CN1908984A - Coated tongue division and extracting method for colored digital photo of tongue - Google Patents
Coated tongue division and extracting method for colored digital photo of tongue Download PDFInfo
- Publication number
- CN1908984A CN1908984A CN 200610112441 CN200610112441A CN1908984A CN 1908984 A CN1908984 A CN 1908984A CN 200610112441 CN200610112441 CN 200610112441 CN 200610112441 A CN200610112441 A CN 200610112441A CN 1908984 A CN1908984 A CN 1908984A
- Authority
- CN
- China
- Prior art keywords
- threshold
- tongue
- piece
- value
- image
- 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
Images
Landscapes
- Image Processing (AREA)
Abstract
The related color image processing technology for medical diagnosis comprises: pre-processing to remove reflecting points, merging the image to substitute the position with color same as the near position; cutting image with Ostu threshold value to classify the tongue and coated tongue; in color space, cutting the coated tongue with Ostu threshold value to classify the thick coated tongue and the thin; finally, marking with different colors. This invention has fine cutting result benefit to doctor decision.
Description
Technical field
The present invention relates to the division and extracting method of the tongue fur of Evolution of Tongue Inspection of TCM tongue colored digital photo, belong to the field that Color Image Processing and graphical analysis are understood.
Background technology
In the lingual diagnosis of tcm diagnosis method, the thickness of tongue fur, color, area occupied on people's lingual surface, distributing positions etc. are unusual important parameters.In reality diagnosis, the doctor usually judges according to the experience of oneself, thereby can the inconsistent situation of getable diagnosis when causing same patient by several diagnosis very common.The self-dependent judgement diagnosis that each doctor is subjective, so the standard of lingual diagnosis is very fuzzy.In order to address this problem, image processing techniques is applied in the analysis of lingual diagnosis tongue fur, and the parameter of reference is provided during for doctor's lingual diagnosis, helps doctors to determine a standard that objectifies.
Analyzing the proterties of tongue fur with image processing techniques, is the standardized part of Evolution of Tongue Inspection of TCM.Evolution of Tongue Inspection of TCM objectifies from the middle and later periods eighties 20th century and begins one's study.Research mainly concentrates on the calibration of tongue photo color, storage output and the photo of tongue such as is handled at the aspect with modern image processing and analyzing technology.At the analysis cutting techniques of tongue fur, people have adopted a lot of methods to obtain some results.These researchs are based on image processing and analyzing and mode identification technology, and the tongue image of logarithmic code photo is cut apart, and the tongue nature and the tongue fur of difference tongue for the doctor provides quantitative parameter, are put into practice very useful to lingual diagnosis.
But these algorithms also come with some shortcomings, and mainly show as:
(1) algorithm is just studied at single problem, such as only marking off tongue fur from the tongue picture.In fact, the tongue picture of obtaining with digital camera also is not suitable for directly cutting apart, and need carry out pre-service and just can obtain better result.These algorithms generally do not provide pretreated method.
(2) algorithm generally is divided into tongue two parts of tongue nature and tongue fur, ignores the wherein analysis of fine feature.Also have many details that tcm diagnosis is also had effect very much on the tongue fur, for example thickness of tongue fur, scope of thick coating and thin tongue and ratio etc.
(3) algorithm too relies on objective data, and some has ignored the sensation of doctor's subjectivity.
Summary of the invention
At the deficiency that the above-mentioned technology that existed exists, the object of the present invention is to provide a kind of algorithm of can more fit people couple and the visual sense feeling of tongue fur, emphasize this algorithm integrality and application simultaneously and, be applicable to clinical based on this algorithm development application program.
The invention is characterized in that described method is carried out successively according to the following steps in a computing machine:
Step (1) tongue colored digital photo input computing machine, is filled with white the part beyond the tongue, keeps the tongue part;
Step (2), reflective spot is removed in pre-service according to the following steps, and the reflective spot counting in place to go is stored:
The tongue cromogram that step (1) is obtained is mapped as gray-scale map, and by on the tongue that judges the zone in reflective spot:
f(x，y)＞threshold_Flash
Wherein, (x y) is coordinate, and (x is that (span is from 0 to 255 to this gray-scale map for x, the y) gray-scale value of some pixel at coordinate y) to f; It is 160 to 180 that threshold_Flash obtains scope,
F (x, y)＞threshold_Flash person is a reflective spot;
Step (3), the tongue coloured image morsel A behind the reflective spot of place to go, each piece is 8 * 8 or 16 * 16 pixels;
Step (4), the small images A that earlier step (3) is obtained is separated into four sub regions, uses A
1, A
2, A
3, A
4Sign, the size of subregion is 4 * 4 or 8 * 8 pixels, then, again according to following split criterion and merging criterion, judges that described four sub regions should divide by tale quale, still should merge into a small images A:
If: small images A satisfies following any split criterion, and then the A piece that the same color of original usefulness is expressed is divided into four zonule A that color is different
1, A
2, A
3, A
4:
|R(A
k)-R(A)|＞threshold
R
" or "
|G(A
k)-G(A)|＞threshold
G
" or "
|B(A
k)-B(A)|＞threshold
B
Wherein, R (A), G (A), red, green, blue average color in the regional A piece of B (A) representative, span is from 0 to 255,
R (A
k), G (A
k), B (A
k) represent subregion A respectively
kRedness in the piece, green, blue average color, k=1,2,3,4, threshold
R, threshold
G, threshold
B, being expressed as redness, green, blue threshold value successively respectively, span is 10 to 25;
If: regional A piece does not need to divide then by what following merging criterion judged that adjacent four of this A piece comprise the A piece and does not need the zonule X that divides
1, X
2, X
3, X
4Whether should be merged into a big regional X, size is 16 * 16 or 32 * 32:
|R(X
k)-R(X)|＜threshold
R
" and "
|G(X
k)-G(X)|＜threshold
G
" and "
|B(X
k)-B(X)|＜threshold
B
Wherein, R (X
k), G (X
k), B (X
k) represent X successively respectively
kRedness in the piece, green, blue average color, k=1,2,3,4;
R (X), G (X), B (X) represent red in the X piece, green, blue average color successively, and span is from 0 to 255;
Threshold
R, threshold
G, threshold
BWith above-mentioned;
If each piece X
kAverage color and the difference of the average color of four piece X after comprehensive be lower than when setting thresholding, then these four zones are merged, do same zone and handle;
Step (5), the tongue coloured image after the division merging is mapped in the gray space, with the Ostu threshold method this image is cut apart, and distinguishes tongue nature and tongue fur:
If: f (x, y)＞threshold_Coating, then (x y) is tongue nature,
If: f (x, y)≤threshold_Coating, then (x y) is tongue fur,
Wherein, threshold value threshold_Coating tries to achieve according to image automatically with the Ostu method, and can adjust in ± 10 scope;
Step (8), the tongue coloured image that step (5) is obtained is transferred to tone space, and tongue fur is carried out cutting apart the second time, is partitioned into thick coating from tongue fur, and it is as follows that it cuts apart formula:
The expression formula that is transformed into tone space from rgb space is:
H (x, y) be point (x, tone value y), R (x, y), G (x, y), (x y) is respectively the redness of this point, green, blue component to B;
If: H (x, y)＞threshold
H, then (x y) is thick coating,
If: f (x, y)≤threshold
H, then (x y) is thin tongue,
Threshold value threshold
HAutomatically try to achieve according to tone images with the Ostu threshold method;
Step (7), computing machine is represented tongue nature respectively with different colours, thin tongue, the thick coating part is listed the result.These characteristics of putting method are mainly:
1) algorithm is complete.Algorithm comprises the pre-service of comparison film, and quantitative parameter is exported in cutting apart and extracting of tongue fur.
2) tongue fur having been carried out further analysis, tongue fur is divided into thin tongue and two parts of thick coating, especially for the division of thick coating, is not mention in the former algorithm.
The parameter of whole algorithm can be controlled by the doctor, has taken into account doctor's subjective sensation and objective picture two aspects and has analyzed.
The tongue image that several hospitals gather is tested, and the result of this methods analyst has obtained doctor's approval.Sample objects the range of age of the picture of test is extensive, has comprised healthy and ill crowd.
This algorithm and program have been applied to clinical, and be simple to operate, and the result provides important basis for diagnosis.
Description of drawings
Fig. 1: mark picture shining point part, white portion is expressed as shining point among the figure;
Fig. 2: place to go picture shining point;
Fig. 3: cut apart the picture result, above three width of cloth images be followed successively by: original photo image, the image behind the reflective spot of place to go, the result images that classification merges; Below three sub-pictures be followed successively by: cut apart thin tongue result, cut apart the thick coating result, comprehensive segmentation result;
Fig. 4: the parameter result of segmented extraction;
Fig. 5: the controllable adjusting part of doctor;
Fig. 6: programmed algorithm process flow diagram;
Fig. 7: splitting-up method synoptic diagram;
Fig. 8: merge algorithm synoptic diagram.
Embodiment:
This method core is realized by Matlab.Software interface is write with Delphi.The tongue photo is taken with digital camera by traditional Chinese medical science doctor, and patient is flat during shooting stretches out tongue, makes things convenient for the doctor that whole tongue is taken in photo.
This method contains following steps successively:
1) will cut apart good tongue picture input computing machine.
Tongue with photo in the picture partly remains, and other part is filled with white.
2) pre-service: remove reflective spot.
At first coloured image is mapped as gray-scale map, judges with threshold value method whether the zone on the tongue is reflective spot again.Judge that whether this picture element is that the method for reflective spot is:
f(x，y)＞threshold_Flash
Wherein, ((x is that (scope is from 0 to 255 to this gray-scale map for x, the y) gray-scale value of some pixel, and threshold_Flash tries to achieve by experience, and scope is 160 to 180 at coordinate y) to f for x, y) denotation coordination.Determine that default value is 170.The meaning that formula is represented is that pixel is a reflective spot greater than the point of certain setting threshold in the gray-scale map.These points in the gray-scale map are removed, promptly removed reflective spot, in order to avoid have influence on cutting apart and calculating of tongue fur.Simultaneously, the number of pixels of reflective spot is added up, generally speaking, more moistening tongue reflective spot is many.
3) image behind the removal reflective spot is divided joint account, the position more close to color replaces with same color, obtains image more abstract than former figure.
With the image morsel, each piece is 8 * 8 or 16 * 16 pixels earlier.For each piece zonule A, judge whether this piece zone can use same color showing.If not all right, then zonule A is separately become four, with four color showings.The division step as shown in the figure, A is the zonule, A
1, A
2, A
3, A
4Be four should the zone subregion, size is 4 * 4 or 8 * 8.
If regional A satisfies split criterion:
|R(A
k)-R(A)|＞threshold
R
" or "
|G(A
k)-G(A)|＞threshold
G
" or "
|B(A
k)-B(A)|＞threshold
B
Just will be to express for zone that same color is expressed is divided into four different colors originally, diagram be seen Fig. 7.In computing machine, the color of each picture element of a width of cloth coloured image represents that by three bytes three bytes are represented the redness of this picture element, green, blue component respectively.Certain component of each picture element is taken out, form a gray-scale map, promptly can obtain the presentation graphs of this passage according to the arrangement mode of pixel.
Wherein, R (A
k), G (A
k), B (A
k) represent A respectively
kRedness in the piece, green, blue average color, k=1,2,3,4.R (A), G (A), B (A) represent red in the A piece, green, blue average color.R (A), G (A), the span of B (A) is from 0 to 255.Threshold
R, threshold
G, threshold
BBe expressed as the threshold value of redness, green, blueness respectively.According to the experience of our experiment, the span that obtains this threshold value is 10 to 25.The default value value is 20.The meaning of this expression formula is, a rectangular area A is divided into four parts, if the gap that the average color in a zone and whole average color arranged in the rectangle that separates is greater than certain threshold value the time, just separately with this zone, with four colors replacement this part zones.
On the other hand, if four zonules that are the vicinity of sphere of movements for the elephants shape do not need to divide up and down, judging then whether these four zonules can merge becomes a bigger zone.X
1, X
2, X
3, X
4Be four adjacent zonules, size is 8 * 8 or 16 * 16, and X is the big zone after merging, and size is 16 * 16 or 32 * 32.
The merging criterion of four zonules is:
|R(X
k)-R(X)|＜threshold
R
" and "
|G(X
k)-G(X)|＜threshold
G
" and "
|B(X
k)-B(X)|＜threshold
B
Merge synoptic diagram and see Fig. 8.
Wherein, R (X
k), G (X
k), B (X
k) represent X respectively
kRedness in the piece, green, blue average color, k=1,2,3,4.R (X), G (X), B (X) represent red in the X piece, green, blue average color.R (X), G (X), the span of B (X) is from 0 to 255.Threshold
R, threshold
G, threshold
BBe expressed as the threshold value of redness, green, blueness respectively.According to the experience of our experiment, the span that obtains this threshold value is 10 to 25.The default value value is 20.Be lower than a certain threshold value if the meaning is the average color of contiguous four each pieces of zone and the difference of four comprehensive average color, just these four zone merging got up, do same zone and handle.
Tongue body is carried out such processing earlier, is that certain abstract need be carried out in the surface of considering tongue, and some noises are fallen in the place to go; On the other hand tongue nature and tongue fur not only with color light mutually, equally also with its around the color of pixel be relevant.This method can be considered the influence to it of the color of pixel itself and surrounding pixel fully.
4) with the Ostu threshold method picture of classifying after merging is cut apart difference tongue nature and tongue fur part.
Coloured image after the division merging is mapped in the gray space, and division methods is:
f(x，y)＞threshold_Coating
Wherein, threshold_Coating represents threshold value, and (x y) is the coordinate of pixel, and (x y) is gray-scale value to f.Gray-scale value is lower than threshold_Coating's, be divided into this part of tongue fur, and gray-scale value is higher than threshold_Coating, is divided into this part of tongue nature.Threshold_Coating obtains according to image automatically according to the way of Ostu.In addition, can also be near automatic threshold make manual adjustment in positive and negative 10 the scope.
The main idea of Ostu method calculated threshold is: set a threshold value t, with picture according to f (x, y)＞t and f (x, y)≤t is divided into two parts, calculating the inter-class variance of two parts: g=w
0(I
0-I)
2+ w
1(I
1-I)
2Wherein, I
0And I
1The average gray of latter two part is cut apart in expression respectively, and I represents the average gray of entire image, w
0, w
1The expression weight in this program, is selected w
0=w
1=0.5.T from 0 to 255 is traveled through, select to make the t of g maximum as the threshold value of cutting apart.
5) RGB (RGB) image is transferred in H (tone Hue) space, the part secondary splitting to tongue fur is partitioned into thick coating from tongue fur, and other parts are thin tongue.
The formula that is transformed into the H space from rgb space is:
(x y) is the coordinate of point, and (x y) is the tint number of this point to H, and (x, y), (x, y), (x y) is respectively the redness of this point, green, blue component to B to G to R.
The method of dividing thin tongue and thick coating is:
H(x，y)＞threshold
H
Tone value is greater than threshold
HBe thick coating, tone value is less than or equal to threshold
HBe thin tongue.Threshold
HValue still use the Ostu threshold method that H (tone) passage is obtained.Method is: set a threshold value h, with picture according to H (x, y)＞h and H (x, y)≤h is divided into two parts, calculating the inter-class variance of two parts: s=w
0(Hue
0-Hue)
2+ w
1(Hue
1-Hue)
2Wherein, Hue
0And Hue
1The average tone of latter two part is cut apart in expression respectively, and Hue represents the average tone of entire image, w
0, w
1The expression weight in this program, is selected w
0=w
1=0.5.H from 0 to 255 is traveled through, select to make the h of s maximum as the threshold value of cutting apart.
To being cutting apart of thick coating that this algorithm proposes first, starting point is that the tone difference of the tone of thick coating and thin tongue is very big.
6) mark tongue nature respectively with different colors, thin tongue, thick coating part, output result.
When this algorithm of utilization carries out the tongue fur segmented extraction, can be the application program of this algorithm directly with the backstage.Concrete grammar and step method when this application program is used are as follows:
1) start-up routine is clicked " selection image " button, is written into to cut apart good tongue body photo.(see Fig. 1, Fig. 2)
2) after Loading Image, the picture position of " white portion is a shining point " will provide the position of the shining point that program calculates.(see figure 3) is clicked below button " place to go reflective spot ", and what the picture position of " image after the reflective spot of place to go " obtained is to remove reflective spot picture afterwards on the former figure basis.If (see figure 4) is dissatisfied to the number of shining point, adjust the slider bar on the right, can control the number and the scope of reflective spot.
3) behind the reflective spot of place to go, click " separation tongue body and tongue coating " button, at " segmentation result ", " the thin tire result of extraction ", " the thick coating result of extraction ", what show respectively in the picture position of " synthesis result " is the image that obtains after division merges, the thin tongue area image that analysis obtains, thick coating area image and will approach tongue, thick coating and tongue nature are used white respectively, the synthesis result of yellow and red expression.(see figure 5) shows that the thin tongue that calculates accounts for the number percent of tongue area and the number percent (see figure 6) that thick coating accounts for the tongue area simultaneously.
4) if the result is unsatisfied with, at " segmentation result ", " the thin tire result of extraction ", three slider bars on the right of " the thick coating result of extraction " can be regulated the threshold value that the control division merges, and choose the gray threshold and the tone threshold value of choosing thick coating of thin tongue.(see figure 7)
Claims (4)
1. the division and extracting method of the tongue fur of tongue colored digital photo is characterized in that, described method is carried out in a computing machine successively according to the following steps:
Step (1) tongue colored digital photo input computing machine, is filled with white the part beyond the tongue, keeps the tongue part;
Step (2), reflective spot is removed in pre-service according to the following steps, and the reflective spot counting in place to go is stored:
The tongue cromogram that step (1) is obtained is mapped as gray-scale map, and by on the tongue that judges the zone in reflective spot:
F (x, y)＞threshold_Flash; Wherein, (x y) is coordinate, and (x is that (span is from 0 to 255 to this gray-scale map for x, the y) gray-scale value of some pixel at coordinate y) to f; It is 160 to 180 that threshold_Flash obtains scope, and f (x, y)＞threshold_Flash person is a reflective spot;
Step (3), the tongue coloured image morsel A behind the reflective spot of place to go, each piece is 8 * 8 or 16 * 16 pixels;
Step (4), the small images A that earlier step (3) is obtained is separated into four sub regions, uses A
1, A
2, A
3, A
4Sign, the size of subregion is 4 * 4 or 8 * 8 pixels, then, again according to following split criterion and merging criterion, judges that described four sub regions should divide by tale quale, still should merge into a small images A:
If: small images A satisfies following any split criterion, and then the A piece that the same color of original usefulness is expressed is divided into four zonule A that color is different
1, A
2, A
3, A
4:
|R(A
k)-R(A)|＞threshold
R
" or "
|G(A
k)-G(A)|＞threshold
G
" or "
|B(A
k)-B(A)|＞threshold
B
Wherein, R (A), G (A), red, green, blue average color in the regional A piece of B (A) representative, span is from 0 to 255,
R (A
k), G (A
k), B (A
k) represent subregion A respectively
kRedness in the piece, green, blue average color, k=1,2,3,4,
Threshold
R, threshold
G, threshold
B, being expressed as redness, green, blue threshold value successively respectively, span is 10 to 25;
If: regional A piece does not need to divide then by what following merging criterion judged that adjacent four of this A piece comprise the A piece and does not need the zonule X that divides
1, X
2, X
3, X
4Whether should be merged into a big regional X, size is 16 * 16 or 32 * 32:
|R(X
k)-R(X)|＜threshold
R
" and "
|G(X
k)-G(X)|＜threshold
G
" and "
|B(X
k)-B(X)|＜threshold
B
Wherein, R (X
k), G (X
k), B (X
k) represent X successively respectively
kRedness in the piece, green, blue average color, k=1,2,3,4;
R (X), G (X), B (X) represent red in the X piece, green, blue average color successively, and span is from 0 to 255;
Threshold
R, threshod
G, threshold
BWith above-mentioned,
If the difference of the average color of the piece X after the average color of each piece Xk and four are comprehensive is lower than when setting thresholding, then these four zones are merged, do same zone and handle;
Step (5), the tongue coloured image after the division merging is mapped in the gray space, with the Ostu threshold method this image is cut apart, and distinguishes tongue nature and tongue fur:
If: f (x, y)＞threshold_Coating, then (x y) is tongue nature,
If: f (x, y)≤threshold_Coating, then (x y) is tongue fur,
Wherein, threshold value threshold_Coating tries to achieve according to image automatically with the Ostu method, and can adjust in ± 10 scope;
Step (8), the tongue coloured image that step (5) is obtained is transferred to tone space, and tongue fur is carried out cutting apart the second time, is partitioned into thick coating from tongue fur, and it is as follows that it cuts apart formula:
The expression formula that is transformed into tone space from rgb space is:
H (x, y) be point (x, tone value y), R (x, y), G (x, y), (x y) is respectively the redness of this point, green, blue component to B;
If: H (x, y)＞threshold
H, then (x y) is thick coating,
If: f (x, y)≤threshold
H, then (x y) is thin tongue,
Threshold value threshold
HAutomatically try to achieve according to tone images with the Ostu threshold method, and can in ± 10 scope, adjust;
Step (7), computing machine is represented tongue nature respectively with different colours, thin tongue, the thick coating part is listed the result.
2. the tongue fur division and extracting method of the colored digital photo of tongue according to claim 1 is characterized in that, described threshold_Coating value is tried to achieve according to following steps:
Step (a) is got a value t from 0 to 255 and is made threshold value, image be divided into f (x, y)＞t and f (x, y)≤two parts of t; Step (b) is calculated as follows the inter-class variance of described two parts: g=w
0(I
0-I)
2+ w
1(I
1-I)
2, wherein, I
0And I
1The average gray of latter two part is cut apart in expression respectively, and I represents the average gray of entire image, w
0, w
1The expression weight is got w
0=w
1=0.5;
Step (c) travels through t from 0 to 255, selects to make the t of g maximum as the threshold value threshold_Coating of cutting apart.
3. the tongue fur division and extracting method of tongue coloured image digital photograph according to claim 1 is characterized in that described threshold
HValue is tried to achieve according to following steps:
Step (A) is got a value h from 0 to 255 and is made threshold value, image be divided into H (x, y)＞h and H (x, y)≤two parts of h;
Step (B) is calculated as follows the inter-class variance of described two parts: s=w
0(Hue
0-Hue)
2+ w
1(Hue
1-Hue)
2, wherein, Hue
0And Hue
1The average tone of latter two part is cut apart in expression respectively, and Hue represents the average gray of entire image, w
0, w
1The expression weight is got w
0=w
1=0.5;
Step (C) travels through h from 0 to 255, selects to make the h of s maximum as the threshold value threshold of cutting apart
H
4. the tongue condition dividing method of tongue colored digital photo according to claim 1 is characterized in that threshold
R, threshold
G, threshold
BValue be 20.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CNB2006101124416A CN100405402C (en) | 2006-08-18 | 2006-08-18 | Coated tongue division and extracting method for colored digital photo of tongue |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CNB2006101124416A CN100405402C (en) | 2006-08-18 | 2006-08-18 | Coated tongue division and extracting method for colored digital photo of tongue |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN1908984A true CN1908984A (en) | 2007-02-07 |
| CN100405402C CN100405402C (en) | 2008-07-23 |
Family
ID=37700089
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CNB2006101124416A Expired - Fee Related CN100405402C (en) | 2006-08-18 | 2006-08-18 | Coated tongue division and extracting method for colored digital photo of tongue |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN100405402C (en) |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102147921A (en) * | 2011-04-08 | 2011-08-10 | 浙江理工大学 | Graph theory-based Chinese medicinal tongue nature and tongue coat separation algorithm |
| CN102426583A (en) * | 2011-10-10 | 2012-04-25 | 北京工业大学 | Chinese medicine tongue manifestation retrieval method based on image content analysis |
| CN105030200A (en) * | 2015-07-24 | 2015-11-11 | 天津大学 | Traditional Chinese medicine tongue body and tongue coating separation method based on green monochromatic source |
| CN105678758A (en) * | 2015-12-31 | 2016-06-15 | 天津慧医谷科技有限公司 | Image feature automatic identifying and extracting method |
| CN110648336A (en) * | 2019-09-23 | 2020-01-03 | 无限极（中国）有限公司 | Method and device for dividing tongue texture and tongue coating |
| CN113706515A (en) * | 2021-08-31 | 2021-11-26 | 平安科技（深圳）有限公司 | Tongue image abnormality determination method, tongue image abnormality determination device, computer device, and storage medium |
| CN114511567A (en) * | 2022-04-20 | 2022-05-17 | 天中依脉(天津)智能科技有限公司 | Tongue body and tongue coating image identification and separation method |
Family Cites Families (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP3486615B2 (en) * | 2001-05-22 | 2004-01-13 | 畦元 将吾 | Region extraction method for medical images |
| CN1367387A (en) * | 2002-03-25 | 2002-09-04 | 北京工业大学 | Chinese medicine tongue fur humidity computer analysis method based on two spectral reflection models |
| CN1169079C (en) * | 2002-03-25 | 2004-09-29 | 北京工业大学 | Chinese medicine tongue image computer cutting based on sample strip Snakes model |
| CN1162798C (en) * | 2002-03-25 | 2004-08-18 | 北京工业大学 | Chinese medicine tongue colour, fur colour and tongue fur thickness analysis method based on multiclass support vector machine |
| KR100506085B1 (en) * | 2002-12-28 | 2005-08-05 | 삼성전자주식회사 | Apparatus for processing tongue image and health care service apparatus using tongue image |
-
2006
- 2006-08-18 CN CNB2006101124416A patent/CN100405402C/en not_active Expired - Fee Related
Cited By (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102147921A (en) * | 2011-04-08 | 2011-08-10 | 浙江理工大学 | Graph theory-based Chinese medicinal tongue nature and tongue coat separation algorithm |
| CN102147921B (en) * | 2011-04-08 | 2012-10-03 | 浙江理工大学 | Graph theory-based Chinese medicinal tongue nature and tongue coat separation algorithm |
| CN102426583A (en) * | 2011-10-10 | 2012-04-25 | 北京工业大学 | Chinese medicine tongue manifestation retrieval method based on image content analysis |
| CN105030200A (en) * | 2015-07-24 | 2015-11-11 | 天津大学 | Traditional Chinese medicine tongue body and tongue coating separation method based on green monochromatic source |
| CN105678758A (en) * | 2015-12-31 | 2016-06-15 | 天津慧医谷科技有限公司 | Image feature automatic identifying and extracting method |
| CN110648336A (en) * | 2019-09-23 | 2020-01-03 | 无限极（中国）有限公司 | Method and device for dividing tongue texture and tongue coating |
| CN110648336B (en) * | 2019-09-23 | 2022-07-08 | 无限极（中国）有限公司 | Method and device for dividing tongue texture and tongue coating |
| CN113706515A (en) * | 2021-08-31 | 2021-11-26 | 平安科技（深圳）有限公司 | Tongue image abnormality determination method, tongue image abnormality determination device, computer device, and storage medium |
| CN113706515B (en) * | 2021-08-31 | 2023-07-18 | 平安科技（深圳）有限公司 | Tongue image anomaly determination method, tongue image anomaly determination device, computer equipment and storage medium |
| CN114511567A (en) * | 2022-04-20 | 2022-05-17 | 天中依脉(天津)智能科技有限公司 | Tongue body and tongue coating image identification and separation method |
Also Published As
| Publication number | Publication date |
|---|---|
| CN100405402C (en) | 2008-07-23 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN1908984A (en) | Coated tongue division and extracting method for colored digital photo of tongue | |
| US10445557B2 (en) | Learning pixel visual context from object characteristics to generate rich semantic images | |
| CN1162798C (en) | Chinese medicine tongue colour, fur colour and tongue fur thickness analysis method based on multiclass support vector machine | |
| CN108876743B (en) | Image rapid defogging method, system, terminal and storage medium | |
| US20030007687A1 (en) | Correction of "red-eye" effects in images | |
| CN1475969A (en) | Method and system for intensify human image pattern | |
| CN100345159C (en) | Image editing device, image cutting method and program | |
| CN1260680C (en) | Method and appts. for digital image segmentation | |
| CN1761205A (en) | System for detecting eroticism and unhealthy images on network based on content | |
| CN101080024A (en) | A method of processing pseudo color of medical gray image signals | |
| CN101079952A (en) | Image processing method and image processing apparatus | |
| CN101079954A (en) | Method and device for realizing white balance correction | |
| CN1741038A (en) | Mid-face location detecting apparatus, method and program | |
| CN1700238A (en) | Method for dividing human body skin area from color digital images and video graphs | |
| CN1991865A (en) | Device, method, program and media for extracting text from document image having complex background | |
| CN1798237A (en) | Method of and system for image processing and computer program | |
| CN1612164A (en) | Red eye correction method and apparatus for implementing said method | |
| CN1910613A (en) | Method for extracting person candidate area in image, person candidate area extraction system, person candidate area extraction program, method for judging top and bottom of person image, system for j | |
| CN1139898C (en) | Cornea focus image cutting method based on k-mean cluster and information amalgamation | |
| CN1667355A (en) | Image recognition method and image recognition apparatus | |
| CN101075872A (en) | Registration device, collation device, extraction method, and program | |
| CN1195284C (en) | Image processing equipment | |
| CN1920854A (en) | Tongue division and extracting method for tongue diagnosis used colored digital photo of tongue in Chinese traditional medicine | |
| CN1737853A (en) | Method for making and playing interactive cartoon program | |
| Reddy et al. | Retinal fundus image enhancement using piecewise gamma corrected dominant orientation based histogram equalization |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20080723 Termination date: 20170818 |
|
| CF01 | Termination of patent right due to non-payment of annual fee |