-up detection, and the intercepted region is intercepted. The image is recorded as a first image to reduce the amount of data processing during subsequent image processing. Among them, since each person's face may have a difference, in different face images, the area corresponding to the area below the hairline above the hairline may be different. Corresponding to different practical application requirements, the specific implementation manner of intercepting the image corresponding to the forehead region in the face image may also be different.

In some embodiments, in order to obtain a more accurate headline level detection result, the "forehead area" includes all areas of the part below the hairline above the eyebrow in the face image, and then the forehead in the face image is intercepted The specific implementation manner of the image corresponding to the region may be: firstly, the method for locating the key point of the face is used to detect the highest point of the eyebrow of the face and the position of the hairline, and then the horizontal line of the line connecting the highest point of the eyebrows is taken out. An image corresponding to an area between the hairline and recorded as a first image. In this embodiment, for different face images, the shape of the area of the first image that is cut out may be Can vary.

In still other embodiments, in order to improve the efficiency of capturing the first image, the "forehead region" is a portion of the portion of the face image below the hairline above the hairline, which is an area set according to an empirical value. The range does not change with the change of the face image. In this embodiment, the specific embodiment of capturing the image corresponding to the forehead region in the face image may be: first acquiring the coordinate parameter of the eyebrow key point in the face image; and then determining the location based on the coordinate parameter The forehead area of the face image is described; finally, the image corresponding to the forehead area is intercepted and recorded as the first image. The method for obtaining the coordinate parameters of the key points of the eyebrows in the face image may be: first, using a third-party toolkit, such as: dlib, performing face key points on the face image (eg, eyebrow key points, eyes) Positioning key points, facial contour key points, mouth key points, etc., and then selecting preset eyebrow key points and determining coordinate parameters thereof, the preset eyebrow key points may include one or more; or, The key points of the eyebrows preset in the face image are directly extracted, and the coordinate parameters thereof are obtained. For example: the image obtained by using the third-party toolkit dlib to locate the key points of the face is shown in Figure 2. The key points of the eyebrows include the key points 18-26; it can be pre-specified: the coordinates of the key point 19 on the x-axis The intermediate value of the key point 20 in the x-axis coordinate is taken as x1, the key value of the key point 25 on the x-axis and the key point 26 in the x-axis coordinate is taken as x2, and the y-axis coordinate of the key point 20 is taken as y2, The y-axis coordinate value y2+(x2-x1)/3 of the key point 20 is taken as y1, and the area enclosed by the coordinate points (x1, y1), (x1, y2), (x2, y1), and (x2, y2) A is the "forehead area"; then, after obtaining the face image, the coordinate parameters of the key points 19, 20, 25, and 26 in the face image are first acquired, and then the area shown in FIG. 2 is determined based on the coordinate parameters. A is the forehead area, and finally the image corresponding to the area A is cut out (as shown in FIG. 3) and recorded as the first image. It should be understood that, in practical applications, coordinate parameters of other eyebrow key points (such as: 21, 23, 26, etc.) may also be selected as criteria for dividing the forehead area; or, other division methods may be used (eg, specifying y1) = y 2+ 。。。。。。。。。。。。。。。。。。。。。。。。。。。。。。。。。。。。。。。。。。。。。 In this embodiment, the coordinate area of the eyebrow key point can be obtained to determine the forehead area of the face image, which improves the efficiency of intercepting the first image.

130. Perform image processing on the first image to obtain a second image, where the image processing includes binary values. Processing.

In the embodiment, 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 "second image" refers to a binarized image of the first image obtained after performing image processing including binarization processing on the first image. In the second image, the pixel block having the wrinkle texture feature appears as a black pixel block, and the pixel block in the forehead region that does not have the wrinkle texture feature appears as a white pixel block.

In the present embodiment, the texture feature of the wrinkles in the first image (ie, the forehead area of the face) is mapped by binarizing the first image. Generally, the binarization processing of the image may be performed by dividing the image into N windows according to a certain rule, and then dividing the pixels in the window according to a uniform threshold T for each of the N windows. For the two parts, binarization is performed. However, it is difficult to achieve an ideal segmentation effect only by setting a fixed threshold. Therefore, in the present embodiment, the first image is binarized by adaptive threshold binarization processing, which is based on the neighborhood block of the pixel. The pixel value distribution determines the binarization threshold at the pixel location, and a better segmentation effect can be obtained. Specifically, in this embodiment, the Adaptive Threshold function of OpenCV may be invoked to perform adaptive threshold binarization processing on the first image as shown in FIG. 3, wherein the calculated neighbors may be set according to the characteristics of the wrinkles and the empirical parameters. The size of the domain block (for example, set "block size" to 9) and the offset value adjustment (for example, setting parameter "C" to 3). The second image obtained by the above processing is as shown in Fig. 4(a).

In addition, since the skin of the human face is unbalanced from a subtle angle, the binarization of the first image leaves some small shapes that are relatively small in shape and discrete in distribution, and which are largely different from the shape of the wrinkles. Noise point. Therefore, in order to remove these fine noise points left after the binarization processing and improve the accuracy of the headline recognition, in some embodiments, the image processing described above includes an etching process in addition to the binarization process. That is, after the binarization processing of the first image, the etching process is also performed. Wherein, the "corrosion treatment" is one of two basic operations of image morphology, and the most basic effect is to corrode the edge of the foreground region in the image, so that the foreground image region becomes smaller, and the background region inside the foreground image is enlarged. . Performing adaptive threshold binarization processing and rot on the first image as shown in FIG. The second image obtained after the image processing of the etch processing is as shown in Fig. 4(b).

140. Acquire a set of black pixel blocks corresponding to the raised lines in the second image, and record the first active block set.

Since the headline in the face image is represented by the black pixel block in the second image, in the embodiment, the “black pixel block corresponding to the headline” may refer to any black pixel in the second image. The block, and the set of all black pixel blocks in the second image is the first active block set. In this case, the specific implementation of the black pixel block set corresponding to the raised head in the second image may be: directly detecting all the black pixel blocks in the second image by using the image edge detection algorithm, The first block set, the first block set is the first valid block set. The size of each black pixel block may be different.

However, in the above-mentioned first block set, there may be a large noise block such as hair, acne, scar, etc. Therefore, in order to improve the accuracy of quantitatively detecting the head-up pattern, in some embodiments, "corresponding to the head-up pattern" The black pixel block "may also refer to a black pixel block in the second image that does not have characteristics such as hair, acne, scars, and the like. In this case, the specific implementation manner of acquiring the black pixel block set corresponding to the raised head in the second image may be: determining a first reference width threshold u based on the second image; detecting the second All black pixel blocks in the image are recorded as a first block set; the noise block in the first block set is filtered according to the first reference width threshold u to obtain a black pixel block set corresponding to the raised head mark , recorded as the first valid block set.

The “first reference width threshold” is a criterion for determining a noise block in the first block set, and the value is related to the length or width of the second image itself. For example, the first reference width threshold u can be set. It is 1/40 or 1/80 of the total length of the second image in the horizontal direction (or the total width in the vertical direction), mainly because the pixels of the photographing terminal or the distance of the photographing distance are different, resulting in different second images. Some face pixels have large face pixels, and different reference width thresholds need to be given for each second image to facilitate subsequent calculation of the length and vertical width of the black pixel block in the horizontal direction, thereby defining A noise block in the first block set.

Wherein, considering that the noise block in the second image is mostly hair, acne or scar, etc., according to its characteristics, the length w in the horizontal direction can be defined to be less than u/2 and/or the width h on the vertical is greater than w. The black pixel block is a noise block, and thus the first effective block set can be obtained by filtering out the black pixel block whose w is smaller than u/2 and/or h is greater than w in the first block set. It should be understood that the black pixel block in which the noise block is w is less than u/2 and/or h is greater than w is only for explaining the embodiment of the present application and is not used to limit the embodiment of the present application, in other practical applications. The noise block can also be defined by other conditions.

150. Determine a heading level of the face image according to the first valid block set.

In the present embodiment, the "head-up level" refers to the severity of the raised pattern of the detected person, which can be divided into multiple levels such as light, medium, and heavy, and each level corresponds to a corresponding one. Reference range. Wherein, the "reference value" is a parameter R that can determine the severity of the headline, and the division standard of the reference value range corresponding to each level can be set by experimenting and observing the headstrip of a large number of face images. . For example, a batch of face images whose statistical degree of wrinkles looks very serious, whose reference value R is greater than or equal to a certain value, for example, R≥a, the headline level of the face image satisfying R≥a is determined as Heavy wrinkle rating.

The specific implementation manner of determining the heading level of the face image according to the first valid block set may be different according to different "reference values" settings. For example, the ratio of the total length of all black pixel blocks in the first active block set in the horizontal direction to the total area of the second image may be used as a “reference value”, and then determined according to the first valid block set. The specific implementation manner of the header level of the face image may be: calculating a total length of all black pixel blocks in the first active block set in the horizontal direction, and recording the first total length; The ratio of the first total length to the total area of the second image determines the heading level of the face image. Alternatively, in other embodiments, the ratio of the total area of all the black pixel blocks in the first active block set to the total area of the second image may be used as the “reference value”. A specific implementation manner of determining a head-up level of the face image may be: calculating a total area of all black pixel blocks in the first valid block set, and recording the first total area; And determining a headline level of the face image according to a ratio of the first total area to a total area of the second image. Of course, in a practical application, the total length of all black pixel blocks in the first active block set in the horizontal direction may be directly used as a “reference value”, and the face image is determined according to the “reference value”. Head up Levels are not listed here.

According to the foregoing technical solution, the method for quantitatively detecting a face raising pattern provided by the embodiment of the present application intercepts an image corresponding to the forehead area of the face image by acquiring a face image. And recorded as a first image; then performing image processing including binarization processing on the first image to obtain a second image; acquiring a black pixel block set corresponding to the raised head in the second image, which is recorded as An effective block set; finally determining a headline level of the face image according to the first valid block set, so as not only accurate recognition of the face image of the face image, but also quantification of the severity thereof The detection method, at the same time, the identification and quantitative detection method is quick and convenient.

Embodiment 2

In addition to the formed wrinkles in the face-lifting pattern, the false wrinkles (ie, fine lines) caused by the dryness of the skin are included, and the fine lines are more easily treated by the later reasonable care than the formed wrinkles. Improved. Therefore, the second embodiment of the present application provides another method for quantitatively detecting the raised head, which is different from the method for quantitatively detecting the raised mark provided in the first embodiment: in the embodiment, the first image is also Performing a filtering process to obtain a third image; and performing image processing on the third image to obtain a fourth image, and acquiring a black pixel block set corresponding to the raised pattern in the fourth image, and recording the second effective block set; Finally, the headline level of the face image is determined according to the first valid block set and the second effective block set.

Specifically, as shown in FIG. 5, the method includes:

510. Obtain a face image.

520. Capture an image corresponding to the forehead region in the face image, and record the image as the first image.

530. Perform image processing on the first image to obtain a second image, where the image processing includes a binarization process.

540. Acquire a set of black pixel blocks corresponding to the raised lines in the second image, and record the first active block set.

In this embodiment, the above steps 510-540 have the same steps as steps 110-140 in the first embodiment. The same technical features, the specific embodiments thereof are also applicable to the present embodiment, and therefore, will not be described again in this embodiment.

550. Perform filtering processing on the first image to obtain a third image.

In the present embodiment, the "filtering process" is to remove the texture features belonging to the fine lines in the first image, leaving only the texture features of the wrinkles. Therefore, it can be considered that the first image is an image including both fine lines and wrinkles, and the third image is an image including only wrinkles.

Since in the image processing, the Gabor function is a linear filter for edge extraction, its frequency and direction expression is similar to that of the human visual system. The study found that Gabor filters are well suited for texture expression and separation. Therefore, in the present embodiment, the first image can be filtered by the Gabor filter to make the image smoother, thereby achieving the effect of filtering out the fine lines and leaving only the wrinkles. For example, a third image obtained by filtering the first image shown in FIG. 3 by using a Gabor filter is as shown in FIG. 6.

560. Perform image processing on the third image to obtain a fourth image.

In this embodiment, the specific implementation manner of performing image processing on the third image to obtain the fourth image may be the same as the specific implementation manner of performing image processing on the first image to obtain the second image in the first embodiment, and is not here. More details. The fourth image obtained by performing adaptive threshold binarization processing on the third image as shown in FIG. 6 is as shown in FIG. 7( a ), and the adaptive threshold binary value is performed on the third image as shown in FIG. 6 . The fourth image obtained after the chemical treatment and the etching treatment is as shown in Fig. 7(b).

570. Acquire a black pixel block set corresponding to the headstrip in the fourth image, and record the second effective block set.

In this embodiment, the specific implementation manner of acquiring the black pixel block set corresponding to the raised head in the fourth image may be obtained by acquiring the black pixel block corresponding to the raised head in the second image in the first embodiment. The specific implementation of the collection is the same. For example, a specific implementation manner of determining a head-up level of the face image according to the second image and the fourth image may be: determining a second reference width threshold based on the fourth image (where the second reference is The width threshold is the same as the first reference width threshold in the first embodiment, and is not specifically described herein; all black pixel blocks in the fourth image are detected as a second block set; The second reference width threshold filters the second block set In the noise block of the combination, a black pixel block set corresponding to the headstrip is obtained, and is recorded as a second effective block set.

In addition, it should be understood that in practical applications, steps 530, 540, 550, 560, and 570 may also be performed in other orders, such as: first performing step 550 and then performing

steps

530 and 540 and finally performing

steps

560 and 570. Or, after performing step 550,

steps

530 and 560 are performed simultaneously, and then steps 540 and 570 are performed simultaneously.

580. Determine a heading level of the face image according to the first valid block set and the second effective block set.

In this embodiment, the head-up image of the face image is first divided into two categories according to the degree of the depth of the head-up pattern: fine lines and wrinkles, and then the level of the head-up pattern is subdivided in each category, and thus, in this embodiment The "head-up pattern" may include, but is not limited to, no fine lines, no wrinkles, mild fine lines, moderate fine lines, severe fine lines, mild wrinkles, moderate wrinkles, and severe wrinkles.

In this embodiment, the first image includes texture features of fine lines and wrinkles, and the third image includes only texture features of the wrinkles, and correspondingly, the second image also includes texture features of fine lines and wrinkles, and the fourth image only Includes texture features for wrinkles. Therefore, the severity of the wrinkles, that is, the level of the raised head under the wrinkle category, may be first determined by the fourth image; when it is determined that the degree of wrinkles of the facial image is light, the fine image of the facial image is further determined by combining the second image. The severity of the headline level under the fine lines category. Similarly, the reference values Ra and Rb and the reference value range corresponding to each level may be set for the second image and the fourth image by experimenting and observing the headstrip of the plurality of face images in advance, and then based on the reference value. And its range determines the level of the raised face of the face image. For example, it is preset that when Ra=0, it is no wrinkle-free wrinkle level; when Rb<b1: if a2>Ra>=a1, it is a light fine grain level, if a3>Ra>=a2, it is medium Degree of fine lines; if Ra>=a3, it is a fine fine grain level; when b2>Rb>=b1, it is a mild wrinkle level; when b3>Rb>=b2, it is a moderate wrinkle level; when Rb> At =b3, it is a heavy wrinkle level. The manner in which the reference values Ra and Rb are set may be the same or different.

Specifically, the specific implementation manner of determining the heading level of the face image according to the first valid block set and the second effective block set is determined by setting manners of reference values Ra and Rb. For example, if the ratio of the total length of all black pixel blocks in the horizontal direction to the total area of the second/fourth image in the first/second effective block set is used as the reference value Ra/Rb, A specific implementation manner of determining a head-up level of the face image by the first valid block set and the second valid block set may be: calculating all black pixel blocks in the first valid block set in a horizontal direction The total length of the upper limit is recorded as the first total length; the total length of all the black pixel blocks in the second effective block set in the horizontal direction is calculated and recorded as the second total length; according to the first total length and the The ratio of the total area of the second image and the ratio of the second total length to the total area of the fourth image determine the heading level of the face image. Or, if the ratio of the total area of all the black pixel blocks in the first/second effective block set to the total area of the second/fourth image is used as the reference value Ra/Rb, according to the first effective area The specific embodiment of the block set and the second active block set determining the heading level of the face image may further be: calculating a total area of all black pixel blocks in the first valid block set, and recording a first total area; calculating a total area of all black pixel blocks in the second active block set, and recording as a second total area; according to the first total area and the total area of the second image The ratio and a ratio of the second total area to the total area of the fourth image determine a headline level of the face image. Of course, in practical applications, the total length of all black pixel blocks in the first/second effective block set in the horizontal direction may be directly used as the reference value Ra/Rb, and then combined with the values of the reference values Ra and Rb. The head-up level of the face image is determined, and will not be enumerated here.

In addition, it should be understood that in the embodiment of the present application, only the image content in the first image is subjected to subsequent image processing without changing its shape and size. Therefore, in the embodiment of the present application, the first image, The second image, the third image, and the fourth image all have the same size.

According to the foregoing technical solution, the beneficial effects of the embodiment of the present application are: obtaining a third image by performing filtering processing on the first image, and performing image processing including binarization processing on the first image and the third image respectively to obtain a second image. The image and the fourth image, and then determining the head-up level of the face image according to the second image and the fourth image, can divide the face-up pattern into two categories of fine lines and wrinkles, further refining the level of the face-lifting pattern. Improves the accuracy of quantitative detection of face-lifting lines.

Embodiment 3

FIG. 8 is a schematic structural diagram of a device for quantitatively detecting a face raising pattern according to an embodiment of the present application. Referring to FIG. 8, the device 8 includes:

An image obtaining unit 81, configured to acquire a face image;

The intercepting unit 82 is configured to intercept an image corresponding to the forehead region in the face image, and record the image as a first image;

The image processing unit 83 is configured to perform image processing on the first image to obtain a second image, where the image processing includes a binarization process, or the image processing includes a binarization process and an erosion process;

as well as,

The edge detecting unit 84 is configured to acquire a black pixel block set corresponding to the headstrip in the second image, and record the first effective block set;

The determining unit 85 is configured to determine a heading level of the face image according to the first valid block set.

In the embodiment of the present application, when the image acquisition unit 81 acquires the face image, the image corresponding to the forehead region of the face image is first intercepted in the clipping unit 82, and recorded as a first image, and then passed through the image processing unit. 83: performing image processing on the first image to obtain a second image, and acquiring, by the edge detecting unit 84, a black pixel block set corresponding to the raised head in the second image, which is recorded as a first effective block set; Unit 84 determines a heading level of the face image based on the first set of valid blocks.

In some embodiments, the intercepting unit 82 is specifically configured to: acquire a coordinate parameter of a key point of the eyebrow in the face image; determine a forehead area of the face image based on the coordinate parameter; and intercept the forehead area Corresponding image and recorded as the first image. In this embodiment, the coordinate area of the eyebrow key point can be obtained to determine the forehead area of the face image, which improves the efficiency of intercepting the first image.

In some embodiments, the edge detecting unit 84 includes a first reference width threshold determining module 841, a first detecting module 842, and a first noise removing module 843. Specifically, when the second image is obtained, the first reference width threshold determining module 841 determines a first reference width threshold based on the second image, and the first detecting module 842 detects all black pixel blocks in the second image, Recorded as the first block set; And then filtering, by the first noise removing module 843, the noise block in the first block set according to the first reference width threshold to obtain a black pixel block set corresponding to the raised head mark, which is recorded as the first valid block set. .

The determining unit 85 is specifically configured to: calculate a total length of all the black pixel blocks in the first active block set in the horizontal direction, and record the first total length; according to the first total length and the second The ratio of the total area of the image determines the heading level of the face image. Or calculating a total area of all the black pixel blocks in the first active block set and recording the first total area; determining the ratio according to the ratio of the first total area to the total area of the second image. The level of the raised face of the face image.

Moreover, in still other embodiments, apparatus 8 further includes filter processing unit 86 for performing a filtering process on the first image to obtain a third image. In this embodiment, the image processing unit 83 is further configured to: perform the binarization processing on the third image to obtain a fourth image. The edge detecting unit 84 further includes a second reference width threshold determining module 844, a second detecting module 845, and a second noise removing module 846; when receiving the fourth image, the second reference width threshold determining module 845 is based on the fourth image Determining a second reference width threshold, the second detecting module 846 detects all black pixel blocks in the fourth image, denoted as a second block set; and then passes the second noise reduction module 847 according to the second reference width threshold The noise block in the second block set is filtered to obtain a black pixel block set corresponding to the headstrip, and is recorded as a second second valid block set. The determining unit 85 is specifically configured to: determine a heading level of the face image according to the first valid block set and the second effective block set. For example, calculating a total length of all black pixel blocks in the first active block set in the horizontal direction, and recording it as a first total length; calculating all black pixel blocks in the second effective block set in a horizontal direction The total length of the upper portion, and is recorded as a second total length; determining the person according to a ratio of the first total length to a total area of the second image and a ratio of the second total length to a total area of the fourth image The level of the face image is raised. Or calculating a total area of all black pixel blocks in the first active block set and recording it as a first total area; calculating a total area of all black pixel blocks in the second effective block set, and recording a second total area; determining a heading of the face image according to a ratio of the first total area to a total area of the second image and a ratio of the second total area to a total area of the fourth image Grain grade.

According to the foregoing technical solution, the device of the present application has the beneficial effect that the device for quantitatively detecting the face-lifting head image provided by the embodiment of the present application intercepts the image in the intercepting unit 82 when acquiring the face image by the image acquiring unit 81. An image corresponding to the forehead area in the face image is recorded as a first image; then the first image is binarized by the image processing unit 83 to obtain a second image; the second image is acquired by the edge detecting unit 84. The black pixel block set corresponding to the headstrip is recorded as the first valid block set; and finally the determining unit 85 determines the heading level of the face image according to the first valid block set, and the pair of people can be realized. The severity of the raised image of the face image is quantitatively detected, and at the same time, the quantitative detection method is quick and convenient. In addition, in some embodiments, the first image is further filtered by the filtering unit 86 to obtain a third image, and the first image and the third image are respectively subjected to image processing in the image processing unit 83 to obtain a second image and a third image. Four images, the first valid block set and the second effective block set are obtained by the edge detecting unit 84, and finally the determining unit 85 determines the heading level of the face image according to the first effective block set and the second effective block set. It can divide the face-up pattern into two categories: fine lines and wrinkles, further refine the level of face-lifting lines, and improve the accuracy of quantitative detection of face-lifting lines.

It should be noted that, since the apparatus for quantitatively detecting the face raising pattern is based on the same inventive concept as the method for quantitatively detecting the face raising in the first embodiment and the second embodiment, the method embodiment 1 and the implementation are The corresponding content of the second example is also applicable to the embodiment of the device, and will not be described in detail herein.

Embodiment 4

9 is a schematic structural diagram of an intelligent terminal according to an embodiment of the present disclosure. The smart terminal 900 can be any type of electronic device, such as a mobile phone, a tablet computer, a beauty authentication device, etc. Referring to FIG. 9, the smart terminal 900 is shown in FIG. include:

One or more processors 910 and memory 920, one processor 910 is taken as an example in FIG.

The processor 910 and the memory 920 may be connected by a bus or other means, as exemplified by a bus connection in FIG.

The memory 920 is a non-transitory computer readable storage medium, and can be used for storing non-transitory software programs, non-transitory computer executable programs, and modules, such as quantitative detecting persons in the embodiments of the present application. A program instruction/module corresponding to the face raising method (for example, the image acquiring unit 81, the clipping unit 82, the image processing unit 83, the edge detecting unit 84, the determining unit 85, and the filter processing unit 86 shown in FIG. 8). The processor 910 performs various functional applications and data processing of the apparatus for quantitatively detecting face-lifting by performing non-transitory software programs, instructions, and modules stored in the memory 920, that is, realizing the quantification of any of the above method embodiments A method of detecting a raised face of a face.

The memory 920 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 store an operation created according to the use of the device for quantitatively detecting the face-lifting of the face Data, etc. Moreover, memory 920 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. In some embodiments, the memory 920 can optionally include a memory remotely located relative to the processor 910 that can be connected via a network to a device that quantitatively detects face lift. 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 920, and when executed by the one or more processors 910, perform a method of quantitatively detecting face lift in any of the above method embodiments, for example, performing the above The method steps 110 through 150 of FIG. 1 are described, and the method steps 510 through 580 of FIG. 5 implement the functions of the units 81-86 of FIG.

Embodiment 5

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 910 in FIG. 9 may cause the one or more processors to perform the method of quantitatively detecting face lift in any of the above method embodiments, for example, performing the method steps 110 to 110 in FIG. 1 described above. 150, method steps 510 through 580 of FIG. 5, implement the functions of units 81-86 of FIG.

The device embodiments described above are merely illustrative, wherein the described as separate components The units may or may not be physically separate, and the components displayed as units may or may not be physical units, that is, may be located in one place, or may 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.

Through the description of the above embodiments, those skilled in the art can clearly understand that the various embodiments can be implemented by means of software plus a general hardware platform, and of course, by hardware. A person skilled in the art can understand that all or part of the process of implementing the above embodiments can be completed by a computer program to instruct related hardware, and the program can be stored in a non-transitory computer readable storage medium. The program, when executed, may include the flow of an embodiment of the methods as described above. The storage medium may be a magnetic disk, an optical disk, a read-only memory (ROM), or a random access memory (RAM).

The above products can perform the methods provided by the embodiments of the present application, and have the corresponding functional modules and beneficial effects of the execution method. For technical details that are not described in detail in this embodiment, reference may be made to the method provided by the embodiments of the present application.

Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present application, and are not limited thereto; in the idea of the present application, the technical features in the above embodiments or different embodiments may also be combined. The steps may be carried out in any order, and there are many other variations of the various aspects of the present application as described above, which are not provided in the details for the sake of brevity; although the present application has been described in detail with reference to the foregoing embodiments, The skilled person should understand that the technical solutions described in the foregoing embodiments may be modified, or some of the technical features may be equivalently replaced; and the modifications or substitutions do not deviate from the embodiments of the present application. The scope of the technical solution.

Claims

A method for quantitatively detecting a face-lifting face, characterized in that it comprises:

Obtaining a face image;

Intercepting an image corresponding to the forehead region in the face image, and recording the image as a first image;

Performing image processing on the first image to obtain a second image, the image processing including binarization processing;

Obtaining a black pixel block set corresponding to the raised head line in the second image, and recording the first effective block set;

Determining a headline level of the face image according to the first valid block set.
The method according to claim 1, wherein the acquiring a set of black pixel blocks corresponding to the headstrip in the second image is recorded as a first valid block set, comprising:

Determining a first reference width threshold based on the second image;

Detecting all black pixel blocks in the second image, and recording them as a first block set;

Filtering the noise blocks in the first block set according to the first reference width threshold, and obtaining a black pixel block set corresponding to the headstrip, which is recorded as the first valid block set.
The method according to claim 1, wherein the determining the heading level of the face image according to the first valid block set comprises:

Calculating a total length of all black pixel blocks in the first active block set in a horizontal direction, and recording the first total length;

And determining a head-up level of the face image according to a ratio of the first total length to a total area of the second image.
The method according to claim 1, wherein the determining the heading level of the face image according to the first valid block set comprises:

Calculating a total area of all black pixel blocks in the first active block set, and recording the first total area;

And determining a headline level of the face image according to a ratio of the first total area to a total area of the second image.
The method of claim 1 further comprising:

Performing a filtering process on the first image to obtain a third image;

Performing the image processing on the third image to obtain a fourth image;

Obtaining a black pixel block set corresponding to the header in the fourth image, and recording the second effective block set;

And determining, according to the first valid block set, the heading level of the face image, including:

Determining a headline level of the face image according to the first valid block set and the second valid block set.
The method according to claim 5, wherein the acquiring a set of black pixel blocks corresponding to the headstrip in the fourth image is recorded as a second active block set, comprising:

Determining a second reference width threshold based on the fourth image;

Detecting all black pixel blocks in the fourth image, and recording them as a second block set;

Filtering the noise blocks in the second block set according to the second reference width threshold to obtain a black pixel block set corresponding to the headstrip, and recording the second effective block set.
The method according to claim 5, wherein the determining the heading level of the face image according to the first active block set and the second active block set comprises:

Calculating a total length of all black pixel blocks in the first active block set in a horizontal direction, and recording the first total length;

Calculating a total length of all black pixel blocks in the second active block set in the horizontal direction, and recording the second total length;

And determining a head-up level of the face image according to a ratio of the first total length to a total area of the second image and a ratio of the second total length to a total area of the fourth image.
The method according to claim 5, wherein the determining the heading level of the face image according to the first active block set and the second active block set comprises:

Calculating a total area of all black pixel blocks in the first active block set, and recording the first total area;

Calculating a total area of all black pixel blocks in the second active block set, and recording it as a second total area;

And determining a headline level of the face image according to a ratio of the first total area to a total area of the second image and a ratio of the second total area to a total area of the fourth image.
The method of any of claims 1-8, wherein the image processing further comprises: an etching process.
An intelligent terminal, comprising:

At least one processor; and,

a memory communicatively coupled to the at least one processor; wherein

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 the method of any of claims 1-9 Methods.
A non-transitory computer readable storage medium, characterized in that the computer readable storage medium stores computer executable instructions for causing a computer to perform as claimed in any one of claims 1-9 The method described.