CN101071455A - Body image automatic standardizing method - Google Patents
Body image automatic standardizing method Download PDFInfo
- Publication number
- CN101071455A CN101071455A CNA2006100791376A CN200610079137A CN101071455A CN 101071455 A CN101071455 A CN 101071455A CN A2006100791376 A CNA2006100791376 A CN A2006100791376A CN 200610079137 A CN200610079137 A CN 200610079137A CN 101071455 A CN101071455 A CN 101071455A
- Authority
- CN
- China
- Prior art keywords
- body image
- humanoid
- skeleton
- corner
- image automatic
- 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
A method of body image standardized automatically, mainly used in medical imaging in the detection software. The process, including a corner stage, as well as pre-feature point of lofting stage in the corner of feature points in the pre-stage, whether it is human-like shadow of closure shape or branching skeleton, at least through images are loaded, the value of gray-scale, corrosion thread, string point eight temporary domain linked establishment of sawtooth phenomenon prevent, least squares method points lower corner to find the angle between adjacent points features seven points an important step in the calculation of lofting stage. The use of computing technology lofting deformation field work, which has standardized body images, the site can be used to correct the geometric standards fell position to the use of human anatomy database ; various body, the shape of the human body image different input geometry types, lofting unified database to link with the human anatomy and the general standard used to facilitate the follow-up of statistical analysis and / database abnormal collection purposes.
Description
Technical field
What the present invention relates to is a kind of body image automatic standardizing method, refers in particular to the human body whole body thermal imagery automatic standardizing method that is applied in the infrared heat image medical treatment detection software.The invention provides a kind of basic technology of interpretation robotization,, provide the auxiliary doctor of information to make accurate judgment, avoid mistaken diagnosis for follow-up abnormal image location and analysis.
Background technology
In the relevant preceding case technology of tw Taiwan patent database, similar this patent uses high image processing technique to locate wanting that the human dissection constructor pays fully, the technology that unique and this patent use have slightly relative have only " can differentiate the dynamic heat power process of mammal and with its be distinguished into the inhomogeneity group method and apparatus (publication number: 200505389) ", this patent is in the temperature anomaly zone of location human body whole body or part, place a grid during photography between infrared image video camera and patient, see through grid and transmit heat energy to patient, be presented on the thermal imagery with active thermal source projection net-like pattern, dissect the target of location with intelligent's body.
This patent is at many shortcomings of aforementioned patent extra means, with the computing machine full automation computing of new thinking in conjunction with body configuration, anatomy and image processing technique, the patient medical video conversion is become the humanoid image of a standard, for the automatic contraposition in follow-up each position.Humanoid with after anatomical database combines in standard, abnormal area and tissue form and link, and the doctor is provided diagnostic message.Therefore this patent is and the diverse logical and implementation method of aforementioned patent.
Summary of the invention
The objective of the invention is to, when using medical infrared heat image human body whole body thermal imagery, with various height, human body whole body thermal imagery that fat or thin, crooked profile is different, drill the computing of translation for how much by mathematics, unified setting-out is humanoid to the standard whole body, is just reaching in order to carrying out follow-up temperature statistics analysis/collection of abnormal temperature distributed data base.
For realizing this purpose, the technical solution used in the present invention is that the body image automatic standardizing flow process mainly comprises calculating of corner unique point and Lofting Calculation two-stage; In the corner unique point in the preposition stage, the setting-out computing of considering subordinate phase very expends system resource, no matter be sealing profile or branch skeleton, all need will put the catena data and delete, but be enough to describe whole profile topology profile to reduce the setting-out operand to lowermost turn corner characteristics point in the phase one.With the infra-red heat image is example, after at first flow process loads the thermal map grey-tone image, promptly carry out the GTG binaryzation, to emphasize to examine the humanoid background reflectance of side person's whole body, then profile utilize the cross shielding once to expand difference set former figure in back produces closed outline or skeleton carries out continuous corrosion to produce the fine rule of humanoid intermediate shaft, and then face the territory rule with eight and set up profile or skeleton point catena in proper order, delete that with least square method feature counts, avoid the zigzag dancing of fine rule pests occurrence rule; In Lofting Calculation in the stage, the corner unique point of humanoid and skeleton is linked to each other, use field deformation (Beier-Neely Field Morphing Algorithm) computing to carry out the standardization setting-out, resulting standardization people body heat shadow sample can drop in the humanoid profile of standard, and meets the requirement on the anatomy.
The maximum characteristics of the present invention are, utilize human body whole body thermal imagery automatic standardizing setting out method of the present invention, can produce the temperature profile at humanoid each position of a standard, not only follow-up temperature statistics analysis and temperature data storehouse are collected and are achieved, also can make medical treatment detect software and possess the data accumulation ability, assist to examine a doctor form the interrogation number the more, temperature just/abnormal area judges more accurately; Image processing technique carries out human body whole body image standardization effort, the image gray scale statistical study after the matching standardization, and computing machine will automatically just precisely pointed out/abnormal area and the unusual organ site of possibility.
All corner unique point geological informations of human body whole body profile and inner skeleton that utilizes carry out the calculating of corner unique point, and utilize field deformation technology (Beier-Neely Field Morphing Algorithm) finish the setting-out of human body whole body image to standard humanoid in, collect in order to follow-up statistical study and human body whole body image database, all belong to the scope of this patent.
Description of drawings
Fig. 1 is a human body whole body image automatic standardizing flow diagram of the present invention;
Fig. 2 is the humanoid setting-out process flow diagram of whole body of the present invention.
Embodiment
Characteristics of the present invention further are described in detail as follows in bar row mode and are illustrated in figure 1 as human body whole body image automatic standardizing flow diagram, are illustrated in figure 2 as the humanoid setting-out process flow diagram of whole body according to shown in the accompanying drawing:
One, corner unique point calculation stages
(1) grey-tone image loads: image format itself is a GTG, by only needing humanoid outline data, carries out the image binary conversion treatment then, to emphasize to examine the humanoid background reflectance of side person's whole body.
(2) GTG binaryzation: with the user from ordering the manual binaryzation of threshold value or with the automatic binaryzation of Otsu method
(3) profile or skeleton graph thinning: if be output as appearance profile, it is humanoid then once to expand with the cross shielding, and the former figure of the humanoid difference set of expansion that then has more a pixel can obtain the appearance profile fine rule; If be output as skeleton branch, then humanoid its remaining trunk and the four limbs skeleton fine rule of making of continuous corrosion.
(4) eight face territory point catena sets up: no matter be appearance profile or skeleton branch, an optional non-background pixel begins scanning, follows the trail of other in proper order and eight faces the non-background pixel in territory, to set up the some catena data structure of profile and skeleton.
(5) crenellated phenomena prevents: for avoiding crenellated phenomena, increase reverse sawtooth (Anti-Zigzag) and calculate.
(6) the least square method reduction of counting: adopt an equation of line reduction to count.
(7) the consecutive point angle is found out the corner unique point: omit tan with 90 * Δ y/ (| Δ x|+| Δ y|) approximate value
-1Calculating, the corner unique point all occurs in the locating of positive and negative transformation of adjacent three two vector angles.
Two, the Lofting Calculation stage
(1) complete input checking: the generation of humanoid intersecting point is necessary, with guarantee after the setting-out each position drop on standard humanoid on; The generation of skeleton branch is an option, to guarantee meeting the anatomy requirement of soft tissue symmetry at trunk and four limbs intermediate shaft after the setting-out.
(2) the corner inspection of counting: in that to carry out good setting-out between examined person's image and the standard video corresponding, examined person counts with the humanoid corner feature of standard must be identical, in like manner the skeleton corner feature of the two count also need the same.
(3) field deformation (Beier-Neely Field Morphing Algorithm) computing: humanoid some catena has only one, but the some catena of skeleton has several branches, so with each bar fine rule is unit, make every thin online unique point of adjacent corners in twos constitute a straight line, cooperate weighted value to drop in the field deformation arithmetical unit.
In sum, the present invention has had the usability and the operation convenience that can detect for industry and medical imaging really, and the statistical study in future than the doctor existing artificially just/abnormal area interpretation mode has more novelty, novelty and simplicity.
The above only is preferred embodiment of the present invention, only is illustrative for the purpose of the present invention, and nonrestrictive.Those skilled in the art is understood, and can carry out many changes to it in the spirit and scope that claim of the present invention limited, revise, even equivalence, but all will fall within the scope of protection of the present invention.
Claims (14)
1. one kind with body image automatic standardizing method, it is characterized in that, it is humanoid to the standard whole body that links with the anatomical data of human body storehouse to utilize the anatomical data of human body storehouse that the input geometric data type of various figures, human body whole body image that profile is different is unified setting-out, and follow-up statistical study and just/collect in the abnormal data storehouse to carry out.
2. body image automatic standardizing method according to claim 1, it is characterized in that, the input geometric data type of described human body whole body image is human body whole body profile or inner skeleton branch, and this setting-out comprises corner unique point calculation stages and Lofting Calculation stage.
3. body image automatic standardizing method according to claim 2 is characterized in that, described corner unique point calculation stages comprises following calculation procedure:
A. grey-tone image loads;
B. GTG binaryzation;
C. profile or skeleton graph thinning;
D. eight face the foundation of territory point catena;
E. crenellated phenomena prevents;
F. the least square method reduction of counting;
G. the consecutive point angle is found out the corner unique point.
4. body image automatic standardizing method according to claim 2 is characterized in that, the described Lofting Calculation stage comprises following calculation step:
A. examine the complete input checking of side person and standard corner unique point;
B. examining side person and the humanoid of standard or skeleton corner feature counts and equates to check; And
C. field deformation computing.
5. body image automatic standardizing method according to claim 3 is characterized in that, described grey-tone image loads and is meant: image format itself is a GTG, owing to only need humanoid outline data, carries out the image binary conversion treatment then.
6. body image automatic standardizing method according to claim 3 is characterized in that, described GTG binaryzation is meant: order the manual binaryzation of threshold value certainly with the user, or with the automatic binaryzation of Otsu method.
7. body image automatic standardizing method according to claim 3, it is characterized in that, described profile or skeleton graph thinning are meant: if be output as appearance profile, it is humanoid then once to expand with the cross shielding, and the former figure of the humanoid difference set of expansion that then has more a pixel can obtain the appearance profile fine rule; If be output as skeleton branch, then humanoid its remaining trunk and the four limbs skeleton fine rule of making of continuous corrosion.
8. body image automatic standardizing method according to claim 3, it is characterized in that, described eight face the foundation of territory point catena is meant: no matter be appearance profile or skeleton branch, an optional non-background pixel begins scanning, follow the trail of other in proper order and eight face the non-background pixel in territory, to set up the some catena data structure of profile and skeleton.
9. body image automatic standardizing method according to claim 3 is characterized in that, described crenellated phenomena prevents to be meant: for avoiding influencing the least square method computing of next procedure, increase reverse sawtooth and calculate.
10. body image automatic standardizing method according to claim 3, it is characterized in that, described least square method is counted to reduce and is meant: adopt an equation of line reduction to count, if the error amount of the middle point group corresponding points line-spacing of catena fragment is then deleted middle point group, is kept 2 points end to end.
11. body image automatic standardizing method according to claim 3, it is characterized in that, described consecutive point angle is found out the corner unique point and is meant: for alleviating the computational burden of single vectorial slope angle, omit tan with 90 * Δ y/ (Δ x|+| Δ y|) approximate value
-1Calculating, carry out secondary and delete, be left a little to be the corner unique point of simplifying most at last, it is poor that wherein Δ x, Δ y are respectively the coordinate of consecutive point.
12. body image automatic standardizing method according to claim 4, it is characterized in that, described complete input checking is meant: standard and tester must produce humanoid intersecting point simultaneously, or option produces the skeleton branch simultaneously, drop in the humanoid profile of standard to guarantee the image after the setting-out.
13. body image automatic standardizing method according to claim 4, it is characterized in that, described corner feature is counted and is checked and to be meant: in that to carry out good setting-out between tester's image and the standard video corresponding, the tester counts with the humanoid corner feature of standard must be identical, and in like manner the skeleton corner feature of the two is counted also needs equally.
14. according to the 3rd described body image automatic standardizing method of claim, wherein the field deformation computing is meant: humanoid some catena has only one, but the some catena of skeleton truly has several branches, so with each bar fine rule is unit, make every thin online unique point of adjacent corners in twos constitute a straight line, drop in the field deformation arithmetical unit of weighted value density function.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CNB2006100791376A CN100517347C (en) | 2006-05-10 | 2006-05-10 | Body image automatic standardizing method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CNB2006100791376A CN100517347C (en) | 2006-05-10 | 2006-05-10 | Body image automatic standardizing method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN101071455A true CN101071455A (en) | 2007-11-14 |
| CN100517347C CN100517347C (en) | 2009-07-22 |
Family
ID=38898676
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CNB2006100791376A Expired - Fee Related CN100517347C (en) | 2006-05-10 | 2006-05-10 | Body image automatic standardizing method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN100517347C (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101840584A (en) * | 2010-04-23 | 2010-09-22 | 浙江工商大学 | Two-dimensional digital character operating method based on interactive display |
| CN102106719B (en) * | 2009-12-24 | 2013-09-25 | 财团法人工业技术研究院 | Health information analysis method adopting image recognition technology and system utilizing method |
| CN105184792A (en) * | 2015-09-06 | 2015-12-23 | 江苏科技大学 | Circular saw web wear extent online measuring method |
| CN106037679A (en) * | 2016-06-29 | 2016-10-26 | 广西全民药业有限责任公司 | Skin temperature acupuncture point location method |
-
2006
- 2006-05-10 CN CNB2006100791376A patent/CN100517347C/en not_active Expired - Fee Related
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102106719B (en) * | 2009-12-24 | 2013-09-25 | 财团法人工业技术研究院 | Health information analysis method adopting image recognition technology and system utilizing method |
| CN101840584A (en) * | 2010-04-23 | 2010-09-22 | 浙江工商大学 | Two-dimensional digital character operating method based on interactive display |
| CN101840584B (en) * | 2010-04-23 | 2012-05-23 | 浙江工商大学 | Two-dimensional digital character operating method based on interactive display |
| CN105184792A (en) * | 2015-09-06 | 2015-12-23 | 江苏科技大学 | Circular saw web wear extent online measuring method |
| CN105184792B (en) * | 2015-09-06 | 2018-01-30 | 江苏科技大学 | A kind of saw blade wear extent On-line Measuring Method |
| CN106037679A (en) * | 2016-06-29 | 2016-10-26 | 广西全民药业有限责任公司 | Skin temperature acupuncture point location method |
Also Published As
| Publication number | Publication date |
|---|---|
| CN100517347C (en) | 2009-07-22 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| He et al. | Synergistic learning of lung lobe segmentation and hierarchical multi-instance classification for automated severity assessment of COVID-19 in CT images | |
| US20180260951A1 (en) | Deep Image-to-Image Recurrent Network with Shape Basis for Automatic Vertebra Labeling in Large-Scale 3D CT Volumes | |
| CN109190540B (en) | Biopsy region prediction method, image recognition device, and storage medium | |
| CN109146988A (en) | Non-fully projection CT image rebuilding method based on VAEGAN | |
| Anand et al. | An automated deep learning models for classification of skin disease using Dermoscopy images: A comprehensive study | |
| CN102934126A (en) | Microcalcification detection and classification in radiographic images | |
| Zhang et al. | Multi-task learning with multi-view weighted fusion attention for artery-specific calcification analysis | |
| CN105938513A (en) | Apparatus and method for providing reliability for computer aided diagnosis | |
| CN109685765A (en) | A kind of X-ray pneumonia prediction of result device based on convolutional neural networks | |
| CN112669254A (en) | Deep learning prostate cancer bone metastasis identification system based on whole-body bone scanning image | |
| CN105139409B (en) | Two dimensional image dividing method based on ant group algorithm | |
| CN111709950A (en) | Mammary gland molybdenum target AI auxiliary screening method | |
| CN113610118B (en) | Glaucoma diagnosis method, device, equipment and method based on multitasking course learning | |
| CN100517347C (en) | Body image automatic standardizing method | |
| CN105389821A (en) | Medical image segmentation method based on combination of cloud module and image segmentation | |
| Zhang et al. | Pulmonary nodule detection based on 3D feature pyramid network with incorporated squeeze‐and‐excitation‐attention mechanism | |
| CN110674726A (en) | Skin disease auxiliary diagnosis method and system based on target detection and transfer learning | |
| CN111402249A (en) | Image evolution analysis method based on deep learning | |
| CN103246888A (en) | System and method for diagnosing lung disease by computer | |
| CN113610746A (en) | Image processing method and device, computer equipment and storage medium | |
| Li et al. | Covid-19 detection in chest radiograph based on yolo v5 | |
| CN113469942B (en) | CT image lesion detection method | |
| CN101071503A (en) | Body local image automatic standardizing method | |
| CN112967260A (en) | Eyeground fluorescence radiography image leakage point detection method based on weak supervision learning | |
| CN104077564B (en) | Iris pupil takes turns area information extracting method to crispaturaing |
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 | ||
| C17 | Cessation of patent right | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20090722 Termination date: 20130510 |