CN111724348A - Method for calculating texture attribute of skin image based on texture hillock features - Google Patents
Method for calculating texture attribute of skin image based on texture hillock features Download PDFInfo
- Publication number
- CN111724348A CN111724348A CN202010473377.4A CN202010473377A CN111724348A CN 111724348 A CN111724348 A CN 111724348A CN 202010473377 A CN202010473377 A CN 202010473377A CN 111724348 A CN111724348 A CN 111724348A
- Authority
- CN
- China
- Prior art keywords
- pixel
- value
- image
- point
- texture
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
- 238000000034 method Methods 0.000 title claims abstract description 34
- 238000011156 evaluation Methods 0.000 claims description 13
- 238000012935 Averaging Methods 0.000 claims description 10
- 238000004364 calculation method Methods 0.000 claims description 8
- 238000005070 sampling Methods 0.000 claims description 6
- 238000003384 imaging method Methods 0.000 claims 3
- 210000003717 douglas' pouch Anatomy 0.000 claims 1
- 238000012545 processing Methods 0.000 abstract description 3
- 230000036548 skin texture Effects 0.000 description 11
- 238000001514 detection method Methods 0.000 description 5
- 210000001061 forehead Anatomy 0.000 description 5
- 238000010586 diagram Methods 0.000 description 4
- 238000011496 digital image analysis Methods 0.000 description 3
- 238000004458 analytical method Methods 0.000 description 2
- 230000003247 decreasing effect Effects 0.000 description 2
- 238000005286 illumination Methods 0.000 description 2
- 238000010801 machine learning Methods 0.000 description 2
- 230000000007 visual effect Effects 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 239000011148 porous material Substances 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 230000037303 wrinkles Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/0002—Inspection of images, e.g. flaw detection
- G06T7/0012—Biomedical image inspection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10004—Still image; Photographic image
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10024—Color image
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30004—Biomedical image processing
- G06T2207/30088—Skin; Dermal
Landscapes
- Engineering & Computer Science (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Medical Informatics (AREA)
- Nuclear Medicine, Radiotherapy & Molecular Imaging (AREA)
- Radiology & Medical Imaging (AREA)
- Quality & Reliability (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Image Analysis (AREA)
Abstract
The invention discloses a method for calculating texture attributes of skin images based on texture hillock features, and belongs to the technical field of skin image processing application. Graying the skin image, calculating to obtain a three-valued Cuff image, and calculating to obtain a texture attribute characteristic value for identifying the texture characteristic of the skin image. By adopting the technical scheme provided by the invention, the accuracy of detecting the texture attribute of the skin image is higher, and the method has practical application value.
Description
Technical Field
The invention relates to a skin image texture attribute calculation method, in particular to a method for calculating skin image texture attributes based on texture hillock trench characteristics, and belongs to the technical field of skin image processing application.
Background
Skin surface state detection is one of important indexes for objective evaluation of skin care effect of skin care products, and the most commonly used skin state detection method at present is to shoot skin images at a short distance, statistically analyze the surface characteristics of the skin through an image field algorithm, and calculate the attribute value of the skin characteristics so as to evaluate the skin state.
The skin surface is mainly characterized by texture, wrinkles, pores, roughness, fat points, spots, fluorescent points, glossiness and the like. Skin texture is one of the important attribute indexes of skin surface features, and is also the attribute which is most difficult to identify and calculate.
Skin texture studies currently include two main branches of methods, namely machine learning methods and computer image analysis methods. The machine learning method needs a large number of learning samples, the calculation speed is low, and the accuracy needs to be improved; the difficulty of identifying and calculating the skin texture features by using a computer image analysis method is high, and the accuracy is not high.
Disclosure of Invention
The invention provides a method for calculating texture attributes of skin images based on texture dune characteristics, which is used for performing texture analysis on near-distance forehead and canthus skin digital color images with the same resolution and uniform illumination, and calculating index values of density, depth and width of skin textures according to the texture dune characteristics, wherein the index values can identify the texture characteristics of the skin images. The algorithm for identifying and analyzing the skin texture attributes has high accuracy in detecting the texture attributes of the skin image and has practical application value.
The invention calculates the texture attribute of the skin image based on the texture dune characteristics, which mainly comprises the following steps: (1) reading the skin color image to a computer memory; (2) graying the skin image and calculating an average value; (3) traversing pixels column by column, and calculating to obtain a three-valued hill-groove image img2 (background 127, hill 255, groove 0 value); (4) calculating the characteristic value (width, depth and density) of the texture attribute.
Specifically, the method comprises the following steps:
A. reading the skin color image to a computer memory, and acquiring the RGB value of an image file, wherein the specific operations are as follows:
A1. the only parameter of the algorithm is an image file stored on a hard disk, and each image has a unique ID value;
A2. the image file can be stored on a local, network or other media medium, but the storage path needs to be recorded in a database, and the image file can be accessed according to the unique ID;
A3. reading the RGB value of the image file to a computer memory;
A4. image file formats include, but are not limited to, jpg, bmp, png, etc.;
B. the skin image graying and calculating to obtain a grayscale image mean value specifically comprises the following steps:
B1. graying the color image to obtain a grayed value of each pixel in the image;
averaging the color components of three channels of RGB of each pixel of the read skin color image, and graying the skin image, wherein the calculation formula is as follows:
Vgray＝(Vr+Vg+Vb)/3
wherein Vgray is a grayed value of one pixel, Vr, Vg, Vb are three color components of RGB of the pixel, respectively, and an image after the color image is grayed is a grayscale image img 1;
B2. calculating the average value avg of the gray level image, and calculating the average value of all pixel points of the gray level image img1 to obtain the average value avg of the gray level;
C. calculating to obtain a three-valued hill-furrow image img2 (the background value is 127, the hill value is 255, and the furrow value is 0), which comprises the following steps:
C1. setting the initial values of all pixels of img2 as background color (namely, 127);
C2. each column of pixels of the gray image img1 is processed separately: for each column of pixels on img1, calculating all hill and groove feature points of the column of pixels and identifying on img 2;
the specific operation is as follows:
c2.1 traversing each column of pixels of the grayscale image img 1;
traversing all columns on the gray image img1, wherein each column is independently processed, and executing the steps C2.4 and C2.5;
this treatment was mainly the subject of the present study: forehead or canthus texture, which have obvious transverse texture features, and which are identified as texture at the forehead or canthus;
c2.2, setting a gray value threshold range span of skin background color, and determining the skin background color when the average value avg of the skin gray image is in the gray value threshold range;
the threshold span is determined, the skin texture appears as a bright (hill) and dark (furrow) feature near the mean of the gray values, the threshold range of gray values that differ little from the mean avg of the skin gray image is defined as the skin background color, and if the span is defined as 3, the range [ -3,3] is the skin background color.
C2.3, setting step length, and sampling one pixel point every step pixel points during traversal;
determining step length step, wherein continuous pixel gray scale cannot be strictly increased or decreased in gray scale, and sampling a pixel point at regular intervals in order to reduce gray scale jitter, the method can effectively reduce a plurality of local hillock and trench feature points caused by small jitter of continuous pixel gray scale values when trench hill feature points are subsequently calculated, and if the step length step can be determined to be 5, sampling a pixel point at every 5 pixel points;
c2.4, traversing pixel points p on each column from top to bottom (or from bottom to top) according to a specified step in sequence, and determining that the attribute of the p pixel is one of a hill (255), a trench (0) and a background (127), wherein a column of pixels are represented by the repetition of a section of background pixel, a section of hill pixel, a section of background pixel, a section of trench pixel, a section of background pixel and a section of hill pixel … …, and the specific steps are as follows:
c2.4.1, calculating a difference v between the gray value g of the pixel point p and the average value avg of the gray image, namely v is g-avg;
c2.4.2 determining a starting point of a segment of the dune pixel;
searching a first pixel point p with a v value larger than span, identifying the pixel point p as s1 (the visual expression of the pixel point is lighter than the skin mean value), wherein a point s1 is the starting point of a segment of continuous hill pixels in the row, and assigning a corresponding pixel value of a point s1 on img2 as 255 (hill);
c2.4.3 determining the end point of a segment of a hill of pixels;
starting from a point s1, searching a pixel point p with a first v value smaller than span, wherein the point p at the moment is an end point of a section of a dune pixel, identifying the pixel point p as t1, and identifying that the pixel point on the column starts from a point s1 and a section of the dune pixel ending at a point t1 is already identified;
c2.4.4 finding a segment of trench pixel starting point;
from a point t1, searching a first pixel point p which enables v < -span, identifying the pixel point p as u1, wherein a point u1 is the starting point of a segment of groove pixels in the row, and assigning a corresponding pixel value of a point u1 on img2 as 0 (groove);
c2.4.5 finding a trench pixel end point;
starting from a point u1, searching a pixel point p of a first v > -span, wherein the point p at the moment is an end point of a section of groove pixels, identifying the pixel point p as a point w1, completing the identification of the section of groove pixels from the point u1 to the end point w1, simultaneously calculating the distance w between the two points u1 and w1, taking the distance w as the width of the groove, storing the width w, and solving a point H with the minimum v value in the u1 and w1 sections as the deepest point of the groove; obtaining the mean value d of all pixel gray values on the gray level image img1 corresponding to the u1 and w1 sections, and storing the mean value d as the depth attribute of the section of groove;
c2.5 traversing each row of pixels of the gray level image img1 to obtain all the hill pixel segments and the ditch pixel segments in the gray level image img 1; further calculating the distance s between the deepest points H of two adjacent grooves, and storing the distance s as the distance attribute of the two adjacent grooves;
C3. after all columns of the gray level image img1 are traversed, a three-valued image img2 is obtained: value 127 identifies a background pixel, value
255 identify hill pixels, a value of 0 identifies trench pixels;
D. calculating the characteristic values (including width, depth and density) of the texture attributes, and specifically comprising the following steps:
D1. calculating a texture width attribute value;
step C2.4.5, calculating and storing w values, and averaging all w values to obtain an average value, namely the texture width attribute width;
D2. calculating a texture depth attribute value;
step C2.4.5, calculating and storing d values, averaging all d values to obtain an average value d1, wherein avg-d1-span is a texture Depth attribute Depth, the Depth value is a positive number, and the larger the value is, the deeper the texture is;
D3. calculating texture density attribute values;
step C2.4.6 calculates and stores s values, averages all s values to obtain an average value s1, s1 is the distance between textures, and can identify the density degree of a texture as 1/s1, wherein the larger the density value is, the denser the texture is, and the sparser the texture is otherwise.
Through the steps, the texture attribute of the skin image is calculated based on the texture hill trench characteristics.
Compared with the prior art, the invention has the beneficial effects that:
the invention provides a method for calculating texture attributes of skin images based on texture dune characteristics, which adopts a computer image analysis method to perform texture analysis on near-distance forehead and canthus skin digital color images with the same resolution and uniform illumination, and calculates index values of density, depth and width attributes of skin textures according to the texture dune characteristics, wherein the index values are used for identifying the texture characteristics of the skin images. By adopting the technical scheme provided by the invention, the accuracy of detecting the texture attribute of the skin image is higher, and the method has practical application value.
Drawings
Fig. 1 is a block diagram of an apparatus structure of a skin image texture property evaluation system implemented by an embodiment of the present invention.
Fig. 2 is a flow chart of a method for skin texture evaluation by using a skin image texture attribute evaluation system according to an embodiment of the present invention.
FIG. 3 is a block flow diagram of the skin texture calculation performed by the method of the present invention.
FIG. 4 is a diagram of a row of pixel gray value hill features in a skin image in accordance with an embodiment of the present invention.
Fig. 5 is a screenshot of a partial image interface used by the skin image texture attribute evaluation system to calculate a texture according to the embodiment of the present invention.
Fig. 6 is an interface screenshot of sorting skin images according to a certain attribute value in the system according to the embodiment of the present invention.
Detailed Description
The invention is further described below with reference to the accompanying drawings by way of an embodiment of a set of skin detection systems that has been deployed for implementation. The hardware configuration of the skin image texture attribute evaluation system implemented by deployment is as shown in table 1:
table 1 device configuration of skin image evaluation system according to an embodiment of the present invention
| Name (R) | Device | Number of |
| Skin image collecting device | A microspur skin image collecting device obtains 1000 x 1000 skin color images | 5 |
| Cloud server | Windows server 2012、MySql5.7.16 | 1 |
| Client terminal | Mobile phone Android client | 5 |
The skin image texture attribute evaluation system comprises skin image acquisition hardware equipment, a computer server end and a mobile phone client end, and is shown in figure 1, the skin texture evaluation flow chart of the skin image evaluation system is shown in figure 2, the skin image texture calculation flow chart is shown in figure 3, and the specific implementation steps are as follows:
A. reading the skin color image to a computer memory, wherein the specific contents are as follows:
A1. the only parameter of the algorithm is an image file stored on a hard disk, and each image has a unique ID value;
A2. the image file can be stored on a local, network or other media medium, but the storage path needs to be recorded in a database, and the image file can be accessed according to the unique ID;
A3. reading the RGB value of the image file to a computer memory;
A4. image file formats include, but are not limited to, jpg, bmp, png, etc.;
B. the skin image graying and calculating to obtain a grayscale image mean value specifically comprises the following steps:
B1. graying the color image;
performing simple average method gray scale on the color components of three channels of RGB of each pixel of the read skin color image, and calculating the formula as follows:
Vgray＝(Vr+Vg+Vb)/3
wherein Vgray is a grayed value of a pixel, Vr, Vg and Vb are three RGB color components of the pixel respectively, and an image obtained by graying the color image is a grayscale image img 1;
B2. calculating the average value avg of the gray level image, and calculating the average value of all pixel points of the gray level image img1 to obtain the average value avg of the gray level;
C. calculating to obtain a three-valued hill-furrow image img2 (the background value is 127, the hill value is 255, and the furrow value is 0), which comprises the following steps:
C1. setting the initial values of all pixels of img2 as background color (namely, 127);
C2. each column of pixels of the gray image img1 is processed separately: for each column of pixels on img2, calculating all hill and groove feature points of the column of pixels and identifying on img 2; the specific operation is as follows:
c2.1 traversing each column of pixels of the grayscale image img 1;
all columns are traversed on the gray image img1, each column is operated independently, and the processing is mainly aimed at the research object of the invention: forehead or canthus texture, which have distinct lateral textural features;
c2.2 determines the threshold span, the skin texture appears as a bright (hill) -dark (furrow) feature near the mean of the gray values, the threshold range of gray values that differ little from the mean avg of the skin gray image is defined as the skin background color, and if the span is defined as 3, the range [ -3,3] is the skin background color.
C2.3, determining step length step, wherein continuous pixel gray scale cannot be strictly increased or decreased gradually, in order to reduce gray scale jitter, sampling a pixel point at regular intervals, the method can effectively reduce a plurality of local hillock and trench feature points caused by small jitter of continuous pixel gray scale values when trench-hill feature points are subsequently calculated, and if the step length step can be determined to be 5, sampling a pixel point at every 5 pixel points;
c2.4, traversing pixel points p on each column from top to bottom (or from bottom to top) according to a specified step in sequence, and determining that the attribute of the p pixel is one of a hill (255), a trench (0) and a background (127), wherein a column of pixels are represented by the repetition of a section of background pixel, a section of hill pixel, a section of background pixel, a section of trench pixel, a section of background pixel and a section of hill pixel … …, and the specific steps are as follows:
c2.4.1, calculating a difference v between the gray value g of the pixel point p and the average value avg of the gray image, namely v is g-avg;
c2.4.2 finding a starting point of a segment of a hill pixel;
searching a first pixel point p with a v value larger than span, identifying the pixel point p as s1 (the visual expression of the pixel point is lighter than the skin mean value), wherein a point s1 is the starting point of a segment of continuous dune pixels on the column, and assigning a corresponding pixel value of a point s1 on img2 as 255 (dune);
c2.4.3 finding a segment of a dune pixel end point;
starting from a point s1, searching a pixel point p with a first v value smaller than span, wherein the point p at the moment is an end point of a section of a dune pixel, identifying the pixel point p as t1, and identifying that the pixel point on the column starts from a point s1 and a section of the dune pixel ending at a point t1 is already identified;
c2.4.4 finding a segment of trench pixel starting point;
from the point t1, searching a first pixel point p which enables v < -span, identifying the pixel point p as u1, wherein the point u1 is the starting point of a section of groove pixel on the column, and assigning a corresponding pixel value of the point u1 on img2 as 0 (groove);
c2.4.5 finding a trench pixel end point;
starting from a point u1, searching a pixel point p of a first v > -span, wherein the point p at the moment is an end point of a section of groove pixels, identifying the pixel point p as a point w1, identifying the identification of the section of groove pixels from the point u1 to the end point w1 of the column of pixels, simultaneously calculating the distance w between the two points u1 and w1, using the distance w as the width of the groove and storing the width of the groove, solving a point H with the minimum v value in the u1 and w1 sections as the deepest point of the groove, and using the mean value of all pixel gray values on the gray level image img1 corresponding to the u1 and w1 sections as d as the depth attribute of the section of the groove and storing the mean value;
C2.4.6. repeating the steps C2.4.2-C2.4.5, identifying all the hill pixel segments and the ditch pixel segments on the row until the traversing of the pixel point p on the row is completed, and in the process of traversing the ditch and the hill, calculating the distance s between the deepest points H of two adjacent ditches as the distance attribute of the two adjacent ditches and storing the distance attribute; C3. after all columns of the gray level image img1 are traversed, a three-valued image img2 is obtained: value 127 identifies background pixels, value 255 identifies hill pixels, and value 0 identifies trench pixels;
D. calculating the characteristic values (including width, depth and density) of the texture attributes, and specifically comprising the following steps:
D1. calculating a texture width attribute value;
step C2.4.5, calculating and storing w values, and averaging all w values to obtain an average value, namely the texture width attribute width;
D2. calculating a texture depth attribute value;
step C2.4.5, calculating and storing d values, averaging all d values to obtain an average value d1, wherein avg-d1-span is a texture Depth attribute Depth, the Depth value is a positive number, and the larger the value is, the deeper the texture is;
D. calculating texture density attribute values;
step C2.4.6 calculates and stores s values, averages all s values to obtain an average value s1, s1 is the distance between textures, and can identify the density degree of a texture as 1/s1, wherein the larger the density value is, the denser the texture is, and the sparser the texture is otherwise.
Through the steps, the texture attribute of the skin image is calculated based on the texture hill trench characteristics. FIG. 4 is a diagram illustrating a row of pixel hills, trench identifications and texture density, depth and width attributes.
The embodiment result shows that the method for calculating the texture attribute of the skin image based on the texture dune characteristics, which is realized according to the method, has the advantages of quick detection result and higher accuracy of the detection result. In this embodiment, the texture features of 200 skin images are sorted according to quantitative values of different attributes, specifically, 5 different clients can respectively log in and acquire skin images at a micro distance and then upload the skin images to a computer server, the computer server adopts the method of the present invention to respectively calculate texture density, depth and width attributes of each image, fig. 5 is a partial image therein, which is from a screenshot of a skin image evaluation system interface, and a value (which is processed by a percentile so that a user at a mobile phone client can understand the value meaning) obtained by calculating the corresponding texture attribute is shown in table 2.
TABLE 2 calculation of texture multiple attribute values for skin images using the method of the present invention
In fig. 5, the middle value below each picture is a texture depth value, and the partial images sorted according to the attribute values of the texture obtained by calculation (from the screenshot of the skin image evaluation system interface, the middle value below each picture is a value for texture sorting) have high accuracy in calculating the texture attribute values by the hill-furrow method in the sorting result, and the sorting result is shown in fig. 6.
It is noted that the disclosed embodiments are intended to aid in further understanding of the invention, but those skilled in the art will appreciate that: various substitutions and modifications are possible without departing from the spirit and scope of the invention and appended claims. Therefore, the invention should not be limited to the embodiments disclosed, but the scope of the invention is defined by the appended claims.
Claims (10)
1. A method for calculating the texture attribute of a skin image based on texture dune characteristics comprises the steps of graying the skin image, calculating to obtain a three-valued dune image, and calculating to obtain a texture attribute characteristic value; the method comprises the following steps:
A. reading a color skin image, and acquiring RGB values of the image;
B. graying the skin image and calculating to obtain a gray image mean value, and the specific steps are as follows:
B1. graying the color image to obtain a grayed value of each pixel in the image and a grayscale image img 1;
specifically, averaging color components of three channels of RGB of each pixel of the skin color image, wherein the average value is a gray value of the pixel; the image after the color image graying is a grayscale image img 1;
B2. calculating to obtain a gray level image mean value avg: averaging the grayed values of all the pixel points of the grayscale image img1 to obtain a grayscale image mean value avg;
C. calculating to obtain a three-valued dune ditch image img2, and identifying a pixel point in the image as one of a background, a dune and a ditch, wherein the attribute is that the value of the background is 127, the attribute is that the value of the dune is 255, and the attribute is that the value of the ditch is 0; the method comprises the following operations:
C1. setting the initial values of all pixels of img2 as background color, namely 127;
C2. each column of pixels of the gray image img1 is processed separately: for each column of pixels on img1, calculating all hill and groove feature points of the column of pixels and identifying on img 2; the specific operation is as follows:
C21. setting a skin background color gray value threshold range span, and determining the skin background color when the skin gray image mean value avg is in the gray value threshold range;
C22. setting step, and sampling a pixel point every step pixel points during traversal;
C23. traversing each row of pixels of the gray image img1 according to the sequence and the step length, determining the attribute of a pixel point p, and identifying the pixel point in the image as one of a background, a hill and a ditch; the method specifically comprises the following operations:
c231, calculating the difference between the gray value g of the pixel point p and the average value avg of the gray image to obtain a gray value v, namely v is g-avg;
c232 determines the starting point of a segment of a hill pixel:
searching a first pixel point p with a v value larger than span, identifying the pixel point p as s1, wherein s1 is the starting point of a section of continuous dune pixels in the column, and assigning a corresponding pixel value of s1 points on img2 as 255 to indicate that the pixel attribute is the dune;
c233, determining a segment of a hill pixel end point;
from the s1 point, searching a first pixel point p with a v value smaller than span, wherein the p point at the moment is the end point of a segment of dune pixel, and identifying the pixel point p as t 1; identifying the pixels of the column of pixel points from the point s1 to the point t1 as the hill pixels;
c234 determining the starting point of a segment of groove pixels;
from a point t1, searching a first pixel point p which enables a v value to be smaller than-span, identifying the pixel point p as u1, wherein a point u1 is a starting point of a section of groove pixel in the row, and assigning a corresponding pixel value of a point u1 on img2 as 0, namely identifying the pixel as the groove pixel;
c235 determining an end point of a trench pixel;
starting from a point u1, searching a first pixel point p with a v value larger than-span, wherein the point p at the moment is the end point of the segment groove pixel, and identifying the pixel point p as a point w 1; identifying the pixels of the row of pixel points from the point u1 to the point w1 as groove pixels;
calculating the distance w between the two points of u1 and w1 in the segment of groove pixels, and storing the distance w as the width of the groove;
obtaining a point H with the minimum v value in the section of groove pixels as the deepest point of the groove;
obtaining the mean value d of all pixel gray values on the gray image img1 corresponding to the section of groove pixels, and storing the mean value d as the depth attribute of the section of groove pixels;
C24. traversing each row of pixels of the gray level image img1 to obtain all hill pixel segments and trench pixel segments in the gray level image img 1; further calculating the distance s between the deepest points H of two adjacent grooves, and storing the distance s as the distance attribute of the two adjacent grooves;
C3. after all columns of the gray level image img1 are traversed, a three-valued image img2 is obtained: a pixel point value of 127 in img2 represents a background pixel, a value of 255 represents a hill pixel, and a value of 0 represents a trench pixel;
D. calculating to obtain texture attribute characteristic values including width, depth and density;
through the steps, the texture attribute of the skin image is calculated and obtained based on the texture hillock groove features.
2. The method according to claim 1, wherein the step D of calculating texture attribute feature values comprises:
D1. calculating the texture width attribute value: averaging all the w values to obtain an average value, namely the texture width attribute width;
D2. calculating a texture depth attribute value: averaging all the values of d to obtain an average value d 1;
texture Depth attribute Depth-avg-d 1-span; the Depth value is a positive number, and the larger the value is, the deeper the texture is;
D3. calculating the attribute value of texture density: averaging all the values of s to obtain an average value s1, wherein s1 is the distance between textures;
the density degree of the texture is 1/s 1; the larger the confidence value, the denser the texture and conversely the sparser.
3. The method of claim 1, wherein the skin color image comprises image files in jpg, bmp, png format.
4. The method according to claim 1, wherein the skin image texture attribute calculation formula in step B is as follows:
Vgray＝(Vr+Vg+Vb)/3
wherein Vgray is a gray scale value of one pixel, and Vr, Vg, and Vb are three RGB color components of the pixel, respectively.
5. The method according to claim 1, wherein in step C2.2, the threshold span is 3; the pixel gray value in the range of-3, 3 indicates that the pixel is the skin background color.
6. The method according to claim 1, wherein in step C2.3, the step size step is 5.
7. A skin image texture attribute evaluation system implementing the method of calculating skin image texture attributes based on texture cul-de-sac features as claimed in claims 1-6.
8. The system of claim 7, wherein the system comprises a skin imaging hardware device, a computer server side, and a client side.
9. The system of claim 8, wherein the skin imaging hardware device is a macro skin imaging device for acquiring 1000 x 1000 skin color images.
10. The system of claim 8, wherein the computer server is a cloud server, specifically running a system Windows server 2012 and a database mysql 5.7.16; and/or the client is an Android mobile phone.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010473377.4A CN111724348B (en) | 2020-05-29 | 2020-05-29 | Method for calculating texture attribute of skin image based on texture hill and groove features |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010473377.4A CN111724348B (en) | 2020-05-29 | 2020-05-29 | Method for calculating texture attribute of skin image based on texture hill and groove features |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN111724348A true CN111724348A (en) | 2020-09-29 |
| CN111724348B CN111724348B (en) | 2024-03-26 |
Family
ID=72565298
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010473377.4A Active CN111724348B (en) | 2020-05-29 | 2020-05-29 | Method for calculating texture attribute of skin image based on texture hill and groove features |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111724348B (en) |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107169960A (en) * | 2017-05-15 | 2017-09-15 | 精诚工坊电子集成技术（北京）有限公司 | A kind of skin surface pore size detection method based on color of image space |
| CN107157447A (en) * | 2017-05-15 | 2017-09-15 | 精诚工坊电子集成技术（北京）有限公司 | The detection method of skin surface roughness based on image RGB color |
| CN107564079A (en) * | 2017-08-28 | 2018-01-09 | 北京白鹭时代信息技术有限公司 | A kind of undulating topography generation method |
| CN109949273A (en) * | 2019-02-25 | 2019-06-28 | 北京工商大学 | Skin image texture segmentation method and system based on texture symmetry |
| US20190340774A1 (en) * | 2018-05-02 | 2019-11-07 | Canfield Scientific, Incorporated | Skin assessment using image fusion |
| CN110853009A (en) * | 2019-11-11 | 2020-02-28 | 北京端点医药研究开发有限公司 | Retina pathology image analysis system based on machine learning |
-
2020
- 2020-05-29 CN CN202010473377.4A patent/CN111724348B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107169960A (en) * | 2017-05-15 | 2017-09-15 | 精诚工坊电子集成技术（北京）有限公司 | A kind of skin surface pore size detection method based on color of image space |
| CN107157447A (en) * | 2017-05-15 | 2017-09-15 | 精诚工坊电子集成技术（北京）有限公司 | The detection method of skin surface roughness based on image RGB color |
| CN107564079A (en) * | 2017-08-28 | 2018-01-09 | 北京白鹭时代信息技术有限公司 | A kind of undulating topography generation method |
| US20190340774A1 (en) * | 2018-05-02 | 2019-11-07 | Canfield Scientific, Incorporated | Skin assessment using image fusion |
| CN109949273A (en) * | 2019-02-25 | 2019-06-28 | 北京工商大学 | Skin image texture segmentation method and system based on texture symmetry |
| CN110853009A (en) * | 2019-11-11 | 2020-02-28 | 北京端点医药研究开发有限公司 | Retina pathology image analysis system based on machine learning |
Also Published As
| Publication number | Publication date |
|---|---|
| CN111724348B (en) | 2024-03-26 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108921130B (en) | Video key frame extraction method based on saliency region | |
| CN109522908B (en) | Image significance detection method based on region label fusion | |
| CN111709420B (en) | Text detection method, electronic device and computer readable medium | |
| KR101165359B1 (en) | Apparatus and method for analyzing relation with image and image or video | |
| CN104978578B (en) | Mobile phone photograph text image method for evaluating quality | |
| CN111383244B (en) | Target detection tracking method | |
| KR20010042672A (en) | Signal processing method and video/audio processing device | |
| US10657628B2 (en) | Method of providing a sharpness measure for an image | |
| CN116071271B (en) | Analysis method for image capturing of tablet personal computer | |
| CN109255799B (en) | Target tracking method and system based on spatial adaptive correlation filter | |
| CN115511907B (en) | Scratch detection method for LED screen | |
| CN111709914A (en) | Non-reference image quality evaluation method based on HVS characteristics | |
| CN117541582A (en) | IGBT insulation quality detection method for high-frequency converter | |
| CN115294035A (en) | Bright point positioning method, bright point positioning device, electronic equipment and storage medium | |
| CN112188309B (en) | Shot boundary detection method and device based on accumulated difference degree and singular value decomposition | |
| CN111724348B (en) | Method for calculating texture attribute of skin image based on texture hill and groove features | |
| CN111784636B (en) | Skin image texture quantitative evaluation method based on texture growth | |
| CN117455911A (en) | Production test method and system for network cable crystal head | |
| CN115965613A (en) | Cross-layer connection construction scene crowd counting method based on cavity convolution | |
| Le Moan et al. | Towards exploiting change blindness for image processing | |
| CN112926676B (en) | False target identification method and device and computer equipment | |
| CN114998876A (en) | Sea-land transition phase shale streak layer structure identification method based on rock slice image | |
| CN111861981B (en) | Image color space fat point detection and evaluation method based on dotted line and plane | |
| KR102556350B1 (en) | Method and Apparatus for Calculating Ratio of Lesion Area | |
| KR20220057026A (en) | Fine dust detecting solution and system by computing saturation residual based on AI |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |